Add JSON prefix O to V

Full_text_JSON/prefixO/json/O00/O00-1001.json

@@ -0,0 +1,543 @@
+{
+    "paper_id": "O00-1001",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:58:59.419741Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-1001",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "\u82e5\u4e09\u5b57\u7d44(c 1 c 2 c 3 )\uff0c\u5176\u4e2d c 1 c 2 \u5c6c\u65bc\u96d9\u97f3\u8a5e\u4e14 c 3 \u5c6c\u65bc\u884d\u751f\u5f8c\u7db4\u6216\u63a5\u5c3e\u8a5e\u8005\uff0c\u4e26\u4e14 \u4e09\u5b57\u7d44(c 1 c 2 c 3 )\u51fa\u73fe\u65bc\u8a9e\u6599\u4e2d\uff0c\u5176\u90e8\u5206\u5b57\u7d44\u4e0d\u5f97\u6bcf\u6b21\u8207\u76f8\u9130\u5b57\u5143\u69cb\u6210\u96d9\u97f3\u8a5e\u6216\u4e09 \u97f3\u8a5e\u3002 (\u4e09\u97f3\u8a5e\u8403\u53d6\u6cd5\u5247\u4e8c): \u82e5\u4e09\u5b57\u7d44(c 1 c 2 c 3 )\uff0c\u5176\u4e2d c 2 c 3 \u5c6c\u65bc\u96d9\u97f3\u8a5e\u4e14 c 1 \u5c6c\u65bc\u884d\u751f\u524d\u7db4\u6216\u63a5\u982d\u8a5e\u8005\uff0c\u4e26\u4e14 \u4e09\u5b57\u7d44(c 1 c 2 c 3 )\u51fa\u73fe\u65bc\u8a9e\u6599\u4e2d\uff0c\u5176\u90e8\u5206\u5b57\u7d44\u4e0d\u5f97\u6bcf\u6b21\u8207\u76f8\u9130\u5b57\u5143\u69cb\u6210\u96d9\u97f3\u8a5e\u6216\u4e09 \u97f3\u8a5e\u3002 3-3",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u2211 = j ij ij ij G T G T G P ) ( ) ( ) ( \u5176\u4e2d T(G ij )\u8868\u793a\u9577\u5ea6\u70ba i \u7684\u7b2c j \u500b\u5b57\u7d44 G ij \u7684\u51fa\u73fe\u6b21\u6578\u3002\u5b57\u7d44\u9593\u7684\u7279\u5fb5\u6709 (1) \u76f8\u5c0d\u983b\u7387(relative frequency count)[Wu 93]\u662f\u5c07\u5b57\u7d44\u7684\u51fa\u73fe\u6b21\u6578\u9664\u4ee5\u6240\u6709\u5b57\u7d44 \u7684\u5e73\u5747\u51fa\u73fe\u6b21\u6578\u5982\u516c\u5f0f(4.2)\u3002 \u5176\u4e2d r ij \u662f\u6307\u9577\u5ea6\u70ba i \u7684\u5b57\u7d44\u5eab\u4e2d\u7684\u7b2c j \u500b\u5b57\u7d44\uff0cf ij \u662f r ij \u7684\u51fa\u73fe\u6b21\u6578\uff0cK i \u662f\u6307\u9577\u5ea6\u70ba i \u7684\u5b57\u7d44\u5eab\u4e2d\u6240\u6709\u5b57\u7d44\u7684\u5e73\u5747\u51fa\u73fe\u6b21\u6578\u3002\u4e00\u822c\u7684\u60c5\u6cc1\u4f86\u8aaa\uff0c\u76f8\u5c0d\u983b\u7387\u8d8a\u9ad8\u7684\u5b57\u7d44\uff0c\u53ef\u80fd \u662f\u5c6c\u65bc\u8a5e\u985e\u7684\u6a5f\u7387\u8d8a\u9ad8\u3002 (4.1) i ij ij K f r =",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "= = = = z P y P x P z y x P z y x D \u5176\u4e2d P(x=1,y=1,z=1) \u662f\u4e2d\u6587\u5b57 z \u7dca\u8ddf\u8457 xy \u51fa\u73fe\u7684\u6a5f\u7387\uff0cP(x=1)\u3001P(y=1)\u8207 P(z=1)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u2211 \u2208 \u2212 = ) ( ) ( ) ( _ log ) ( _ ) ( _ i j i G LN C j G T j i C L P C L P G L H \u2211 \u2208 \u2212 = ) ( ) ( ) ( _ log ) ( _ ) ( _ i j i G RN C j G T j i C R P C R P G R H (4.7a) (4.7b) \u5176\u4e2d H_L(G i )\u8207 H_R(G i )\u5206\u5225\u4ee3\u8868\u5b57\u7d44 G i \u7684\u5de6\u71b5\u8207\u53f3\u71b5\uff0cLN(G i )\u8207 RN(G i )\u5206\u5225\u4ee3\u8868\u5b57 \u7d44 G i \u7684\u5de6\u76f8\u9130\u5b57\u5143\u96c6\u5408\u8207\u53f3\u76f8\u9130\u5b57\u5143\u96c6\u5408\uff0cP_L(C j )\u8207 P_R(C j )\u5247\u5206\u5225\u4ee3\u8868\u5b57\u5143 C j \u5728 G i \u7684\u5de6\u76f8\u9130\u5b57\u5143\u96c6\u5408\u7684\u51fa\u73fe\u6a5f\u7387\uff0c\u8207\u53f3\u76f8\u9130\u5b57\u5143\u96c6\u5408\u7684\u51fa\u73fe\u6a5f\u7387\u3002 \u5f9e\u5be6\u9a57\u4e2d\u6211\u5011\u767c\u73fe\u5e7e\u4e4e\u6240\u6709\u5b57\u7d44\u7684\u76f8\u5c0d\u983b\u7387\u8207\u9ab0\u5b50\u77e9\u9663\u7684\u7279\u5fb5\u503c\u90fd\u843d\u5728\u503c\u57df\u7684 \u6700\u5c0f\u767e\u5206\u4e4b\u4e94\uff0c\u5c24\u5176\u9ab0\u5b50\u77e9\u9663\u5e7e\u4e4e\u5168\u90fd\u843d\u5728 0 \u5230 0.05 \u4e4b\u9593\uff0c\u9019\u6a23\u7684\u5206\u4f48\u5e7e\u4e4e\u986f\u4e0d\u51fa\u5b57 \u7d44\u7684\u5dee\u7570\u6027\u3002\u800c\u4e8c\u5b57\u7d44\u7684\u76f8\u95dc\u5ea6\u5206\u4f48\u60c5\u5f62\u76f8\u7576\u63a5\u8fd1\u9ad8\u65af\u5206\u4f48(Normal Distribution) \uff0c \u5de6\u71b5\u8207\u53f3\u71b5\u9664\u4e86\u7279\u5fb5\u503c\u70ba 0 \u7684\u500b\u6578\u8f03\u591a\u4e4b\u5916\uff0c\u5176\u9918\u7684\u5206\u4f48\u8f03\u70ba\u5e73\u5747\u3002\u56e0\u6b64\u6211\u5011\u9996\u5148\u4ee5 \u76f8\u95dc\u5ea6\u3001\u5de6\u71b5\u8207\u53f3\u71b5\u4f5c\u70ba\u7cfb\u7d71\u7684\u8f38\u5165\u7279\u5fb5\u3002\u82e5\u8981\u8003\u616e\u81ea\u52d5\u7279\u5fb5\u9078\u53d6\u7684\u554f\u984c\u53ef\u4ee5\u53c3\u8003\u5faa \u5e8f \u5411 \u524d \u9078 \u53d6 (Sequential Forward Selection) \u3001 Generalized \"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": ") ( ) )( ( 2 ) ( ) 1 ( 2 1 exp 1 2 1 ) | , ( \u03c3 \u00b5 \u03c3 \u03c3 \u00b5 \u00b5 \u03c3 \u00b5 \u03c3 \u03c0\u03c3 ) ( ) ( : ) ( _ i i j j G T G C T C L P \u2212 (4.7c) ) ( ) ( : ) ( _ i j i j G T C G T C R P \u2212 (4.7d) (4.8a) \uf8f4 \uf8fe \uf8f4 \uf8fd \uf8fc \uf8f4 \uf8f3 \uf8f4 \uf8f2 \uf8f1 \uf8f7 \uf8f7 \uf8f8 \uf8f6 \uf8ec \uf8ec \uf8ed \uf8eb \u2212 + \u2212 \u2212 \u2212 \u2212 \u2212 \u2212 \u2212 = \u2212 r r r a r a m a r a R R A r A r r Word Non R A f 2 ' 2 ' ' ' ' ' ' 2 ' 2 ' 2 ' 2 ' ' ' ) ( ) )( ( 2 ) ( ) 1 ( 2 1 exp 1 2 1 ) | , ( \u03c3 \u00b5 \u03c3 \u03c3 \u00b5 \u00b5 \u03c3 \u00b5 \u03c3 \u03c0\u03c3 \u5176\u4e2d A \u548c R \u662f\u4ee3\u8868\u76f8\u95dc\u5ea6\u8207\u76f8\u5c0d\u983b\u7387\u7684\u8b8a\u6578\u3002\u5047\u8a2d A \u548c R \u90fd\u662f\u5c6c\u65bc\u9ad8\u65af\u5206\u4f48\uff0c\u800c\u03bc a \u662f\u8a5e\u985e\u5b57\u7d44\u7684\u76f8\u95dc\u5ea6\u5e73\u5747\u6578\u3001\u03bc ' a \u662f\u8a5e\u985e\u5b57\u7d44\u7684\u76f8\u95dc\u5ea6\u5e73\u5747\u6578\uff0c\u03bc a \u662f\u8a5e\u985e\u5b57\u7d44\u7684\u76f8\u5c0d \u983b\u7387\u5e73\u5747\u6578\u3001\u03bc ' a \u662f\u8a5e\u985e\u5b57\u7d44\u7684\u76f8\u5c0d\u983b\u7387\u5e73\u5747\u6578\uff0c\u03c3 a \u662f\u8a5e\u985e\u5b57\u7d44\u7684\u76f8\u95dc\u5ea6\u6a19\u6e96\u5dee\uff0c\u03c3 ' a \u662f\u975e\u8a5e\u985e\u5b57\u7d44\u7684\u76f8\u95dc\u5ea6\u6a19\u6e96\u5dee\uff0c\u03c3 r \u662f\u8a5e\u985e\u5b57\u7d44\u7684\u76f8\u5c0d\u983b\u7387\u6a19\u6e96\u5dee\uff0c\u03c3 ' r \u662f\u975e\u8a5e\u985e\u5b57\u7d44 \u7684\u76f8\u5c0d\u983b\u7387\u6a19\u6e96\u5dee\uff0cr \u662f\u8a5e\u985e\u5b57\u7d44\u76f8\u95dc\u5ea6\u8207\u76f8\u5c0d\u983b\u7387\u7684\u76f8\u95dc\u4fc2\u6578\uff0cr ' \u662f\u975e\u8a5e\u985e\u5b57\u7d44\u76f8\u95dc \u5ea6\u8207\u76f8\u5c0d\u983b\u7387\u7684\u76f8\u95dc\u4fc2\u6578\u3002\u5b9a\u7fa9\u4e86\u6a5f\u7387\u51fd\u6578\u5f8c\uff0c\u5c07\u6a5f\u7387\u51fd\u6578\u5957\u5165\u5c0d\u6578\u53ef\u80fd\u5ea6\u6bd4\u7387\u6a21\u7d44\uff0c \u82e5\u662f log\u03bb\u5c0f\u65bc\u9580\u6abb\u503c T lrm \u5247\u662f\u5c6c\u65bc\u975e\u8a5e\u985e\uff0c\u82e5\u662f log\u03bb\u5927\u65bc T lrm \u5247\u5c6c\u65bc\u8a5e\u985e\uff0c\u5728\u672c\u7cfb \u7d71 T lrm \u7684\u9810\u8a2d\u503c\u662f 0\u3002 \u6211\u5011\u6240\u9078\u53d6\u7684\u7279\u5fb5\u662f\u76f8\u95dc\u5ea6\u3001\u5de6\u71b5\u8207\u53f3\u71b5\uff0c\u56e0\u6b64\u6211\u5011\u4f7f\u7528\u591a\u8b8a\u6578\u7684\u9ad8\u65af\u51fd\u6578\u4f86\u4f5c \u70ba\u53ef\u80fd\u6a5f\u7387\u51fd\u6578[Duda 73]\uff1a \uf8fa \uf8fb \uf8f9 \uf8ef \uf8f0 \uf8ee \u2212 \u03a3 \u2212 \u2212 \u03a3 = \u2212 ) ( ) ( 2 1 exp | | ) 2 ( 1 ) | ( 1 2 / 1 2 / \u00b5 \u00b5 \u03c0 x x word x f t d ] | [ word x E = \u00b5 ] ) )( [( t x x E \u00b5 \u00b5 \u2212 \u2212 = \u03a3 \uf8fa \uf8fb \uf8f9 \uf8ef \uf8f0 \uf8ee \u2212 \u03a3 \u2212 \u2212 \u03a3 = \u2212 \u2212 ) ( ) ( 2 1 exp | | ) 2 ( 1 ) | ( ' 1 ' ' 2 / 1 ' 2 / \u00b5 \u00b5 \u03c0 x x word Non x f t d ] | [ ' word Non x E \u2212 = \u00b5 ] ) )( [( ' ' ' t x x E \u00b5 \u00b5 \u2212 \u2212 = \u03a3 \u5176\u4e2d x \u4ee3\u8868\u4e00\u500b\u884c\u5411\u91cf[A LH RH] t \uff0cA\u3001LH \u8207 RH \u5206\u5225\u4ee3\u8868\u76f8\u95dc\u5ea6\u3001\u5de6\u71b5\u8207\u53f3\u71b5\u7684\u8b8a \u6578\uff0c\u4e26\u5047\u8a2d\u6b64\u4e09\u8b8a\u6578\u662f\u5c6c\u65bc\u9ad8\u65af\u5206\u4f48\uff0c\u03bc\u662f\u4ee3\u8868\u8a5e\u985e\u5b57\u7d44\u7684\u7279\u5fb5\u5e73\u5747\u503c\uff0c\u03bc=[\u03bc A \u03bc LH \u03bc RH ] t \uff0c\u03bc A \u03bc LH \u03bc RH \u5206\u5225\u4ee3\u8868\u8a5e\u985e\u5b57\u7d44\u7684\u76f8\u95dc\u5ea6\u3001\u5de6\u71b5\u8207\u53f3\u71b5\u7684\u5e73\u5747\u503c\uff0c\u03bc ' \u662f \u4ee3\u8868\u975e\u8a5e\u985e\u5b57\u7d44\u7684\u7279\u5fb5\u5e73\u5747\u503c\uff0c\u03bc ' =[\u03bc ' A \u03bc ' LH \u03bc ' RH ] t \uff0c\u03bc ' A \u03bc ' LH \u03bc ' RH \u5206\u5225\u4ee3\u8868\u975e (4.9d)",
+                        "eq_num": "("
+                    }
+                ],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Mandarin Chinese",
+                "authors": [
+                    {
+                        "first": "Charles",
+                        "middle": [
+                            "N"
+                        ],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Sandra",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Thompson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Li, Charles N. and Thompson, Sandra A., \"Mandarin Chinese,\" University of California Press, New York, 1992.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Rule-Based Word Identification for Mandarin Chinese Sentences -A Unification Approach",
+                "authors": [
+                    {
+                        "first": "Ching-Long",
+                        "middle": [],
+                        "last": "Yeh",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Lee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "His-Jian",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "Computer Processing of Chinese & Oriental Languages",
+                "volume": "5",
+                "issue": "2",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Yeh, Ching-Long and Lee, His-Jian, \"Rule-Based Word Identification for Mandarin Chinese Sentences -A Unification Approach,\" Computer Processing of Chinese & Oriental Languages, Vol. 5, No.2, March 1991.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Retrieving Collocations from Text: Xtract",
+                "authors": [
+                    {
+                        "first": "Frank",
+                        "middle": [],
+                        "last": "Smadja",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Computational Linguistics",
+                "volume": "19",
+                "issue": "1",
+                "pages": "143--177",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Smadja, Frank, \"Retrieving Collocations from Text: Xtract,\" Computational Linguistics, Vol. 19, No. 1, 1993, pp. 143-177.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Translating Collocations for Bilingual Lexicons",
+                "authors": [
+                    {
+                        "first": "Frank",
+                        "middle": [],
+                        "last": "Smadja",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Mckeown",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "V",
+                        "middle": [],
+                        "last": "Hatzivasiloglou",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Computational Linguistics",
+                "volume": "22",
+                "issue": "1",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Smadja, Frank, McKeown, K.R. and Hatzivasiloglou, V. \"Translating Collocations for Bilingual Lexicons,\" A Statistical Approach,\" Computational Linguistics, Vol. 22, No. 1, 1996.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Introduction to Artificial Neural Systems",
+                "authors": [
+                    {
+                        "first": "Jacek",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Zurada",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Zurada, Jacek M., \"Introduction to Artificial Neural Systems\", West Publishing Company, USA, 1992.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Combining Dictionary, Rules and Statistical Information in Segmentation of Chinese",
+                "authors": [
+                    {
+                        "first": "Jian",
+                        "middle": [],
+                        "last": "Nie",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Yun",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Marie-Louise",
+                        "middle": [],
+                        "last": "Hannan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Wanying",
+                        "middle": [],
+                        "last": "Hannan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Computer Processing of Chinese and Oriental Languages",
+                "volume": "9",
+                "issue": "2",
+                "pages": "125--143",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Nie, Jian Yun, Hannan, Marie-Louise and Hannan, Wanying, \"Combining Dictionary, Rules and Statistical Information in Segmentation of Chinese,\" Computer Processing of Chinese and Oriental Languages, Vol. 9, No. 2, December 1995, pp. 125-143.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Automatic Lexicon Acquisition and Precision-Recall Maximization for Untagged Text Corpora",
+                "authors": [
+                    {
+                        "first": "Jing",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Shin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chang, Jing Shin, \"Automatic Lexicon Acquisition and Precision-Recall Maximization for Untagged Text Corpora\", National Tsing-Hua University, P.h.D. thesis, 1997.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Chinese Word Segmentation through Constraint Satisfaction and Statistical Optimization",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Chang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "Chinese) Proceedings of ROCLING-IV, R.O.C. Computational Linguistics Conferences",
+                "volume": "",
+                "issue": "",
+                "pages": "147--165",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chang, J. S., Chen, C. D. and Chen, S. D., \"Chinese Word Segmentation through Constraint Satisfaction and Statistical Optimization,\" (in Chinese) Proceedings of ROCLING-IV, R.O.C. Computational Linguistics Conferences, Taiwan ROC, 1991, pp. 147-165.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Word Association Norms, Mutual Information and Lexicography",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Church",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Hanks",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Computational Linguistics",
+                "volume": "16",
+                "issue": "",
+                "pages": "22--29",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Church, K. and Hanks, P., \"Word Association Norms, Mutual Information and Lexicography,\" Computational Linguistics, Vol.16, March. 1990, pp. 22-29.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Unknown Word Detection for Chinese by a Corpus-based Learning Method",
+                "authors": [
+                    {
+                        "first": "Keh",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Jiann",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ming",
+                        "middle": [],
+                        "last": "Bai",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Hong",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of ROCLING X",
+                "volume": "",
+                "issue": "",
+                "pages": "159--174",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, Keh Jiann, Bai, Ming Hong, \"Unknown Word Detection for Chinese by a Corpus-based Learning Method,\" Proceedings of ROCLING X, Taipei, Taiwan, ROC, 1997, pp. 159-174.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "A preliminary study on unknown word problem in Chinese word segmentation",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [
+                            "Y"
+                        ],
+                        "last": "Lin",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Chang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "Y"
+                        ],
+                        "last": "Su",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Proceedings of 1993 R.O.C. Computational Linguistics Conference",
+                "volume": "",
+                "issue": "",
+                "pages": "119--137",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lin, M.Y., Chang, T. H. and Su, K. Y., \"A preliminary study on unknown word problem in Chinese word segmentation,\" Proceedings of 1993 R.O.C. Computational Linguistics Conference, Taiwan, 1993, pp.119-137.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Corpus-based Automatic Compound Extraction with Mutual Information and Relative Frequency Count",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Wu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "Y"
+                        ],
+                        "last": "Su",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Proceedings of ROCLING VI",
+                "volume": "",
+                "issue": "",
+                "pages": "207--216",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Wu, M. W. and Su, K. Y. \"Corpus-based Automatic Compound Extraction with Mutual Information and Relative Frequency Count,\" Proceedings of ROCLING VI, Nantou, Taiwan, ROC, Sep. 1993pp. 207-216.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "A Statistical Method For Finding Word Boundaries In Chinese Text",
+                "authors": [
+                    {
+                        "first": "Richard",
+                        "middle": [],
+                        "last": "Sproat",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Chilin",
+                        "middle": [],
+                        "last": "Shin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Computer Processing of Chinese & Oriental Language",
+                "volume": "4",
+                "issue": "4",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Sproat, Richard and Shin, Chilin \"A Statistical Method For Finding Word Boundaries In Chinese Text,\" Computer Processing of Chinese & Oriental Language, Vol. 4, No. 4, March 1990.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "The Processing of English Compound and Complex Words in an English-Chinese Machine Translation System",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "Y"
+                        ],
+                        "last": "Su",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1988,
+                "venue": "Proceedings of ROCLING I",
+                "volume": "",
+                "issue": "",
+                "pages": "87--98",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, S. C. and Su, K. Y. \"The Processing of English Compound and Complex Words in an English-Chinese Machine Translation System,\" Proceedings of ROCLING I, Nantou, Taiwan, 1988, pp. 87-98.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "type_str": "figure",
+                "text": "\u76f8\u95dc\u5ea6(Association)[Sproat 90]\u5b9a\u7fa9\u5982\u4e0b\uff1a \u5176\u4e2d P(a)\u3001P(b) \u5206\u5225\u4ee3\u8868\u4e2d\u6587\u5b57 a \u8207 b \u7684\u51fa\u73fe\u6a5f\u7387\u3002P(ab)\u4ee3\u8868\u96d9\u5b57\u7d44 ab \u7684\u51fa\u73fe\u6a5f\u7387\u3002 \u6b64\u7d71\u8a08\u7279\u5fb5\u6709\u4e00\u7f3a\u9ede\uff0c\u7576 P(a)\u3001P(b)\u90fd\u5f88\u5c0f\u7684\u6642\u5019\uff0cA(ab)\u5bb9\u6613\u8b8a\u5f97\u5f88\u5927\u3002\u4e09\u5b57\u7d44\u7684\u76f8 \u95dc\u5ea6 A(abc)\u5b9a\u7fa9\u70ba\uff1a \u5176\u4e2d P(a)\u3001P(b) \u3001P(c) \u5206\u5225\u4ee3\u8868\u4e2d\u6587\u5b57 a\u3001b \u8207 c \u7684\u51fa\u73fe\u6a5f\u7387\uff0cP(abc)\u5247\u4ee3\u8868\u4e09\u5b57\u7d44 abc \u7684\u51fa\u73fe\u6a5f\u7387\u3002 x=1,y=1) \u662f\u4e2d\u6587\u5b57 y \u7dca\u8ddf\u8457\u4e2d\u6587\u5b57 x \u51fa\u73fe\u7684\u6a5f\u7387\uff0cP(x=1)\u8207 P(y=1)\u5247\u5206\u5225\u662f\u4e2d \u6587\u5b57 x\u3001y \u51fa\u73fe\u7684\u6a5f\u7387\u3002\u7531\u4e0a\u5f0f\u53ef\u767c\u73fe\u9ab0\u5b50\u77e9\u9663\u8207\u76f8\u95dc\u5ea6\u5f88\u50cf\uff0c\u7576 P(x=1)\u8207 P(y=1)\u90fd",
+                "num": null,
+                "uris": null
+            },
+            "TABREF3": {
+                "text": "1\uff0c1]\uff0c\u82e5\u662f\u8f38\u51fa\u503c\u5927\u65bc T mlff \u5247\u8996\u70ba\u8a5e\u5f59\uff0c \u82e5\u662f\u8f38\u51fa\u503c\u5c0f\u65bc T mlff \u5247\u8996\u70ba\u975e\u8a5e\u5f59\u985e\u5225\uff0c\u7cfb\u7d71\u9810\u8a2d\u7684 T mlff \u662f 0\u3002 \u5716 4-1 \u662f\u8abf\u6574\u4e0d\u540c\u9580\u6abb\u503c T mlff \u8207 T lrm \u6642\uff0c\u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44\u8207\u53ef\u80fd\u5ea6\u6bd4\u7387\u6a21\u7d44\u96d9\u97f3",
+                "content": "<table><tr><td colspan=\"10\">\u8a5e\u985e\u5b57\u7d44\u7684\u76f8\u95dc\u5ea6\u3001\u5de6\u71b5\u8207\u53f3\u71b5\u7684\u5e73\u5747\u503c\uff0c\u03a3\u662f\u4ee3\u8868\u8a5e\u985e\u5b57\u7d44\u7279\u5fb5\u7684\u76f8\u95dc\u4fc2\u6578\u77e9\u9663\uff0c \u8a5e\u8403\u53d6\u6b63\u78ba\u7387\u8207\u53ec\u56de\u7387\u7684\u8b8a\u5316\u60c5\u5f62\u3002\u5728\u96d9\u5b57\u7d44\u7684\u65b0\u8a5e\u8403\u53d6\u65b9\u9762\uff0c\u7576\u9ad8\u53ec\u56de\u7387\u7684\u60c5\u5f62\u6642\uff0c</td></tr><tr><td colspan=\"10\">\u03a3 ' \u4ee3\u8868\u975e\u8a5e\u985e\u5b57\u7d44\u7684\u7279\u5fb5\u76f8\u95dc\u4fc2\u6578\u77e9\u9663\u3002 \u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44\u7684\u6b63\u78ba\u7387\u512a\u65bc\u53ef\u80fd\u5ea6\u6bd4\u7387\u6a21\u7d44;\u800c\u7576\u4f4e\u53ec\u56de\u7387\u7684\u60c5\u5f62\u6642\uff0c\u53ef\u80fd\u5ea6\u6bd4\u7387\u6a21</td></tr><tr><td colspan=\"10\">4.8b) \u6211\u5011\u5229\u7528\u5728 3.4 \u7bc0\u5b9a\u7fa9\u7684\u6b63\u78ba\u7387\u8207\u53ec\u56de\u7387\u4f86\u8a55\u4f30\u7cfb\u7d71\u7684\u6548\u80fd\uff0c\u53e6\u5916\u4ee5\u52a0\u6b0a\u5f0f\u6b63\u78ba \u7d44\u7684\u6b63\u78ba\u7387\u5247\u512a\u65bc\u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44\u3002</td></tr><tr><td colspan=\"7\">\u53ec\u56de\u7387(weighted precision recall,WPR)\u505a\u70ba\u8861\u91cf\uff0c</td><td/><td/><td/></tr><tr><td>\u52a0\u6b0a\u5f0f\u6b63\u78ba\u53ec\u56de\u7387</td><td>=</td><td>W 1</td><td>\u00d7</td><td>\u6b63\u78ba\u7387</td><td>+</td><td>W 2</td><td>\u00d7</td><td>, \u53ec\u56de\u7387</td><td>(4.10)</td></tr><tr><td colspan=\"3\">\u5176\u4e2d W 1 \u8207 W 2 \u7686\u8a2d\u5b9a\u70ba\u4e8c\u5206\u4e4b\u4e00\u3002</td><td/><td/><td/><td/><td/><td/><td/></tr><tr><td colspan=\"10\">\u6211\u5011\u4ee5\u4e82\u6578\u9078\u53d6\u4e09\u5206\u4e4b\u4e8c\u7684\u5b57\u7d44\u4f5c\u70ba\u8a13\u7df4\u8cc7\u6599\uff0c\u5206\u5225\u4f7f\u7528\u53ef\u80fd\u5ea6\u6bd4\u7387\u6a21\u7d44\u8207\u985e\u795e</td></tr><tr><td colspan=\"10\">\u7d93\u7db2\u8def\u6a21\u7d44\u9032\u884c\u65b0\u8a5e\u8403\u53d6\u3002\u6211\u5011\u4f7f\u7528\u76f8\u95dc\u5ea6\u3001\u5de6\u71b5\u8207\u53f3\u71b5\u4f5c\u70ba\u7d71\u8a08\u5f0f\u7279\u5fb5\uff0c\u4e26\u4e14\u5229\u7528</td></tr><tr><td colspan=\"10\">\u591a\u8b8a\u6578\u9ad8\u65af\u51fd\u6578\u4f5c\u70ba\u53ef\u80fd\u5ea6\u6bd4\u7387\u6a21\u7d44\u7684\u6a5f\u7387\u5206\u4f48\uff0c\u8a08\u7b97\u8a13\u7df4\u8cc7\u6599\u7684\u5e73\u5747\u503c\u8207\u76f8\u95dc\u4fc2\u6578</td></tr><tr><td colspan=\"9\">\u77e9\u9663\uff0c\u53ef\u5f97\u5230\u9ad8\u65af\u51fd\u6578\u7684\u53c3\u6578\uff0c\u5957\u5165\u9ad8\u65af\u51fd\u6578\u5f8c\u53ef\u5f97\u5230\u53ef\u80fd\u5ea6\u6bd4\u7387\u6a21\u7d44\u3002</td><td/></tr><tr><td colspan=\"10\">\u5728\u985e\u795e\u7d93\u7db2\u8def\u8403\u53d6\u6a21\u7d44\u65b9\u9762\uff0c\u7531\u65bc\u76f8\u95dc\u5ea6\u7684\u503c\u57df\u6bd4\u5de6\u71b5\u8207\u53f3\u71b5\u5927\u8a31\u591a\uff0c\u56e0\u6b64\u6211\u5011</td></tr><tr><td colspan=\"10\">\u5148\u5c07\u6b64\u4e09\u7a2e\u7279\u5fb5\u4f5c\u4e00\u7c21\u55ae\u7684\u503c\u57df\u8f49\u63db\uff0c\u5c07\u7279\u5fb5\u7684\u503c\u57df\u8f49\u63db\u5230[0.05\uff0c0.95]\u53ca[-0.95\uff0c \u5716 4-2\uff1a\u4e09\u97f3\u8a5e\u8403\u53d6\u6548\u80fd\u6bd4\u8f03\u5716 \u5716 4-3\uff1a\u4e09\u97f3\u8a5e\u8403\u53d6\u6548\u80fd\u6bd4\u8f03\u5716</td></tr><tr><td colspan=\"10\">-0.05]\u3002\u56e0\u70ba 0 \u5728\u5012\u50b3\u905e\u7db2\u8def\u4e2d\uff0c\u662f\u6c92\u6709\u4f5c\u7528\u7684\uff0c\u56e0\u6b64\u907f\u958b 0\u3002\u82e5\u662f\u7279\u5fb5 X \u7684\u503c\u6c38\u9060\u5927</td></tr><tr><td>\u65bc\u96f6\uff0c\u5247\u4f7f\u7528\u4ee5\u4e0b\u7684\u8f49\u63db\u51fd\u6578 5. \u7d50\u8ad6</td><td/><td/><td/><td/><td/><td/><td/><td/><td/></tr><tr><td colspan=\"10\">\u5716 4-1 \u96d9\u97f3\u8a5e\u8403\u53d6\u6548\u80fd\u6bd4\u8f03\u5716 \u672c\u8ad6\u6587\u63d0\u51fa\u4e00\u500b\u5169\u968e\u6bb5\u7684\u4e2d\u6587\u65b0\u8a5e\u8403\u53d6\u6280\u8853\uff0c\u53ef\u61c9\u7528\u65bc\u4e2d\u6587\u6587\u4ef6\u8655\u7406\u7cfb\u7d71\uff0c\u5c07\u8a9e\u6599\u4e2d</td></tr><tr><td colspan=\"10\">(4.9a) (4.9b) (4.9c) (4.9e) (4.9f) \u56e0\u70ba\u6211\u5011\u9996\u5148\u53ea\u4f7f\u7528\u4e09\u7a2e\u7279\u5fb5\uff0c\u6240\u4ee5\u8f38\u5165\u5c64\u7684\u7bc0\u9ede\u500b\u6578\u662f\u4e09\u500b\uff0c\u8f38\u51fa\u503c\u4ea6\u53ea\u6709\u4e00\u500b\uff0c 05 . 0 min) ( min max 05 . 0 95 . 0 ) ( + \u2212 \u2212 \u2212 = x x f \u5426\u5247\u4f7f\u7528\u6b64\u51fd\u6578 05 . 0 ) ( max 05 . 0 95 . 0 ) ( + \u2212 = x x f , if x \u2267 0 05 . 0 ) ( max 05 . 0 95 . 0 ) ( \u2212 \u2212 = x x f , if x &lt; 0 (4.11b) (4.11c) \u5728\u4e09\u97f3\u8a5e\u7684\u8403\u53d6\u65b9\u9762\uff0c\u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44\u5728\u9580\u6abb\u503c\u70ba\u9810\u8a2d\u503c\u6642\uff0c\u7565\u512a\u65bc\u53ef\u80fd\u5ea6\u6bd4\u7387 \u6a21\u7d44\u3002\u89c0\u5bdf\u5716 4-2\uff0c\u767c\u73fe\u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44\u8207\u53ef\u80fd\u5ea6\u6bd4\u7387\u6a21\u7d44\u65bc\u4e09\u97f3\u8a5e\u7684\u8403\u53d6\u80fd\u529b\u4e26\u7121 \u660e\u986f\u7684\u512a\u52a3\u5206\u5225\u3002 \u7531\u65bc\u4e09\u5b57\u7d44\u4e2d\u5176\u4e8c\u5b57\u7d44\u7684\u8cc7\u8a0a\u662f\u6709\u610f\u7fa9\u7684\u56e0\u6b64\u5728\u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44\u4e09\u5b57\u7d44 c 1 c 2 c 3 \u65b0 \u8a5e\u8403\u53d6\u4e2d\u9664\u4e86\u539f\u672c\u4f7f\u7528\u7684\u76f8\u95dc\u5ea6\u3001\u5de6\u71b5\u8207\u53f3\u71b5\u7684\u4e09\u500b\u7279\u5fb5\u5916\uff0c\u6211\u5011\u53e6\u52a0\u5165\u5176\u90e8\u5206\u5b57\u7d44 c 1 c 2 \u8207 c 2 c 3 \u7684\u76f8\u95dc\u5ea6\u3001\u5de6\u71b5\u8207\u53f3\u71b5\u4f5c\u70ba\u7279\u5fb5\uff0c\u7279\u5fb5\u500b\u6578\u589e\u52a0\u70ba\u4e5d\u500b\u3002\u5728\u8868 4.1 \u548c\u5716 4-3 \u53ef\u77e5\u4ee5\u7279\u5fb5\u6578\u7684\u589e\u52a0\u78ba\u5be6\u53ef\u63d0\u9ad8\u5206\u8fa8\u7684\u6b63\u78ba\u7387\u3002 \u53ef\u80fd\u5ea6\u6bd4\u7387\u6a21\u7d44 \u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44 (\u4e09\u7a2e\u7279\u5fb5) \u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44 (\u4e5d\u7a2e\u7279\u5fb5) \u6b63\u78ba\u7387 16.32\uff05 13.68\uff05 18.97\uff05 \u53ec\u56de\u7387 59.3\uff05 63.32\uff05 77.89\uff05 \u52a0\u6b0a\u5f0f\u6b63\u78ba\u53ec\u56de\u7387 37.81\uff05 38.5\uff05 48.83\uff05 \u8868 4-1 \u4e09\u97f3\u8a5e\u8403\u53d6\u6548\u80fd\u6bd4\u8f03\u8868 \u6709\u610f\u7fa9\u7684\u65b0\u8a5e\u8403\u53d6\u51fa\u4f86\u3002\u5be6\u9a57\u6578\u64da\u7684\u5206\u6790\u986f\u793a\u5229\u7528\u69cb\u8a5e\u5b78\u7684\u65b9\u6cd5\u78ba\u5be6\u80fd\u6709\u6548\u7684\u5c07\u4e09\u97f3 \u65b0\u8a5e\u8403\u53d6\u51fa\u4f86\uff0c\u4e26\u4e14\u6b63\u78ba\u5730\u5c07\u5927\u90e8\u5206\u975e\u8a5e\u5f59\u5b57\u7d44\u904e\u6ffe\u6389\u3002\u53e6\u4e00\u65b9\u9762\u5229\u7528\u985e\u795e\u7d93\u7db2\u8def\u7d50 \u5408\u5404\u7a2e\u7d71\u8a08\u5f0f\u8cc7\u8a0a\u4f86\u8403\u53d6\u65b0\u8a5e\uff0c\u53ef\u5f4c\u88dc\u69cb\u8a5e\u6cd5\u5247\u7684\u4fb7\u9650\u6027\u3002\u6700\u5f8c\u6211\u5011\u4ea6\u63a2\u8a0e\u7279\u5fb5\u7684\u9078 \u53d6\u5c0d\u65bc\u8403\u53d6\u7684\u5f71\u97ff\uff0c\u4e26\u8207\u53ef\u80fd\u5ea6\u6a21\u7d44\u6bd4\u8f03\u3002\u5f9e\u5be6\u9a57\u7684\u7d50\u679c\u6211\u5011\u5f97\u77e5\u4e09\u97f3\u8a5e\u4e2d\u4e8c\u5b57\u7d44\u7279 \u5fb5\u7684\u52a0\u5165\u78ba\u5be6\u80fd\u63d0\u9ad8\u4e09\u97f3\u8a5e\u65b0\u8a5e\u7684\u6b63\u78ba\u7387\u8207\u53ec\u56de\u7387\u3002 \u672c\u8ad6\u6587\u7684\u5f8c\u7e8c\u7814\u7a76\u65b9\u5411\u4e3b\u8981\u6709\u7279\u5fb5\u7684\u81ea\u52d5\u9078\u53d6\u3002\u5728\u4f7f\u7528\u985e\u795e\u7d93\u8403\u53d6\u6a21\u7d44\u6642\uff0c\u9078\u53d6 \u7684\u7279\u5fb5\u7684\u597d\u58de\u6703\u76f4\u63a5\u5f71\u97ff\u5230\u7cfb\u7d71\u7684\u6548\u80fd\uff0c\u5728\u672c\u8ad6\u6587\u4f7f\u7528\u5206\u6790\u5176\u503c\u57df\u5206\u4f48\u60c5\u5f62\u4f86\u4f5c\u7279\u5fb5 \u9078\u53d6\u3002\u4f46\u662f\u7576\u53ef\u4f7f\u7528\u7279\u5fb5\u5f88\u591a\uff0c\u5c0e\u81f4\u96e3\u4ee5\u9010\u500b\u5206\u6790\u6642\uff0c\u5247\u9700\u5229\u7528\u7279\u5fb5\u81ea\u52d5\u9078\u53d6\u4f86\u89e3\u6c7a \u6211\u5011\u4f7f\u7528\u7684\u8f49\u63db\u51fd\u6578\u662f\u96d9\u5f4e\u66f2\u51fd\u6578\u5176\u503c\u57df\u70ba[-(4.11a) \u9019\u500b\u554f\u984c\u3002\u8207\u6b64\u554f\u984c\u76f8\u95dc\u7684\u9084\u6709\u7279\u5fb5\u7684\u503c\u57df\u8f49\u63db\u554f\u984c\uff0c\u7576\u5404\u7a2e\u7279\u5fb5\u8cc7\u8a0a\u7684\u503c\u57df\u7bc4\u570d\u76f8</td></tr><tr><td colspan=\"9\">\u5dee\u592a\u5927\u6642\uff0c\u5c31\u9700\u8981\u7279\u5fb5\u7684\u503c\u57df\u8f49\u63db\uff0c\u907f\u514d\u7cfb\u7d71\u88ab\u5c11\u6578\u5e7e\u500b\u7279\u5fb5\u6240\u4e3b\u5bb0\u3002</td><td/></tr></table>",
+                "html": null,
+                "num": null,
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-1002.json

@@ -0,0 +1,22 @@
+{
+    "paper_id": "O00-1002",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:09.126020Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-1002",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [],
+        "back_matter": [],
+        "bib_entries": {},
+        "ref_entries": {}
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-1004.json

@@ -0,0 +1,656 @@
+{
+    "paper_id": "O00-1004",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:07.991201Z"
+    },
+    "title": "Building A Chinese Text Summarizer with Phrasal Chunks and Domain Knowledge",
+    "authors": [
+        {
+            "first": "Weiquan",
+            "middle": [],
+            "last": "Liu",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Intel China Research Center",
+                "location": {
+                    "addrLine": "601 North Tower, Beijing Kerry Center #1 Guanghua Road",
+                    "postCode": "10002",
+                    "settlement": "Beijing",
+                    "country": "China"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Joe",
+            "middle": [],
+            "last": "Zhou",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Intel China Research Center",
+                "location": {
+                    "addrLine": "601 North Tower, Beijing Kerry Center #1 Guanghua Road",
+                    "postCode": "10002",
+                    "settlement": "Beijing",
+                    "country": "China"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Joe",
+            "middle": [
+                "F"
+            ],
+            "last": "Liu",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Intel China Research Center",
+                "location": {
+                    "addrLine": "601 North Tower, Beijing Kerry Center #1 Guanghua Road",
+                    "postCode": "10002",
+                    "settlement": "Beijing",
+                    "country": "China"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "@intel",
+            "middle": [],
+            "last": "Zhou}",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Intel China Research Center",
+                "location": {
+                    "addrLine": "601 North Tower, Beijing Kerry Center #1 Guanghua Road",
+                    "postCode": "10002",
+                    "settlement": "Beijing",
+                    "country": "China"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "",
+            "middle": [],
+            "last": "Com",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Intel China Research Center",
+                "location": {
+                    "addrLine": "601 North Tower, Beijing Kerry Center #1 Guanghua Road",
+                    "postCode": "10002",
+                    "settlement": "Beijing",
+                    "country": "China"
+                }
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "This paper introduces a Chinese summarizier called ThemePicker. Though the system incorporates both statistical and text analysis models, the statistical model plays a major role during the automated process. In addition to word segmentation and proper names identification, phrasal chunk extraction and content density calculation are based on a semantic network pre-constructed for a chosen domain. To improve the readability of the extracted sentences as auto-generated summary, a shallow parsing algorithm is used to eliminate the semantic redundancy.",
+    "pdf_parse": {
+        "paper_id": "O00-1004",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "This paper introduces a Chinese summarizier called ThemePicker. Though the system incorporates both statistical and text analysis models, the statistical model plays a major role during the automated process. In addition to word segmentation and proper names identification, phrasal chunk extraction and content density calculation are based on a semantic network pre-constructed for a chosen domain. To improve the readability of the extracted sentences as auto-generated summary, a shallow parsing algorithm is used to eliminate the semantic redundancy.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "Due to the overwhelming amount of textual resources over Internet people find it increasingly difficult to grasp targeted information without any adjunctive tools. One of these tools is automatic summarization and abstraction. When coupled with general search and retrieval systems, text summarization can contribute to alleviating the effort in accessing these abundant information resources. It is capable of condensing the amount of original text, enabling the user to quickly capture the main theme of the text.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "Based on the techniques employed (Hovy, 1998) , existing summarization systems can be divided into three categories, i.e., word-frequency-based, cohesion-based, or information-extraction-based.",
+                "cite_spans": [
+                    {
+                        "start": 33,
+                        "end": 45,
+                        "text": "(Hovy, 1998)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "Comparing to the other two techniques the first one is statistical oriented, fast and domain independent (Brandow et al, 1995) . The quality, however, is often questionable. Cohesion-based techniques (or sometimes called as being linguistic oriented) can generate more fluent abstracts, but the sentence-by-sentence computation against the entire raw text is often quite expensive. Even the most advanced part of speech (POS) tagging or syntactic parsing algorithms are unable to handle all the language phenomena emerged from giga-bytes of naturally running text. Summarization based on information extraction relies on the predefined templates. It is domain dependent. The unpredictable textual content over Internet, however, may let the templates suffer from incompletion or intra-contradiction no matter how well they might be predefined.",
+                "cite_spans": [
+                    {
+                        "start": 105,
+                        "end": 126,
+                        "text": "(Brandow et al, 1995)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "In this paper we introduces a Chinese summarization system. Though it is a hybrid system incorporating some natural language techniques, considering the speed and efficiency of text processing we still adapted a statistical oriented algorithm and allowed it to play a major role during the automatic process. After pre-processing, the system first extracts phrasal chunks from the input.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "The phrasal chunks normally refer to meaningful terms and proper names existing in the text that are difficult to capture using simple methods. Then, we use a domain specific concept network to calculate the content density, i.e. measuring the significance score of each individual sentence.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "Finally, a Chinese dependency grammar applies as a shallow parser to process the extracted sentences into bracketed frames so as to achieve further binding and embellishment for the final output.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "The system, hereafter referred to as ThemePicker, works as a plug-in to web browsers. When surfing among some selected Chinese newspaper web sites, ThemePicker monitors the content of the browser s window. When the number of domain words or terms exceeds a pre-defined threshold, it will kick off the summary generation process and display the output in a separate window. Currently, we chose economic news as our specific domain.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "System Overview",
+                "sec_num": "2"
+            },
+            {
+                "text": "The system consists of four components (see Fig. 1 ). The first component is a pre-processor dealing with the layout of the news web pages and removing unnecessary HTML tags while keeping the headline, title and paragraph hierarchy. The retained information will provide the location of the extracted sentences for later manipulation.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 44,
+                        "end": 50,
+                        "text": "Fig. 1",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "System Overview",
+                "sec_num": "2"
+            },
+            {
+                "text": "The second component performs two tasks in parallel, resolving Chinese word segmentation and identifying and extracting phrasal chunks. As it is known to all, Chinese is an ideographical character based language with no spaces or delimiting symbols between adjacent words. After breaking the input sentence into a chain of separate character strings we use a lexical knowledge base to look up each word and parse the sentence appropriately. Person names and other proper names are also recognized during the segmentation process. Phrasal chunks are lexical units larger than words but not idioms. They are content oriented special terms (Zhou, 1999) . We examined hundreds of documents and frequently encountered these phrasal chunks in the text that bear important information about the document. Since the meaning of a phrasal chunk is by no means the simple aggregation of the meanings of all the words in it, the word segmentation can not handle it. ThemePicker uses a statistical algorithm for phrasal chunk identification, aiming at the larger lexical unit that consists of two or more words always occurring in the same sequence.",
+                "cite_spans": [
+                    {
+                        "start": 637,
+                        "end": 649,
+                        "text": "(Zhou, 1999)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "System Overview",
+                "sec_num": "2"
+            },
+            {
+                "text": "The third component in sequence computes the degrees of sentence content density. The computation assigns a significance score to each sentence. The concept net that contains of more than 2000 concept nodes on economic news domain is used to define the semantic similarities between different sentences and adjust the significance scores of sentences across the input text.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "System Overview",
+                "sec_num": "2"
+            },
+            {
+                "text": "Sentences with high scores are selected for the inclusion in the candidate summary.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "System Overview",
+                "sec_num": "2"
+            },
+            {
+                "text": "The fourth component analyzes the candidate sentences using a Chinese dependency grammar. The purpose is to improve the readability of the output summary. In the remaining sections of this paper we will describe in some details the major system components, i.e., word segmentation and proper name identification (Section 3), phrasal chunk extraction (Section 4), domain knowledge for summary generation (Section 5), and the dependency grammar (Section 6). The final section (Section 7) devotes to the system evaluation.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "System Overview",
+                "sec_num": "2"
+            },
+            {
+                "text": "The segmentation algorithm is a single scan Reverse Maximum Matching (RMM). One major difference from other RMMs is the special lexicon it uses. The lexicon consists of two parts, the indexing pointers and the main body of lexical entries (see Fig 2) . ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 244,
+                        "end": 250,
+                        "text": "Fig 2)",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Word Segmentation and Proper Name Identification",
+                "sec_num": "3"
+            },
+            {
+                "text": "The algorithm works efficiently. The average number of comparisons needed to segment each word is only 2.89 (Liu et al, 1998) . The unregistered single characters that are left behind the word segmentation will become the target of proper name recognition. To fulfil the task of recognizing Chinese person names we built a surname and a given name databases. Intuitively, any given Chinese person name is formed by a lead surname and followed by 1 or 2 given names. The surname has only one character and rarely has two, therefore the length of each person name ranges from 2 to 4 characters. In the surname and given name databases, each character is given a possibility value that is obtained by calculating its frequency over a large name bank. Our person name recognition algorithm works as follows.",
+                "cite_spans": [
+                    {
+                        "start": 108,
+                        "end": 125,
+                        "text": "(Liu et al, 1998)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Output summary",
+                "sec_num": null
+            },
+            {
+                "text": "When an unregistered single character word is encountered during the scan of the segmented text, the algorithm will check a) whether the character is a surname, and b) whether the character is followed by one or two single character words. When calculating the possibilities, the title words, such as Mr., Mrs. etc. that immediately before n and verbs that follow n are also considered heuristically.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Output summary",
+                "sec_num": null
+            },
+            {
+                "text": "The difference between Chinese person name and transliterated foreign name is that the latter uses only a limited set of characters. The number of characters that allow to be used to denote foreign origin names is about 400 to 500 (Sun, 1998) . Within this set, a portion of it can only be used as the first character and another subset can only be the tail ones. Using this principle we defined a set of rules to label the margins of foreign names resulting in satisfactory precision and recall.",
+                "cite_spans": [
+                    {
+                        "start": 231,
+                        "end": 242,
+                        "text": "(Sun, 1998)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Output summary",
+                "sec_num": null
+            },
+            {
+                "text": "Company name identification is also statistical and heuristic in nature. Based on the observation and analysis of a large quantity of collected Chinese text, we concluded that most company names can be denoted by the following BNF:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Output summary",
+                "sec_num": null
+            },
+            {
+                "text": "<Geographical Loc> + [<Ordinal Number>] + {<Product Name>|<Trade Name>} + <Appellative Noun>",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Output summary",
+                "sec_num": null
+            },
+            {
+                "text": "Thus, we built a FSM in which heuristic rules are introduced to allow the system capture such text strings as company names.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Output summary",
+                "sec_num": null
+            },
+            {
+                "text": "Our initial evaluation of some sample text databases indicates that approximately 3% of the original text are proper names of various kinds, among whom the above two categories constitute more than 95%. This means that we would lose 2.85% of the segmentation accuracy if no action were taken to handle these two names. The above procedure now achieves more than 96% in accuracy. The improvement to the segmentation is 2.74%.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Output summary",
+                "sec_num": null
+            },
+            {
+                "text": "As mentioned above, proper names denote critical information in the original document. Their incorporation can make the summary more informative. Improved segmentation helps identify domain words more accurately. The identification of proper names also benefits the shallow parsing and improves the coherence and cohesion of summary output. Though phrasal chunk identification is independent to the segmentation, it is character based not word based.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Output summary",
+                "sec_num": null
+            },
+            {
+                "text": "The phrasal chunk identification algorithm is to locate new terms formed by two or more words that frequently occur in the input text. For the words , and found in the input text, if their frequencies all exceed a pre-defined threshold, we can say that they are key words in the original text. But, this does not mean the whole phrasal chunk is also a key word. To determine such a long term or a phrase chunk is also a key word we have to prove that these three words or 6 characters frequently appear in exactly the same sequence.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Phrasal Chunk Identification",
+                "sec_num": "4"
+            },
+            {
+                "text": "Our phrasal chunk identification algorithm uses a data structure used called Association Tree (A-Tree). A unique A-Tree can be constructed for each individual character using itself as the root of the respective tree. \u2022 Repeat step 4 until no leaf can be expanded, then the A-Tree of C i is complete.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Phrasal Chunk Identification",
+                "sec_num": "4"
+            },
+            {
+                "text": "Once all A-Trees are constructed, new phrasal chunks can be extracted using entropy measurement.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 3: Phrasal chunk identification and an A-Tree",
+                "sec_num": null
+            },
+            {
+                "text": "By tracking from the root node to each leaf node we can get a string of characters. For example,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 3: Phrasal chunk identification and an A-Tree",
+                "sec_num": null
+            },
+            {
+                "text": "given a string a 1 a 2 a n b 1 b 2 b m that denotes two sub-strings A=a 1 a 2 a n and B=b 1 b 2 b m with a 1 as the root, the entropy in B given A is:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 3: Phrasal chunk identification and an A-Tree",
+                "sec_num": null
+            },
+            {
+                "text": ") | ( log ) | ( A B p A B H \u2212 = .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 3: Phrasal chunk identification and an A-Tree",
+                "sec_num": null
+            },
+            {
+                "text": "For an A-Tree the ratio |b m | / |a n | is an estimation of p(B|A). The smaller the H value the closer the relationship between these two sub-strings. A zero value means B always follows A, suggesting that AB is a meaningful phrasal chunk.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 3: Phrasal chunk identification and an A-Tree",
+                "sec_num": null
+            },
+            {
+                "text": "For a string \u0393=C 0 C 1 C 2 C n , the entropy in C 1 given C 0 is H C1 = -log P(C 1 |C 0 ). Given C 0 C 1 , entropy in C 2 is H C2 = -log p(C 2 |C 0 C 1 ). Thus, the total entropy measurement of \u0393 is defined as:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 3: Phrasal chunk identification and an A-Tree",
+                "sec_num": null
+            },
+            {
+                "text": ") ( log where , ) ... ( log 0 0 0 0 C p H C C p H H C n n i Ci \u2212 = \u2212 = = \u2211 = \u0393",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 3: Phrasal chunk identification and an A-Tree",
+                "sec_num": null
+            },
+            {
+                "text": "As shown in Fig. 3 there are three phrasal chunks that have been listed with their respective H values with the first one bearing the lowest. The chunk identification algorithm will collect all the phrasal chunks with H value less than a certain threshold among all the A-Trees built from the input text. These phrasal chunks are larger than a word and likely express the key content of the input.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 12,
+                        "end": 18,
+                        "text": "Fig. 3",
+                        "ref_id": "FIGREF3"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Figure 3: Phrasal chunk identification and an A-Tree",
+                "sec_num": null
+            },
+            {
+                "text": "The significance score of a sentence is determined based on the sum of two measurements, the density of domain concepts and the density of phrasal chunks.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "Suppose a sentence denoted as",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "S=U 1 U 2 U 3 U L , U i \u2208[F | W |K], 1<i<L (here F: function words, W:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "domain concept words and K: phrasal chunks), for those U i that belong to F, no contribution will be made to the significance score. For other U i that belong to W, their contribution to the significance score is gained from the domain knowledge contained in a ConceptNet. The ConceptNet is a graphic network constructed semi-automatically with nodes as various concepts and arcs as relations between concepts. The current version of our ConceptNet contains more than 2,000 nodes all collected from a large economic news database (see Fig. 4 ). The relations between concepts are of several types, such as a-kind-of, a-part-of, abbreviation-of, product-of, member-of, etc. The density of domain concepts \u03b1 w is calculated as follows:",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 535,
+                        "end": 541,
+                        "text": "Fig. 4",
+                        "ref_id": "FIGREF4"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "\u2211 \u2211 \u2208 \u2208 \u2212 = W U j i W U i w i j U U U R | | / )) , ( 1 ( \u03b3 \u03b1",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": ", \u03b3 i is a heuristic coefficient.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "R(w 1 ,w 2 )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "is a function that determines the semantic relations between w 1 and w 2 . For those U i that belong to K, their contribution to the significance score is calculated as",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "\u2211 \u2208 = K U i i K i U U H | | / ) ( \u03b3 \u03b1",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "(referring to the previous section on the calculation of H(U i ), the entropy of U i .). Thus, the final significance score for the sentence S is:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": ")) 1 ( ( 2 1 T S S w \u03b1 \u03b2 \u03b1 \u03b2 \u03bb \u03b1 \u2212 + = .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "Conceptually, we give special treatment to domain concept words and phrasal chunks that appear in the title and headline. Some cue words or phrases are also detected that may bring positive or negative contributions to the significance score depending on their properties. \u03b2 1 and \u03b2 2 are balance factors for \u03b1 w and \u03b1 K . \u03bbs is determined by the location of S in the paragraph.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "After all the input sentences receive the significance scores, those having values greater than a predefined threshold are chosen for the possible inclusion in the generated summary. The default length of the output summary is within 10~20% of the original text.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Sentence Extraction Using Domain Knowledge",
+                "sec_num": "5"
+            },
+            {
+                "text": "Though they receive higher significance scores, the extracted sentences cannot be treated as the abstract of the original text. The readability is low even if they are strung together in the order as they occur in the input. The duplication in meaning and the appearance of improper conjunction words often make readers confused. Anaphora without contextual reference also poses difficulty in comprehension.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Dependency Grammar",
+                "sec_num": "6"
+            },
+            {
+                "text": "To bind and embellish the output summary we employed a Chinese dependency grammar to parse the extracted sentence into Dependency Relation Tree (DRT). Based on the methodology introduced in Liu et al, 1998, DRT can further be bracketed into cells. One of the cells is called the core with others being dominated by the core. There exist unique mappings between dependency relations in DRT and the dominating relations among cells. Fig. 5 illustrates such an example. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 431,
+                        "end": 437,
+                        "text": "Fig. 5",
+                        "ref_id": "FIGREF5"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Dependency Grammar",
+                "sec_num": "6"
+            },
+            {
+                "text": "\u2211 = i i i S slot S slot diff S S D )) ( ), ( ( ) , ( 2 1 2 1",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Dependency Grammar",
+                "sec_num": "6"
+            },
+            {
+                "text": "If Core(S 1 ) and Core(S 2 ) are different, D(S 1 , S 2 ) is indefinite. If Core(S 1 ) and Core(S 2 ) are the same, diff( . ) is used to denote the semantic similarities between slot i (S 1 ) and slot i (S 2 ). The more similar the contents in the two slots, the smaller the value of diff( . ), thus the smaller the distance D(S 1 , S 2 ).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Dependency Grammar",
+                "sec_num": "6"
+            },
+            {
+                "text": "A special case of the semantic distance is D(S 1 , S 2 )=0, that means S 1 and S 2 are basically identical in meaning, so one of them can be deleted. In most cases, D(S 1 , S 2 ) is greater than zero. A distance threshold is pre-defined in order to determine which extracted sentence can be eliminated. After the redundancy elimination the remaining portion of extracted sentences is reorganized to assemble the final output summary.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Dependency Grammar",
+                "sec_num": "6"
+            },
+            {
+                "text": "In this paper we introduced a Chinese summarizier called ThemePicker. It is a hybrid system incorporating both statistical and text analysis models. For the sake of speed and efficiency, the algorithm was implemented in a way that allows the statistical model to take the major role during the automated process. We built a semantic network (ConceptNet), a knowledge base that contains more than 2000 concept nodes with arcs indicating the conceptual relationships between or across nodes. Our experiments have showed that the content density measured based on ConceptNet can be more valid than an algorithm purely based on key terms. To achieve higher degrees of readability of the auto-generated summary, we adapted a shallow parsing algorithm to eliminate the semantic redundancy between the extracted sentences. While enhancing the summary cohesion and coherence, the computational overhead is restricted.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Performance Evaluation",
+                "sec_num": "7"
+            },
+            {
+                "text": "As pointed out in the literature, due to the lack of the evaluation standards for auto summaries, it remains to be an open research topic regarding how to compare the performance of a text summarizer with any concrete and solid measurement (Paice, 1990) . We conducted a preliminary system evaluation against the database that contains 2800 news articles (2.4M words in total) on the economic domain. First, two human analysts manually screened 1200 articles and identifies 80 specific topics like Euro, Fortune Forum, RMB won't be depreciated, etc. Then, they manually generated summaries for several selected documents from each of the 40 topics. After that, they compared the automatically generated summaries with those they manually composed. The benchmark uses three grading scales, comparing to the manually generated summary the auto counterpart was assigned as either, good or acceptable or non-acceptable. The results indicated that the total documents that received either good or acceptable grades constitute more than two-thirds of the total documents evaluated. Evaluation using more rigid methodology will be performed in the future. ",
+                "cite_spans": [
+                    {
+                        "start": 240,
+                        "end": 253,
+                        "text": "(Paice, 1990)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Performance Evaluation",
+                "sec_num": "7"
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {},
+        "ref_entries": {
+            "FIGREF0": {
+                "num": null,
+                "type_str": "figure",
+                "text": "System overview and process flow",
+                "uris": null
+            },
+            "FIGREF1": {
+                "num": null,
+                "type_str": "figure",
+                "text": "Lexicon structure and segmentation process",
+                "uris": null
+            },
+            "FIGREF3": {
+                "num": null,
+                "type_str": "figure",
+                "text": "shows an example of A-Trees. Each node consists of a character and an associated integer shows in parentheses. The integer refers to the number of occurrences of the character in the input text. The integers associated with other child nodes denote the number of occurrences that particular character follows its parent node. An A-Tree is constructed in the following way: \u2022 Scan the input and record the position of each individual character C. Define \u03c8 = {C i | C i \u2208 \u2211} as the set of all possible characters found in the input. | C i | is the number of occurrence of C i . Delete all C i when |C i | < T with T as a predefined threshold \u2022 For each remaining individual character C i \u2208 \u03c8, create a A-tree and place C i (n) at the root of the tree and n as the associated integer \u2022 Add all the descendants of C i to the leaf node set \u03d5 = {d j | d j \u2208\u2211 }. Delete those d j where |d i | < T with T as a predefined threshold \u2022 For each node d j in \u03d5, add its descendant characters as described in step 3 and remove d j after it gets expanded",
+                "uris": null
+            },
+            "FIGREF4": {
+                "num": null,
+                "type_str": "figure",
+                "text": "A partial snapshot of ConceptNet for economic news domain",
+                "uris": null
+            },
+            "FIGREF5": {
+                "num": null,
+                "type_str": "figure",
+                "text": "A sample sentence and its DRT To eliminate the redundancy between two extracted sentences, we defined a semantic distance between them. Suppose that the bracketed cells of sentence S are represented as:Core(S):=[slot 1 (S), slot 2 (S), slot n (S)], then we can define the semantic distance between S 1 and S 2 as D(S 1 , S 2 ):",
+                "uris": null
+            },
+            "FIGREF6": {
+                "num": null,
+                "type_str": "figure",
+                "text": "Brandow et al, 1995)  Brandow R. Mitze K. and Rau L F. Automatic Condensation of Electronic Publication by Sentence Selection. Information Processing & Management, 31(5): 675-68, 1995 (Cohen, 1995) Cohen J D. Highlights: Language and Domain Independent Automatic Indexing Terms for Abstracting. Journal of the American Society for Information Science, 46(3): 162-174, 1995 (Hovy, 1998) Hovy E. and Marcu D. Automatic Text Summarization. Tutorial of CONLING/ACL'98. 1998 (Liu et al, 1998) Liu W. Wang M. and Zhong Y. Implementation of a Field Non-specific Hybrid Automatic Abstracting System, in the Proceedings of 2 nd Intl Conf on Information Infrastructure (ICOII 98), Beijing pp275-278 (Paice, 1990) Paice C D. Constructing Literature Abstracts by Computer: Techniques and Prospects. Information Processing & Management, 26(1):171-186, 1990 (Sun, 1998) Sun, M S et al. Identifying Chinese names in Unrestricted texts. Journal of Chinese Information Processing 9(2):16-27, 1998 (Zhou, 1999) Zhou F J. Phrasal Terms in Real-world IR Applications. In Strzalkowski T. eds Natural Language Information Retrieval, pp215-260. Kluwer Academic Publishers, 1999",
+                "uris": null
+            },
+            "TABREF0": {
+                "html": null,
+                "type_str": "table",
+                "text": "Proper names in Chinese carry no signals like capitalization, hyphenation, and interpunction in English, to indicate that they are special and different from other noun phrases. Our algorithm currently can handle two types of proper names, people names and organization names. People names include Chinese person names and names of foreign origin (though treated differently). The majority of organization names are company names due to the nature of the selected domain economic news.",
+                "content": "<table/>",
+                "num": null
+            },
+            "TABREF1": {
+                "html": null,
+                "type_str": "table",
+                "text": "If both conditions are met, these two to three consecutive character string may likely be a person name, denoted as n=sc 1 c 2 . (four-character names are temporarily omitted since they are rare). Here is the calculation of the possibility of n:",
+                "content": "<table><tr><td>p</td><td>(</td><td>n</td><td>)</td><td/><td>log</td><td>p</td><td>(</td><td>s</td><td>)</td><td>p</td><td>(</td><td>1 c</td><td colspan=\"2\">),</td><td colspan=\"2\">if</td><td colspan=\"4\">there</td><td>is</td><td>a</td><td colspan=\"3\">single</td><td>given</td><td>name,</td><td>or</td></tr><tr><td>p</td><td>(</td><td>n</td><td>)</td><td>=</td><td>log</td><td>p</td><td>(</td><td>s</td><td>)</td><td>p</td><td>(</td><td>1 c</td><td>)</td><td colspan=\"2\">p</td><td>(</td><td>c</td><td>2</td><td>),</td><td>if</td><td colspan=\"3\">there</td><td>are</td><td>double</td><td>given</td><td>names.</td></tr></table>",
+                "num": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-1005.json

@@ -0,0 +1,352 @@
+{
+    "paper_id": "O00-1005",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:02.977719Z"
+    },
+    "title": "Similarity Measure in Backward Transliteration between Different Character Sets and Its Application to CLIR",
+    "authors": [
+        {
+            "first": "Wei-Hao",
+            "middle": [],
+            "last": "Lin",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Taiwan University Taipei",
+                "location": {
+                    "country": "TAIWAN, R.O.C"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Hsin-Hsi",
+            "middle": [],
+            "last": "Chen",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Taiwan University Taipei",
+                "location": {
+                    "country": "TAIWAN, R.O.C"
+                }
+            },
+            "email": "hh_chen@csie.ntu.edu.tw"
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "This paper classifies the problem of machine transliteration into four types, i.e., forward/backward transliteration between same/different character sets, based on transliteration direction and character sets. A phoneme-based similarity measure is proposed to deal with backward transliteration between different character sets. Chinese-English information retrieval is taken as an example. The experiments show that phoneme-based approach is better than grapheme-based approach. In a mate matching of 1,261 candidates, the average rank is 7.80 and 57.65% of candidates are ranked as number one.",
+    "pdf_parse": {
+        "paper_id": "O00-1005",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "This paper classifies the problem of machine transliteration into four types, i.e., forward/backward transliteration between same/different character sets, based on transliteration direction and character sets. A phoneme-based similarity measure is proposed to deal with backward transliteration between different character sets. Chinese-English information retrieval is taken as an example. The experiments show that phoneme-based approach is better than grapheme-based approach. In a mate matching of 1,261 candidates, the average rank is 7.80 and 57.65% of candidates are ranked as number one.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": ")\u4f86\u8aaa\uff0c\u56e0\u70ba\u8a9e\u8a00\u4e0d\u540c\u6240\u5f62\u6210 \u7684\u95b1\u8b80\u8207\u8655\u7406\u969c\u7919\uff0c\u4e5f\u65e5\u6f38\u589e\u52a0\u3002\u5728\u9019\u7a2e\u591a\u8a9e\u7684\u5927\u74b0\u5883\u4e0b\uff0c\u6a5f\u5668\u7ffb\u8b6f(machine translation)\u8207\u8de8\u8a9e\u8a00\u8cc7\u8a0a\u6aa2\u7d22(cross language information retrieval)\u7b49\u76f8\u95dc\u81ea\u7136\u8a9e \u8a00\u8655\u7406\u7cfb\u7d71\u7814\u7a76\uff0c\u5c31\u6975\u70ba\u53d7\u5230\u91cd\u8996\u3002 \u6240\u8b02\u8de8\u8a9e\u8a00\u8cc7\u8a0a\u6aa2\u7d22 (Chen, 1997) (2) P(e|w)\uff1a\u82f1\u6587\u8a5e\u5f59\u767c\u97f3\u3002",
+                "cite_spans": [
+                    {
+                        "start": 145,
+                        "end": 157,
+                        "text": "(Chen, 1997)",
+                        "ref_id": "BIBREF1"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u7cfb \u7d71 \u3002 \u5728 \u5c07 \u82f1 \u6587 \u5f62 \u7d20 \u8f49 \u6210 \u97f3 \u7d20 \u7684 \u904e \u7a0b \u4e2d \uff0c \u9019 \u500b \u7cfb \u7d71 \u5148 \u5c07 \u82f1 \u6587 \u5b57 \u6bcd \u97f3 \u7bc0 \u5316",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(3) P(j|e)\uff1a\u5c07\u82f1\u6587\u767c\u97f3\u8f49\u6210\u65e5\u6587\u767c\u97f3\u3002 (4) P(k|j)\uff1a\u5c07\u65e5\u6587\u767c\u97f3\u8f49\u6210\u7247\u5047\u540d\u3002 5 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u2211 = l i i S i S s 1 ' 2 ' 1 )) ( ), ( ( \u3002 \u6211\u5011\u4ee5\u4e00\u500b\u4f8b\u5b50\u8aaa\u660e\u4e0a\u8ff0\u5b9a\u7fa9\u3002\u5982\u524d\u4f8b\uff0c \u300c\u4e9e\u745f\u300d\u7d93\u67e5\u8868\u5f8c\uff0c\u5f97\u5230\u7684\u767c\u97f3\u662f \u300cIY AA S r\u300d \uff0c\u800c Arthur \u7684\u767c\u97f3\u70ba \u300cAA R TH ER\u300d \uff0c\u6240\u4ee5\u6b64\u6642\u7684\u03a3={AA, ER, IY, R, r, S, TH}\uff0c\u800c\u97f3\u7d20\u9593\u5f7c\u6b64\u7684\u5206\u6578\uff0c\u4ee5\u4e0b\u9762\u7684\u5c0d\u7a31\u77e9\u9663\u8868\u793a\uff1a",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "S AA ER IY R r S TH _ AA 5 0 0 -10 0 -10 -10 -5 ER 0 5 0 -10 8 -10 -10 -5 IY 0 0 5 -10 0 -10 -10 -5 R -10 -10 -10 10 -10 -10 -10 -5 r 0 8 0 -10 5 -10 -10 -5 S -10 -10 -10 -10 -10 10 8 -5 TH -10 -10 -10 -10 -10 8 10 -5 _ -5 -5 -5 -5 -5 -5 -5 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u7d66\u5b9a\u4e00\u500b\u5b57\u6bcd\u96c6\u5408\u03a3'\uff0c\u548c\u6210\u5c0d\u7684\u5206\u6578\u77e9\u9663\u3002\u5b57\u4e32 S 1 \u8207 S 2 \u7684\u76f8\u4f3c\u5ea6\uff0c\u5b9a\u7fa9 \u6210 S 1 \u8207 S 2 \u7684\u6700\u4f73\u5c0d\u9f4a\u65b9\u5f0f A \u7684\u503c\uff0c\u4e5f\u5c31\u662f\u6700\u5927\u7684\u5b57\u4e32\u5c0d\u9f4a\u76f8\u4f3c\u5ea6\u503c\u3002 \u76f8\u4f3c\u5ea6\u8ddf\u76f8\u95dc\u7684\u6700\u4f73\u5c0d\u9f4a\u65b9\u5f0f\uff0c\u53ef\u4ee5\u7528 dynamic programming \u7684\u65b9\u5f0f\u4f86\u627e\u51fa\u3002 Gusfield (1997)\u66fe\u5b9a\u7fa9\u57fa\u5e95\u689d\u4ef6(base condition)\u70ba )) ( (_, ) , 0 ( _) ), ( ( ) 0 , ( 1 2 1 1 k S s j V k S s i V j k i k \u2211 = \u2211 = \u2264 \u2264 \u2264 \u2264 \u4e00\u822c\u7684 recurrence \u5f0f\u53ef\u4ee5\u5beb\u6210\uff1a ))] ( (_, ) 1 , ( _), ), ( ( ) , 1 ( )), ( ), ( ( ) 1 , 1 ( max[ ) , ( 2 1 2 1 j S s j i V i S s j i V j S i S s j i V j i V + \u2212 + \u2212 + \u2212 \u2212 = 0\u2266i\u2266length(S 1 )\uff0c0\u2266j\u2266length(S 2 )\uff0cV(0, 0) = 0\u3002\u5176\u4e2d V(i, j)\u70ba S 1 [1..i]\u8207 S 2 [1..j] \u9019\u5169\u500b\u524d\u5b57\u4e32(prefix)\uff0c\u6700\u4f73\u5c0d\u9f4a\u65b9\u5f0f\u7684\u503c\u3002\u5047\u8a2d S 1 \u8207 S 2 \u7684\u9577\u5ea6\u5404\u70ba n \u8207 m\uff0c\u5247 \u6700\u4f73\u5c0d\u9f4a\u65b9\u5f0f\u7684\u503c\u5c31\u662f V(n, m)\u3002\u5982\u679c\u5229\u7528 dynamic programming \u7684\u65b9\u5f0f\u4f86\u6c42\uff0c\u9019 \u500b\u503c\u53ef\u4ee5\u5728 O(nm)\u7684\u6642\u9593\u5167\u7b97\u51fa\u4f86\u3002 \u5728\u6211\u5011\u7684\u53cd\u5411\u7570\u6587\u5b57\u97f3\u8b6f\u8a9e\u97f3\u76f8\u4f3c\u5ea6\u8a55\u91cf\u4e2d\uff0c\u03a3'\u70ba 63 \u500b IPA \u97f3\u6a19\u7b26\u865f(\u542b \u7a7a\u767d)\uff0c\u5176\u4e2d\u82f1\u6587\u6709 39 \u500b\uff0c\u4e2d\u6587\u9664\u4e86\u5171\u7528\u7684\u4e4b\u5916\uff0c\u53e6\u5916\u9084\u6709 24 \u500b\u4e2d\u6587\u6240\u7368\u7528\u7684 \u7b26\u865f\uff0c\u6240\u4ee5\u6574\u500b\u5206\u6578\u77e9\u9663\u7684\u5927\u5c0f\u70ba 63x63\u3002\u6211\u5011\u5c0d\u5206\u6578\u77e9\u9663\u4e2d\u7684\u5206\u6578\u6307\u5b9a\u65b9\u5f0f\u5982 \u4e0b\uff1a (1) \u539f\u5247\u4e0a\uff0cIPA \u5339\u914d(match)\u7d66 10 \u5206\uff0c\u4e0d\u5339\u914d(mismatch)\u6263 10 \u5206\u3002\u4f46\u82e5\u5339 \u914d\u7684\u70ba\u6bcd\u97f3\uff0c\u5247\u53ea\u7d66 5 \u5206\uff0c\u800c\u6bcd\u97f3\u4e0d\u5339\u914d\u4e0d\u6263\u5206\u3002\u9019\u88e1\u6211\u5011\u5e0c\u671b\u5229\u7528\u6bcd\u97f3\u4f86\u6355\u6349 \u97f3\u7bc0\u7684\u5c0d\u9f4a\uff0c\u6240\u4ee5\u6bcd\u97f3\u4e0d\u5c0d\u9f4a\u4e0d\u6263\u5206\u3002\u4f46\u7531\u65bc\u6bcd\u97f3\u5728\u4e0d\u540c\u8a9e\u8a00\u9593\u7684\u5339\u914d\uff0c\u610f\u7fa9\u8f03 \u4e0d\u986f\u8457\uff0c\u56e0\u6b64\u76f8\u540c\u7684\u6bcd\u97f3\u53ea\u7d66 5 \u5206\u3002 (2) \u8207\u7a7a\u767d\u5b57\u5143(_)\u5c0d\u9f4a\uff0c\u53ef\u4ee5\u770b\u505a insertion \u6216\u662f deletion\u3002\u7531\u65bc\u4e0d\u5339\u914d\u53ef\u4ee5 \u770b\u6210\u662f\u4e00\u500b insertion \u52a0\u4e0a\u4e00\u500b deletion\u3002\u4f8b\u5982 abcdfgh \u548c abcdigh\uff0c\u5176\u4e2d f \u8207 i \u672a \u5339\u914d\uff0c\u7576\u8981\u5c0d\u9f4a\u6642\uff0c\u53ef\u4ee5\u63a1\u7528\u5982\u4e0b\u7684\u65b9\u5f0f\uff1a abcdf_gh abcd_igh \u6240\u4ee5\u672a\u5339\u914d\u8981\u6263\u7684\u5206\u6578\uff0c\u8ddf\u5169\u500b\u5b57\u5143\u5c0d\u4e0a\u7a7a\u767d\uff0c\u4ea6\u5373\u505a\u4e00\u6b21 insertion \u548c\u4e00\u6b21 deletion \u8981\u76f8\u540c\uff0c\u9019\u6a23\u624d\u6c92\u6709\u504f\u597d\u3002\u56e0\u6b64\uff0c\u70ba\u4e86\u516c\u5e73\u8d77\u898b\uff0c\u6211\u5011\u8b93 insertion \u6216\u662f",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "deletion \u7684\u6263\u5206\uff0c\u7b49\u65bc\u4e0d\u76f8\u540c\u914d\u5c0d\u7684\u4e00\u534a\uff0c\u4e5f\u5c31\u662f 10/2=5 \u5206\u3002\u53e6\u5916\uff0c\u95dc\u65bc\u7a7a\u767d\u5c0d \u7a7a\u767d\u5206\u6578\u9084\u662f\u8a2d-5(\u53c3\u8003\u5206\u6578\u77e9\u9663\u7bc4\u4f8b)\uff0c\u539f\u56e0\u662f\u5169\u500b\u5b57\u4e32 ab \u8207 ac \u5728\u5c0d\u9f4a\u6642\uff0c\u5982 \u679c a \u5c0d a \u5339\u914d\u7d66 10 \u5206\uff0c\u4e0d\u5339\u914d\u6263 10 \u5206\uff0c\u5247 ab \u548c ac \u5b57\u4e32\u5c0d\u9f4a\u76f8\u4f3c\u5ea6\u70ba\uff1a10 + (-10) = 0 \u5206\u3002\u5982\u679c\u52a0\u4e0a\u7a7a\u767d\uff0c\u518d\u9032\u884c\u5c0d\u9f4a\uff0c\u5982 a_b \u548c a_c\uff0c\u9019\u6a23\u7684\u5206\u6578\u70ba 10 + (-5) + (-10) ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "= -5 \u5206\u3002\u4e5f\u5c31\u662f\u5728\u5c0d\u5217\u6642\uff0c\u540c\u6642\u52a0\u4e0a\u7a7a\u767d\u662f\u6c92\u6709\u7528\u7684\uff0c\u53ea\u662f\u6703\u628a\u5206\u6578\u62c9\u4f4e\uff0c\u6240 \u4ee5\u7a7a\u767d\u5c0d\u7a7a\u767d\u662f-5 \u5206\u3002 (3) \u5176\u4ed6\u6839\u64da\u767c\u97f3\u4f4d\u7f6e\u8207\u767c\u97f3\u65b9\u5f0f\u7684\u76f8\u8fd1\uff0c\u4e2d\u82f1\u6587\u5728\u97f3\u8b6f\u4e0a\u7684\u7fd2\u6163\u3001\u4e2d\u82f1\u6587 \u5404\u81ea\u7684\u767c\u97f3\u7279\u6027\u3001\u5c07\u67d0\u4e9b\u97f3\u6a19\u4e4b\u9593\u7684\u914d\u5c0d\u5206\u6578\u8a2d\u70ba 8 \u5206\uff0c\u5982\u8868\u4e09\u6240\u5217\u3002 \u8868\u4e09\u2022\u5176\u4ed6\u97f3\u6a19\u4e4b\u914d\u5c0d \u7406\u7531 \u4f8b\u5b50 \u4e2d\u6587\u4e0d\u5206\u6e05\u6fc1 P\u8207 B\u3001D \u8207 T\u3001F \u8207 V\u3001G \u8207 K\u3001S \u8207 Z \u767c\u97f3\u65b9\u5f0f\u8207\u4f4d\u7f6e\u76f8\u8fd1 B \u8207 Ph\u3001K \u8207 Kh\u3001D \u8207 Th\u3001P \u8207 Ph \u767c\u97f3\u4f4d\u7f6e\u76f8\u8fd1 L\u8207 R\u3001DH \u8207 Th \u767c\u97f3\u65b9\u5f0f\u76f8\u8fd1 CH \u8207 Tch\u3001CH \u8207 TSch\u3001H \u8207 Th\u3001G \u8207 Tc\u3001JH \u8207 Tc\uff0cL \u8207 R\u3001M \u8207 ANG\u3001N \u8207 AN\u3001N \u8207 AHN\u3001N \u8207 ANG\u3001NG \u8207 ANG\u3001NG \u8207 AN\u3001NG \u8207 AHNG\u3001S \u8207 Sc\u3001S \u8207 c\u3001S \u8207 TH\u3001S \u8207 TS\u3001Z \u8207 Sc\u3001Z \u8207 TS\u3001Z \u8207 TSc \u97f3\u8b6f\u7fd2\u6163\u4ee5\u53ca\u8de8\u8a9e\u8a00 \u6240\u9020\u6210\u7684\u97f3\u6a19\u7a7a\u7f3a K \u8207 Tc\u3001L \u8207 e\u3001R \u8207 e\u3001TH \u8207 Th\u3001ZH \u8207 Tch\u3001ER \u8207 r\u3001 ER \u8207 L\u3001ER \u8207 e\u3001UW \u8207 V\u3001JH \u8207 TSc\u3001G \u8207 Tch \u4e2d\u6587\u4e0d\u5206\u9577\u77ed\u6bcd\u97f3 IH \u8207 IY\u3001UW \u8207 W \u534a\u6bcd\u97f3\u8207\u6bcd\u97f3 IY \u8207",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Cross-Language Information Retrieval",
+                "authors": [
+                    {
+                        "first": "Hsin-Hsi",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of ROCLING Workshop on ED/MT/IR, Academic Sinica",
+                "volume": "",
+                "issue": "",
+                "pages": "4--5",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, Hsin-Hsi (1997) \"Cross-Language Information Retrieval,\" Proceedings of ROCLING Workshop on ED/MT/IR, Academic Sinica, Taipei, June 2, 1997, pp. 4-1~4-27.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Proper Name Translation in Cross-Language Information Retrieval",
+                "authors": [
+                    {
+                        "first": "Yung-Wei",
+                        "middle": [],
+                        "last": "Ding",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Shih-Chung",
+                        "middle": [],
+                        "last": "Tsai",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Tsai",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of 17 th International Conference on Computational Linguistics and 36 th Annual Meeting of the Association for Computational Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "232--236",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ding, Yung-Wei and Tsai, Shih-Chung Tsai (1998) \"Proper Name Translation in Cross-Language Information Retrieval,\" Proceedings of 17 th International Conference on Computational Linguistics and 36 th Annual Meeting of the Association for Computational Linguistics, Montreal, Quebec, Canada, August 10-14 1998, pp. 232-236.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "A Multilingual News Summarizer",
+                "authors": [
+                    {
+                        "first": "Hsin-Hsi And",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Chuan-Jie",
+                        "middle": [],
+                        "last": "Lin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Proceedings of 18th International Conference on Computational Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, Hsin-Hsi and Lin, Chuan-Jie (2000) \"A Multilingual News Summarizer,\" Proceedings of 18th International Conference on Computational Linguistics, July 31-August 4 2000, University of Saarlandes.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Machine Transliteration",
+                "authors": [
+                    {
+                        "first": "Kevin",
+                        "middle": [],
+                        "last": "Knight",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Jonathan",
+                        "middle": [],
+                        "last": "Graehl",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Computational Linguistics",
+                "volume": "24",
+                "issue": "4",
+                "pages": "599--612",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Knight, Kevin and Graehl, Jonathan (1998) \"Machine Transliteration,\" Computational Linguistics, Vol. 24, No. 4, 1998, pp. 599-612.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "The Art of Computer Programming",
+                "authors": [
+                    {
+                        "first": "Donald",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Knuth",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1973,
+                "venue": "",
+                "volume": "3",
+                "issue": "",
+                "pages": "391--392",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Knuth, Donald E. (1973) The Art of Computer Programming, Volume 3, Sorting and Searching, Addison-Wesley, Reading, Mass, 1973, pp. 391-392.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Name Searching and Information Retrieval",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Thompson",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Dozier",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of Second Conference on Empirical Methods in Natural Language Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Thompson, P. and Dozier, C. (1997) \"Name Searching and Information Retrieval,\" Proceedings of Second Conference on Empirical Methods in Natural Language Processing, Providence, Rhode Island, 1997.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Automatic English-Chinese Name Transliteration for Development of Multilingual Resources",
+                "authors": [
+                    {
+                        "first": "Stephen",
+                        "middle": [],
+                        "last": "Wan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Cornelia",
+                        "middle": [],
+                        "last": "Verspoor",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Maria",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of 17 th COLING and 36 th ACL",
+                "volume": "",
+                "issue": "",
+                "pages": "1352--1356",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Wan, Stephen and Verspoor, Cornelia Maria (1998) \"Automatic English-Chinese Name Transliteration for Development of Multilingual Resources,\" Proceedings of 17 th COLING and 36 th ACL, 1998, pp. 1352-1356.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "text": "p\uff0c\u5982\u679c\u6211\u5011\u60f3\u8981\u627e\u51fa\u9019\u500b\u767c\u97f3\u53ef\u80fd\u7684\u82f1\u6587\u5b57 \u6642\uff0c\u6211\u5011\u5c31\u53ef\u4ee5\u5c0b\u627e\u770b\u770b\u54ea\u4e00\u500b\u82f1\u6587\u5b57 w \u53ef\u4ee5\u8b93 P(w|p)\u9019\u500b\u6a5f\u7387\u6709\u6700\u5927\u503c\u3002\u6839 \u64da\u8c9d\u5f0f\u5b9a\u7406(Bayes' Theorem) \uff0c\u9019\u76f8\u7576\u65bc\u5c0b\u627e P(w)\u2022P(p|w)\u3002\u9019\u500b\u7cfb\u7d71\u7528\u5230\u5982\u4e0b \u4e94\u500b\u6a5f\u7387\u5206\u4f48\uff0c\u5176\u4e2d w \u70ba\u82f1\u6587\u5b57\u3001e \u70ba\u82f1\u6587\u767c\u97f3\u3001j \u70ba\u65e5\u6587\u767c\u97f3\u3001k \u70ba\u7247\u5047\u540d\u3001o \u70ba\u5149\u5b78\u8fa8\u8b58\u51fa\u4f86\u7684\u5b57\u5143\uff1a (1) P(w)\uff1a\u7522\u751f\u82f1\u6587\u8a5e\u5f59\u3002",
+                "uris": null,
+                "type_str": "figure",
+                "num": null
+            },
+            "TABREF1": {
+                "content": "<table><tr><td>\u97f3\u8b6f\u901a\u5e38\u4fdd\u6301\u539f\u59cb\u8a9e\u8a00\u7684\u6587\u5b57\u62fc\u6cd5\uff0c\u800c\u76ee\u7684\u8a9e\u8a00\u7684\u4f7f\u7528\u8005\u5247\u4ee5\u76ee\u7684\u8a9e\u8a00\u7684\u767c\u97f3\u898f</td></tr><tr><td>\u4ee5\u610f\u8b6f\u4f86\u8868\u793a\u6642\uff0c\u6703\u63a1\u7528\u6b63\u5411\u97f3\u8b6f\uff0c\u5c07\u5176\u97f3\u5448\u73fe\u51fa\u4f86\u3002\u4f8b\u5982\u7fa9\u5927\u5229\u7684\u89c0\u5149\u52dd\u5730 \u5247\uff0c\u6216\u662f\u4ee5\u539f\u59cb\u8a9e\u8a00\u7684\u767c\u97f3\u898f\u5247\u4f86\u767c\u97f3\u3002\u4f8b\u5982 Beethoven \u96d6\u7136\u662f\u5fb7\u570b\u540d\u5b57\uff0c\u4f46\u662f</td></tr><tr><td>Firenze\uff0c\u4e2d\u6587\u5c31\u97f3\u8b6f\u6210\u300c\u7fe1\u51b7\u7fe0\u300d \uff0c\u6b64\u70ba\u6b63\u5411\u97f3\u8b6f\u3002\u53cd\u904e\u4f86\u8aaa\uff0c\u7576\u6211\u5011\u770b\u5230\u4e00\u500b \u5728\u82f1\u6587\u7684\u6587\u672c\u4e2d\uff0c\u9084\u662f\u76f4\u63a5\u4f7f\u7528\u76f8\u540c\u7684\u6587\u5b57\u62fc\u6cd5\u3002\u5373\u4f7f\u5728\u4f7f\u7528\u76f8\u540c\u62fc\u97f3\u5b57\u6bcd\u7684\u8a9e</td></tr><tr><td>\u4e2d\u6587\u7684\u97f3\u8b6f\u4eba\u540d\u300c\u963f\u8afe\u53f2\u74e6\u8f9b\u683c\u300d \uff0c\u5982\u679c\u60f3\u8981\u627e\u51fa\u539f\u6587\u662f Arnold Schwarzenegger\uff0c \u8a00\u4e2d\uff0c\u9084\u662f\u53ef\u80fd\u5b58\u5728\u97f3\u8b6f\u3002\u4f8b\u5982\u7fa9\u5927\u5229\u89c0\u5149\u52dd\u5730 Firenze(\u7fa9\u5927\u5229\u6587)\uff0c\u82f1\u6587\u5247\u97f3\u8b6f</td></tr><tr><td>\u5c31\u662f\u53cd\u5411\u97f3\u8b6f\u3002\u4e00\u822c\u4f86\u8aaa\uff0c\u4f7f\u7528\u7f85\u99ac\u5b57\u6bcd\u7684\u62fc\u97f3\u6587\u5b57\u8a9e\u8a00\uff0c\u6703\u4fdd\u6301\u539f\u8a5e\u8a9e\u5b57\u6bcd\u7684 \u70ba Florence\u3002</td></tr><tr><td>\u62fc\u6cd5\uff0c\u7136\u5f8c\u4ee5\u539f\u8a9e\u8a00\u7684\u767c\u97f3\u898f\u5247\uff0c\u6216\u662f\u81ea\u5df1\u8a9e\u8a00\u7684\u767c\u97f3\u898f\u5247\u4f86\u767c\u97f3\u3002\u4f46\u5982\u679c\u5728\u8c61 \u4e0d\u540c\u8a9e\u8a00\u4f7f\u7528\u8005\u5728\u767c\u97f3\u6642\uff0c\u6703\u63a1\u7528\u81ea\u5df1\u8a9e\u8a00\u7684\u767c\u97f3\u898f\u5247\u3002\u4f8b\u5982\u82f1\u8a9e\u4f7f\u7528\u8005\u53ef</td></tr><tr><td>\u5f62\u6587\u5b57\u8207\u62fc\u97f3\u6587\u5b57\u8a9e\u8a00\u4e4b\u9593\u4f5c\u97f3\u8b6f\u6642\uff0c\u5247\u9700\u8981\u5c07\u8072\u97f3\u7531\u539f\u8a9e\u8a00\u76e1\u91cf\u7528\u53e6\u5916\u4e00\u7a2e\u8a9e \u80fd\u6703\u4f9d\u82f1\u8a9e\u7684\u767c\u97f3\u898f\u5247\u4f86\u767c\u97f3\uff0c\u9019\u6a23\u5c31\u8ddf\u539f\u4f86\u5fb7\u6587\u7684\u767c\u97f3\u4e0d\u540c\u3002\u4f46\u5927\u9ad4\u4f86\u8aaa\u5728\u97f3</td></tr><tr><td>\u8a00\u76f8\u8fd1\u7684\u97f3\u7d20\u4f86\u8868\u793a\uff0c\u800c\u4e14\u8981\u7b26\u5408\u76ee\u7684\u8a9e\u8a00(target language)\u7684\u8a9e\u97f3\u7d44\u5408\u898f\u5247\u3002\u5f88 \u7d20\u4e0a\u7684\u767c\u97f3\u8f03\u70ba\u63a5\u8fd1\uff0c\u800c\u4e14\u8d8a\u4f86\u8d8a\u591a\u7684\u4eba\u6703\u9078\u64c7\u4ee5\u539f\u59cb\u8a9e\u8a00\u4f86\u767c\u97f3\uff0c\u4ee5\u5c0a\u91cd\u539f\u59cb</td></tr><tr><td>\u986f\u7136\u5730\uff0c\u62fc\u97f3\u6587\u5b57\u8207\u8c61\u5f62\u6587\u5b57\u4e4b\u9593\u7684\u97f3\u8b6f\u8655\u7406\u76f8\u5c0d\u4f86\u8aaa\u8f03\u70ba\u56f0\u96e3\uff0c\u53cd\u5411\u97f3\u8b6f\u6bd4\u6b63 \u8a9e\u8a00\u3002\u53e6\u5916\uff0c\u65e5\u6587\u4e2d\u7684\u6f22\u5b57\u96d6\u7136\u8207\u4e2d\u6587\u76f8\u901a\uff0c\u4f46\u7531\u65bc\u5728\u767c\u97f3\u4e0a\u5dee\u8ddd\u751a\u5927\uff0c\u6240\u4ee5\u901a</td></tr><tr><td>\u5411\u97f3\u8b6f\u66f4\u96e3\u3002\u6b63\u5411\u97f3\u8b6f\u5141\u8a31\u67d0\u7a2e\u7a0b\u5ea6\u7684\u5931\u771f\uff0c\u6240\u80fd\u5920\u63a5\u53d7\u7684\u932f\u8aa4\u7bc4\u570d\u8f03\u5927\uff1b\u4f46\u53cd \u5e38\u65e5\u6587\u6f22\u5b57\u7ffb\u8b6f\u6210\u4e2d\u6587\u6642\uff0c\u8868\u9762\u4e0a\u8207\u7f85\u99ac\u62fc\u97f3\u6587\u5b57\u4e00\u6a23\uff0c\u4fdd\u6301\u539f\u4f86\u65e5\u6587\u6f22\u5b57\u7684\u5beb</td></tr><tr><td>\u5411\u97f3\u8b6f\u5247\u4e0d\u662f\u3002\u53cd\u5411\u97f3\u8b6f\u8f03\u4e0d\u5141\u8a31\u932f\u8aa4\uff0c\u4e5f\u5c31\u662f\u5728\u627e\u51fa\u539f\u6587\u7684\u904e\u7a0b\u4e2d\uff0c\u5fc5\u9808\u8981\u76f8 \u6cd5\uff0c\u4f46\u4e2d\u6587\u4f7f\u7528\u8005\u901a\u5e38\u6703\u4ee5\u4e2d\u6587\u7684\u5ff5\u6cd5\u4f86\u5c0d\u65e5\u6587\u6f22\u5b57\u767c\u97f3\u3002\u9664\u975e\u9019\u4f4d\u4f7f\u7528\u8005\u5b78\u7fd2</td></tr><tr><td>\u7576\u6e96\u78ba\uff0c\u5426\u5247\u53cd\u5411\u97f3\u8b6f\u7684\u7d50\u679c\u61c9\u7528\u6027\u5c31\u8f03\u4f4e\u3002 \u904e\u65e5\u6587\uff0c\u624d\u6709\u8fa6\u6cd5\u4ee5\u6b63\u78ba\u7684\u65e5\u6587\u6f22\u5b57\u4f86\u767c\u97f3\u3002</td></tr><tr><td>\u672c\u6587\u7b2c\u4e8c\u7bc0\u7531\u97f3\u8b6f\u7684\u6b63\u5411\u548c\u53cd\u5411\uff0c\u4ee5\u53ca\u662f\u5426\u8de8\u6587\u5b57\u4f86\u5206\u6790\u97f3\u8b6f\u554f\u984c\uff0c\u4e26\u4ecb\u7d39 2-2 \u6b63\u5411\u7570\u6587\u5b57\u9593\u97f3\u8b6f</td></tr><tr><td>\u904e\u53bb\u76f8\u95dc\u7684\u7814\u7a76\u3002\u7b2c\u4e09\u7bc0\u7531\u76f8\u4f3c\u5ea6\u7684\u89c0\u5ff5\uff0c\u4f86\u57f7\u884c\u6a5f\u5668\u53cd\u5411\u97f3\u8b6f\u7684\u7a0b\u5e8f\u3002\u7b2c\u56db\u7bc0 \u5728\u6b63\u5411\u7570\u6587\u5b57\u9593\u97f3\u8b6f\u6642\uff0c\u4e3b\u8981\u7684\u5de5\u4f5c\u5728\u65bc\u5c07\u539f\u59cb\u8a9e\u8a00\u7684\u97f3\u7d20\uff0c\u4ee5\u76ee\u7684\u8a9e\u8a00\u7684</td></tr><tr><td>\u63d0\u51fa\u4e00\u7a2e\u4ee5\u97f3\u7d20\u9032\u884c\u76f8\u4f3c\u5ea6\u6bd4\u8f03\u7684\u65b9\u6cd5\u3002\u7b2c\u4e94\u7bc0\u4ecb\u7d39\u5be6\u9a57\u898f\u756b\uff0c\u4e26\u5c0d\u5be6\u9a57\u7d50\u679c\u9032 \u97f3\u7d20\u4f86\u5448\u73fe\uff0c\u4e26\u914d\u5408\u76ee\u7684\u8a9e\u8a00\u7684\u7d44\u5408\u898f\u5247\u8868\u793a\u3002\u5982\u679c\u61c9\u7528\u5728\u66f8\u5beb\u7cfb\u7d71\u4e0a\uff0c\u9084\u8981\u9032</td></tr><tr><td>\u884c\u8a0e\u8ad6\uff0c\u6700\u5f8c\u662f\u7d50\u8ad6\u3002 \u4e00\u6b65\u5c07\u4e4b\u524d\u97f3\u8b6f\u5f8c\u7684\u7d50\u679c\uff0c\u9078\u64c7\u76ee\u7684\u8a9e\u8a00\u9069\u7576\u7684\u66f8\u5beb\u6587\u5b57\uff0c\u4f86\u5448\u73fe\u6700\u5f8c\u97f3\u8b6f\u7684\u7d50</td></tr><tr><td>2. \u97f3\u8b6f\u5206\u985e\u8207\u76f8\u95dc\u7814\u7a76 \u679c\u3002Wan \u8207 Verspoor(1998)\u767c\u5c55\u51fa\u4e00\u5957\u81ea\u52d5\u5c07\u82f1\u6587\u5c08\u6709\u540d\u8a5e\uff0c\u6b63\u5411\u97f3\u8b6f\u6210\u4e2d\u6587\u7684</td></tr><tr><td>\u6839\u64da\u97f3\u8b6f\u7684\u65b9\u5411\uff0c\u6211\u5011\u5c07\u97f3\u8b6f\u554f\u984c\u5340\u5206\u70ba\u300c\u6b63\u5411\u97f3\u8b6f\u300d\u8207\u300c\u53cd\u5411\u97f3\u8b6f\u300d\u5169\u7a2e\u3002</td></tr><tr><td>\u53e6\u5916\uff0c\u6839\u64da\u97f3\u8b6f\u7684\u539f\u59cb\u8a9e\u8a00\u8207\u76ee\u6a19\u8a9e\u8a00\u6240\u63a1\u7528\u7684\u5b57\u6bcd\u7cfb\u7d71\uff0c\u9084\u53ef\u4ee5\u5c07\u97f3\u8b6f\u5340\u5206\u70ba</td></tr><tr><td>\u300c\u540c\u6587\u5b57\u7cfb\u7d71\u9593\u97f3\u8b6f\u300d\u8207\u300c\u7570\u6587\u5b57\u7cfb\u7d71\u9593\u97f3\u8b6f\u300d\u5169\u7a2e\u3002\u4ee5\u4e0b\u5404\u5c0f\u7bc0\uff0c\u5c31\u91dd\u5c0d\u9019\u56db</td></tr><tr><td>\u7a2e\u7d44\u5408\u4f86\u4ecb\u7d39\u76f8\u95dc\u554f\u984c\u3002</td></tr><tr><td>2-1 \u6b63\u5411\u540c\u6587\u5b57\u9593\u97f3\u8b6f</td></tr><tr><td>\u76f8\u540c\u6587\u5b57\u7cfb\u7d71\u4e4b\u9593\u7531\u65bc\u5171\u7528\u540c\u4e00\u7a2e\u6587\u5b57\uff0c\u5c24\u5176\u4ee5\u7f85\u99ac\u5b57\u6bcd\u70ba\u57fa\u790e\u7684\u62fc\u97f3\u6587</td></tr><tr><td>\u5b57\uff0c\u4e0d\u540c\u8a9e\u8a00\u5728\u5f62\u7d20(grapheme)\u8207\u97f3\u7d20(phoneme)\u9593\u7684\u7d44\u5408\u898f\u5247\u96d6\u4e0d\u4e00\u6a23\uff0c\u4f46\u662f\u4e00</td></tr><tr><td>\u500b\u8a9e\u8a00\u7684\u8a5e\u8a9e\uff0c\u8981\u8868\u9054\u6210\u53e6\u5916\u4e00\u500b\u540c\u6587\u5b57\u7cfb\u7d71\u7684\u8a9e\u8a00\uff0c\u901a\u5e38\u6c92\u6709\u554f\u984c\u3002\u9019\u985e\u578b\u7684</td></tr></table>",
+                "num": null,
+                "html": null,
+                "text": "\u5c31\u662f\u4ee5\u4e00\u7a2e\u8a9e\u8a00\u6240\u8868\u9054\u7684\u67e5\u8a62(query)\uff0c\u53bb \u6aa2\u7d22\u53e6\u4e00\u7a2e\u8a9e\u8a00\u6240\u5448\u73fe\u7684\u5167\u5bb9\u3002\u56e0\u70ba\u8a9e\u8a00\u4e0a\u7684\u5dee\u7570\uff0c\u901a\u5e38\u9700\u8981\u5c07\u67e5\u8a62\u8f49\u63db\u6210\u8ddf\u5167 \u5bb9\u4e00\u6a23\u7684\u8a9e\u8a00\u3002\u6b67\u7fa9\u5206\u6790(disambiguation)\uff0c\u662f\u67e5\u8a62\u7ffb\u8b6f(query translation)\u4e00\u9805\u91cd \u8981\u7684\u7814\u7a76(Bian and Chen, 2000)\u3002\u6839\u64da 1995 \u5e74\u7db2\u8def\u4f7f\u7528\u8005\uff0c\u5c0d Wall Street Journal\u3001 Los Angeles Times \u548c Washington Post \u7b49\u65b0\u805e\u8a9e\u6599\u6aa2\u7d22\u7684\u7d71\u8a08(Thompson and Dozier, 1997)\uff0c\u5206\u5225\u6709 67.8%\u300183.4%\u3001\u548c 38.8%\u7684\u6aa2\u7d22\u8a5e\u542b\u5c08\u6709\u540d\u8a5e\u3002\u6211\u5011\u77e5\u9053 \u8fad\u5178\u7684\u8986\u84cb\u5ea6\uff0c\u4e00\u76f4\u662f\u67e5\u8a62\u7ffb\u8b6f\u7684\u91cd\u8981\u554f\u984c\uff0c\u5728\u5c08\u6709\u540d\u8a5e\u7684\u7ffb\u8b6f\u66f4\u662f\u6311\u6230\u3002Chen \u7b49\u4eba(1998)\uff0cKnight \u548c Graehl(1998)\uff0cWan \u548c Verspoor(1998)\u90fd\u76f8\u7e7c\u63d0\u51fa\u6a5f\u5668\u97f3\u8b6f (machine transliteration)\u7684\u65b9\u6cd5\uff0c\u4f86\u8655\u7406\u9019\u500b\u554f\u984c\u3002 \u97f3\u8b6f\u53ef\u4ee5\u6839\u64da\u8655\u7406\u7684\u65b9\u5411\uff0c\u5340\u5206\u6210\u6b63\u5411\u97f3\u8b6f(forward transliteration)\u8207\u53cd\u5411\u97f3 \u8b6f(backward transliteration)\u3002\u7576\u4e00\u500b\u8a9e\u8a00\u7684\u5c08\u6709\u540d\u8a5e\uff0c\u56e0\u70ba\u6c92\u6709\u9069\u7576\u6216\u662f\u4e0d\u5bb9\u6613",
+                "type_str": "table"
+            },
+            "TABREF2": {
+                "content": "<table><tr><td>\u8207\u767c\u97f3\u76f8\u4f3c\u6642\uff0c\u8868\u793a\u9019\u5169\u500b\u8a5e\u5f59\u7684\u767c\u97f3\u5c31\u53ef\u80fd\u8d8a\u76f8\u4f3c\u3002\u800c Chen \u7b49\u4eba(1998)\u7684\u7814 \u7d71\u4f86\u505a\u5f62\u7d20\u4e0a\u7684\u6bd4\u8f03\u6642\uff0c\u53ef\u4ee5\u767c\u73fe\u5176\u4ed6\u7cfb\u7d71\u5b8c\u5168\u914d\u5c0d\u5931\u6557(er\u2260l) \uff0c\u53ea\u6709\u7d93\u904e\u8a13 \u4e32\u5c0d\u9f4a\u76f8\u4f3c\u5ea6\uff0c\u6700\u5f8c\u5230\u5b57\u4e32\u76f8\u4f3c\u5ea6\u3002</td></tr><tr><td>\u7a76\uff0c\u53ef\u4ee5\u8996\u70ba\u5728\u5f62\u7d20\u4e0a\u6bd4\u8f03\u76f8\u4f3c\u5ea6\u7684\u53cd\u5411\u97f3\u8b6f\u7cfb\u7d71\u3002\u7531\u65bc\u6240\u8a0e\u8ad6\u7684\u4e2d\u6587\u97f3\u8b6f\u5b57\uff0c \u7df4\u968e\u6bb5\u6240\u7522\u751f\u5c0d\u61c9\u624d\u80fd\u6b63\u78ba\u914d\u5c0d(l=l) \u3002\u63db\u53e5\u8a71\u8aaa\uff0c\u9019\u500b\u7cfb\u7d71\u7684\u5c0d\u7167\u8868\u662f\u6bd4\u8f03\u6709 \u7b26\u865f\u548c IPA \u7b26\u865f\u5c0d\u7167\u3002\u4f8b\u5982\uff0c \u300c\u03b9\u300d\u5c0d\u61c9\u300cIY\u300d \u3002 \u5b9a\u7fa9 1\uff1a\u5b57\u6bcd\u5c0d\u9f4a\u76f8\u4f3c\u5ea6</td></tr><tr><td>\u8207\u539f\u59cb\u8a9e\u8a00\u82f1\u6587\u7684\u66f8\u5beb\u7cfb\u7d71\u4e0d\u540c\uff0c\u4ed6\u5011\u5148\u5c07\u97f3\u8b6f\u5b57\u8f49\u63db\u6210\u7f85\u99ac\u5b57\u6bcd\uff0c\u9019\u500b\u52d5\u4f5c\u7a31 \u6548\u7684\uff0c\u6240\u4ee5\u80fd\u5920\u5728\u5f62\u7d20\u5c64\u6b21\u4e0a\u7684\u76f8\u4f3c\u5ea6\u6bd4\u8f03\uff0c\u6709\u66f4\u597d\u7684\u6548\u80fd\u3002 \u82f1\u4e2d\u97f3\u8b6f\u5b57 \u82f1\u6587\u5019\u9078\u5b57 \u5047\u8a2d S 1 \u8207 S 2 \u9019\u5169\u500b\u5b57\u4e32\u7684\u5b57\u6bcd\u96c6\u5408\u70ba\u03a3\uff0c\u800c\u03a3'\u8868\u793a\u03a3\u52a0\u4e0a\u300c_\u300d(_\u8868\u793a\u7a7a</td></tr><tr><td>P(o|k)\uff1a\u52a0\u5165\u56e0\u70ba\u5149\u5b78\u5b57\u5143\u8fa8\u8b58\u6240\u7522\u751f\u7684\u932f\u8aa4\u3002 \u7576 OCR \u53d6\u5f97\u4e00\u500b\u7247\u5047\u540d\u5b57\u4e32 o \u6642\uff0c\u53cd\u5411\u97f3\u8b6f\u4f7f\u7528\u4e0b\u9762\u7684\u516c\u5f0f\uff0c\u627e\u51fa\u82f1\u6587\u5b57\u4e32 w\u3002 \u70ba\u300c\u7f85\u99ac\u62fc\u97f3\u5316\u300d(romanization)\u3002\u4ed6\u5011\u6240\u63a1\u7528\u7684\u6a19\u6e96\u62fc\u97f3\u7cfb\u7d71\uff0c\u6709\u5a01\u7fdf\u8207\u6f22\u8a9e\u62fc \u97f3\uff0c\u4e26\u52a0\u4e0a\u4e00\u4e9b\u7d93\u9a57\u6cd5\u5247\u4fee\u6b63\uff0c\u4f86\u63d0\u9ad8\u7cfb\u7d71\u6548\u80fd\u3002 \u767d\u5b57\u5143)\uff0c\u7d66\u4e88\u03a3'\u4e2d\u7684\u5169\u500b\u5b57\u5143 x \u8207 y\uff0cs(x, y)\u8868\u793a x \u8207 y \u5c0d\u9f4a\u5f8c\uff0c\u6240\u5f97\u5230 \u8868\u4e00\u2022\u8a13\u7df4\u7d50\u679c\u8207\u7f85\u99ac\u62fc\u97f3\u7cfb\u7d71 \u65b7\u5b57 \u767c\u97f3\u67e5\u8868 \u7684\u5206\u6578\uff0c\u7a31\u70ba\u5b57\u6bcd\u5c0d\u9f4a\u76f8\u4f3c\u5ea6\u3002</td></tr><tr><td>) \u8036\u9b6f \u6f22\u8a9e\u62fc\u97f3 \u6ce8\u97f3\u7b26\u865f\u7b2c\u4e8c\u5f0f \u300c \u81ea \u52d5 \u7522 \u751f \u7f85 \u99ac \u62fc \u97f3 | ( ) | ( ) | ( ) | ( k o P j k P e j P w e P \u00d7 \u00d7 \u00d7 \u7531\u65bc\u7f85\u99ac\u62fc\u97f3\u7cfb\u7d71\uff0c\u4e3b\u8981\u4e26\u4e0d\u662f\u8003\u616e\u8a9e\u97f3\u4e0a\u7684\u76f8\u8fd1\u4f86\u8a2d\u8a08\uff0c\u4f8b\u5982\u6f22\u8a9e\u62fc\u97f3\u5c31 ) ( max arg w P w \u6ce8\u97f3\u7b26\u865f \u5a01\u7fdf \u00d7 \u5c0d\u7167\u8868\u300d\u7cfb\u7d71\u7684\u7d50\u679c \u6f22\u5b57\u8f49\u6ce8\u97f3\u7b26\u865f \u570b\u969b\u97f3\u6a19 \u5b9a\u7fa9 2\uff1a\u5b57\u4e32\u5c0d\u9f4a\u76f8\u4f3c\u5ea6</td></tr><tr><td>Chen \u7b49\u4eba(1998)\u63d0\u51fa\u4e00\u500b\u5c07\u82f1\u6587\u97f3\u8b6f\u6210\u4e2d\u6587(\u76ee\u7684\u8a9e\u8a00)\u7684\u97f3\u8b6f\u5b57\uff0c\u53cd\u5411\u97f3\u8b6f erh er Er Er l \u56de\u82f1\u6587(\u539f\u59cb\u8a9e\u8a00)\u7684\u6a21\u7d44\uff0c\u4e26\u61c9\u7528\u65bc\u4e2d\u82f1\u8de8\u8a9e\u8a00\u8cc7\u8a0a\u6aa2\u7d22\u7cfb\u7d71\u3002\u9019\u500b\u7cfb\u7d71\u662f\u5c07\u53ef \u3126 \u7528\u5230\u4e86 Zh\u3001Q \u8207 X \u7b49\u7f85\u99ac\u5b57\u6bcd\uff0c\u4f86\u8868\u793a\u8207\u5b57\u6bcd\u767c\u97f3\u5b8c\u5168\u7121\u95dc\u7684\u6f22\u8a9e\u8a9e\u97f3\uff0c\u6240\u4ee5 \u82f1\u6587\u97f3\u8b6f\u6210\u4e2d\u6587\u7684\u97f3\u8b6f\u5b57\uff0c\u5728\u5229\u7528\u7f85\u99ac\u62fc\u97f3\u7cfb\u7d71\u8f49\u63db\u6210\u7f85\u99ac\u62fc\u97f3\u5b57\u6bcd\u5f8c\uff0c\u9019\u4e9b\u7f85 \u311e ai ai Ai Ai \u5047\u8a2d A \u70ba\u5b57\u4e32 S 1 \u8207 S 2 \u7684\u67d0\u4e00\u7a2e\u5c0d\u9f4a\u65b9\u5f0f(alignment)\uff0cS 1 '\u8207 S 2 '\u70ba\u63d2\u5165\u7a7a\u767d e \u6ce8\u97f3\u7b26\u865f\u8f49 IPA \u8a9e\u97f3\u6bd4\u8f03 \u5f8c\u7684\u5b57\u4e32\u3002\u5982\u679c S 1 '\u8207 S 2 '\u7684\u9577\u5ea6\u70ba l\uff0c\u5247\u5c0d\u9f4a\u65b9\u5f0f A \u7684\u5206\u6578\u5982\u4e0b\uff1a</td></tr><tr><td>\u80fd\u7684\u97f3\u8b6f\u5b57\u8fa8\u8b58\u51fa\u4f86\uff0c\u518d\u9032\u884c\u53cd\u5411\u97f3\u8b6f\u3002\u9996\u5148\u5229\u7528\u6f22\u5b57\u7f85\u99ac\u62fc\u97f3\u7cfb\u7d71(\u4f8b\u5982 Wade \u99ac\u62fc\u97f3\u5b57\u6bcd\uff0c\u8ddf\u539f\u4f86\u8a5e\u5f59\u7684\u62fc\u97f3\u5b57\u6bcd\uff0c\u5728\u767c\u97f3\u4e0a\u4e26\u4e0d\u5341\u5206\u76f8\u8fd1\u3002 \u3109\u3128\u3123 tun dwei Duan duan don \u570b\u969b\u97f3\u6a19 \u76f8\u4f3c\u5ea6</td></tr><tr><td>Giles (\u5a01\u7fdf)\uff0c\u6216\u662f\u6f22\u8a9e\u62fc\u97f3(Pinyin))\uff0c\u628a\u53ef\u80fd\u7684\u97f3\u8b6f\u5b57(\u4e2d\u6587)\u8f49\u6210\u7f85\u99ac\u5b57\u6bcd\u3002\u63a5 \u6709\u9451\u65bc\u5728\u5f62\u7d20\u5c64\u6b21\u4e0a\u505a\u7f85\u99ac\u62fc\u97f3\u5316\u6642\uff0c\u975e\u5e38\u9700\u8981\u4e00\u500b\u4ee5\u5f62\u7d20\u76f8\u8fd1\u70ba\u51fa\u767c\u9ede\u800c \u5716\u4e00\u2022\u97f3\u7d20\u6bd4\u5c0d</td></tr><tr><td>\u8457\u5c07\u9019\u500b\u8a5e\u5f59\u8207\u4e00\u4e32\u53ef\u80fd\u7684\u5c08\u6709\u540d\u8a5e\u9032\u884c\u6bd4\u5c0d\uff0c\u85c9\u6b64\u627e\u51fa\u53ef\u80fd\u7684\u539f\u6587(\u82f1\u6587)\u3002 \u8a2d\u8a08\u7684\u7684\u7f85\u99ac\u62fc\u97f3\u7cfb\u7d71\u3002\u4f8b\u5982\u5728\u4e2d\u6587\u548c\u82f1\u6587\u9019\u5169\u7a2e\u66f8\u5beb\u7cfb\u7d71\u5b8c\u5168\u4e0d\u540c\u7684\u8a9e\u8a00\uff0c\u6211 4. \u97f3\u7d20\u76f8\u4f3c\u5ea6\u8a55\u91cf \u8868\u4e8c\u2022\u6ce8\u97f3\u7b26\u865f\u8207 IPA \u5c0d\u7167\u8868</td></tr><tr><td>3. \u8a9e\u97f3\u76f8\u4f3c\u5ea6 \u5011\u53ef\u4ee5\u8a2d\u8a08\u4e00\u500b\u300c\u81ea\u52d5\u5efa\u7acb\u7f85\u99ac\u62fc\u97f3\u5c0d\u7167\u8868\u300d\u7684\u7cfb\u7d71\u3002\u9019\u500b\u7cfb\u7d71\u5206\u70ba\u5169\u500b\u968e\u6bb5\uff1a \u8003\u91cf\u7269\u7406\u767c\u97f3\u5728\u8de8\u8a9e\u8a00\u8cc7\u8a0a\u6aa2\u7d22\u7684\u5be6\u7528\u6027\uff0c\u4ee5\u53ca\u5f62\u7d20\u5c64\u6b21\u4e0a\u6bd4\u5c0d\u7684\u4e8b\u524d\u8a13 (a) \u5b50\u97f3\u90e8\u4efd</td></tr><tr><td>\u7b2c\u4e00\u500b\u968e\u6bb5\u662f\u8a13\u7df4\uff0c\u6211\u5011\u5f9e\u5df2\u77e5\u7684\u82f1-\u4e2d\u97f3\u8b6f\u5b57\u8207\u539f\u6587\u8a5e\u5f59\u7684\u914d\u5c0d\u4e2d\uff0c\u5b78\u7fd2\u82f1\u4e2d \u6ce8\u97f3\u7b26\u865f \u3105 \u3106 \u3107 \u3108 \u3109 \u310a \u310b \u310c \u310d \u310e \u310f \u3110 \u3111 \u3112 \u3113 \u3114 \u7df4\uff0c\u56e0\u6b64\u97f3\u7d20\u5c64\u6b21\u4e0a\u7684\u76f8\u4f3c\u5ea6\u6bd4\u8f03\u6613\u986f\u91cd\u8981\u3002\u800c\u8861\u91cf\u5169\u500b\u8a5e\u5f59\u7684\u97f3\u7d20\u76f8\u4f3c\u5ea6\uff0c\u6211 \u672c\u7bc7\u8ad6\u6587\u628a\u97f3\u8b6f\u554f\u984c\u8996\u70ba\u76f8\u4f3c\u5ea6\u7684\u8861\u91cf\u3002\u6b63\u5411\u97f3\u8b6f\u5373\u662f\u5728\u4e0d\u540c\u8a9e\u8a00\u4e4b\u9593\uff0c\u8b93 \u97f3\u8b6f\u5f8c\u7684\u7d50\u679c\u80fd\u5920\u4fdd\u6301\u6700\u5927\u7684\u76f8\u4f3c\u5ea6\u3002\u5728\u53cd\u5411\u97f3\u8b6f\uff0c\u5982\u679c\u9810\u5148\u7d66\u4e88\u4e00\u4efd\u5019\u9078\u540d \u97f3\u8b6f\u5b57\u6240\u61c9\u8a72\u8f49\u63db\u7684\u7f85\u99ac\u62fc\u97f3\u5b57\u6bcd\u3002\u4f8b\u5982 Elton \u8207\u300c\u611b\u723e\u9813\u300d\u9019\u500b\u914d\u5c0d\uff0c\u5148\u5c07\u4e2d IPA \u03c0 \u03c0\u0397 \u00b5 \u03c6 \u03c4 \u03c4\u0397 \u03bd \u03bb \u03ba \u03ba\u0397 \u03be \u03c4\uf8fe \u03c4\uf8fe\u0397 \uf8fe \u03c4\u2663 \u03c4\u2663\u0397 \u5011\u63d0\u51fa\u4e00\u500b\u4ee5\u570b\u969b\u97f3\u6a19(International Phonetic Alphabet\uff0cIPA)\u70ba\u57fa\u6e96\u7684\u6bd4\u8f03\uff0c\u5148\u5c07 \u6ce8\u97f3\u7b26\u865f \u3115 \u3116 \u3117 \u3118 \u3119</td></tr><tr><td>\u55ae\uff0c\u5247\u7cfb\u7d71\u6bd4\u8f03\u97f3\u8b6f\u5b57\u8207\u5019\u9078\u540d\u55ae\u4e0a\u7684\u8a5e\u5f59\uff0c\u8a08\u7b97\u5169\u5169\u76f8\u4f3c\u5ea6\u3002\u76f8\u4f3c\u5ea6\u7684\u6bd4\u5c0d\uff0c \u6587\u4ee3\u63db\u6210\u6ce8\u97f3\u7b26\u865f\u5f8c\uff0c\u7136\u5f8c\u5206\u5225\u5c0d\u5169\u500b\u5b57\u505a\u97f3\u7bc0\u62c6\u89e3\u7684\u52d5\u4f5c\uff0c\u5f97\u5230\u300cEl\u2022ton\u300d \u5169\u500b\u8a5e\u5f59\u7684\u570b\u969b\u97f3\u6a19\u5217\u51fa\u4f86\uff0c\u7136\u5f8c\u6bd4\u8f03\u570b\u969b\u97f3\u6a19\u7684\u76f8\u4f3c\u5ea6\uff0c\u9032\u800c\u9054\u5230\u53cd\u5411\u97f3\u8b6f\u7684 IPA \u2663 \uf8e6 \u03c4\u03c3 \u03c4\u03c3\u0397 \u03c3</td></tr><tr><td>\u8207\u300c\u311e\u2022\u3126\u2022\u3109\u3128\u3123\u300d \uff0c\u9019\u88e1\u5ffd\u7565\u82f1\u6587\u91cd\u97f3\u8207\u4e2d\u6587\u8072\u8abf\u7b26\u865f\uff0c\u800c\u2022\u70ba\u97f3\u7bc0\u9593\u9694\u7b26 \u76ee\u7684\u3002\u5716\u4e00\u986f\u793a\u97f3\u7d20\u76f8\u4f3c\u5ea6\u6bd4\u8f03\u7684\u6d41\u7a0b\u3002\u6211\u5011\u5148\u8aaa\u660e\u6d41\u7a0b\u5716\u5de6\u908a\u7684\u90e8\u5206\uff0c\u4e5f\u5c31\u662f \u53ef\u4ee5\u5206\u6210\u4e09\u500b\u5c64\u6b21\uff1a\u5f62\u7d20\u3001\u97f3\u7d20\u3001\u548c\u7269\u7406\u8072\u97f3\u3002 \u865f\u3002\u63a5\u8457\u9032\u4e00\u6b65\u5c07\u82f1\u6587\u97f3\u7bc0\u505a\u6b21\u97f3\u7bc0\u5316\u5f8c\uff0c\u6211\u5011\u5c31\u53ef\u4ee5\u5f97\u5230\u82f1\u6587\u97f3\u7bc0\u8207\u4e2d\u6587\u5b57\u7684 (b) \u6bcd\u97f3\u90e8\u4efd \u82f1\u4e2d\u97f3\u8b6f\u5b57\u8655\u7406\u7684\u90e8\u5206\uff1a \u6ce8\u97f3\u7b26\u865f \u3127 \u3128 \u3129 \u311a \u311b \u311c \u311d \u311e \u311f \u3120 \u3121 \u3122 \u3123 \u3124 \u3125 \u3126 \u97f3\u8b6f\u5f8c\u7684\u8a5e\u5f59\u8207\u539f\u8a5e\u5f59\u4e4b\u9593\uff0c\u6700\u76f4\u63a5\u7684\u6bd4\u8f03\u65b9\u5f0f\uff0c\u5c31\u662f\u8acb\u6bcd\u8a9e\u4f7f\u7528\u8005\u767c\u97f3\uff0c \u5f9e\u4e8b\u5f62\u7d20\u76f8\u4f3c\u5ea6\u8861\u91cf\uff0c\u7cfb\u7d71\u6839\u64da\u524d\u4e00\u500b\u968e\u6bb5\u8a13\u7df4\u6240\u5f97\u5230\u7684\u5c0d\u7167\u8868\uff0c\u5c07\u82f1-\u4e2d\u97f3\u8b6f \u662f\u65b7\u5b57\u3002\u4f8b\u5982\u300c\u4e9e\u745f\u300d\u9019\u500b\u97f3\u8b6f\u5b57\uff0c\u7d93\u904e\u65b7\u5b57\u5f8c\u53d6\u51fa\u300c\u4e9e\u300d\u8207\u300c\u745f\u300d \u3002 \u7136\u5f8c\u4ee5\u7269\u7406\u4e0a\u53ef\u4ee5\u6e2c\u91cf\u5230\u7684\u97f3\u6ce2\u4f86\u6bd4\u8f03\u3002\u5982\u679c\u5f9e\u4eba\u985e\u53ef\u4ee5\u767c\u51fa\u7684\u8a9e\u97f3\u4f86\u770b\uff0c\u97f3\u7d20 \u97f3\u7bc0\u5c0d\u61c9\u5171\u4e09\u7d44\uff0c\u5305\u62ec\u300c\u311e\u2192e\u300d \u3001 \u300c\u3126\u2192l\u300d\u8207\u300c\u3109\u3128\u3123\u2192don\u300d \u3002\u7b2c\u4e8c\u500b\u968e\u6bb5\u5be6\u969b (1) \u65b7\u5b57\uff1a\u5728\u6536\u5230\u82f1\u4e2d\u97f3\u8b6f\u5b57\u6642\uff0c\u7b2c\u4e00\u500b\u6b65\u9a5f\u5373\u662f\u53d6\u51fa\u5176\u4e2d\u7684\u4e2d\u6587\u5b57\uff0c\u4e5f\u5c31 IPA \u03b9 \u03c5 \u03c8 \u03b1 \u03bf \u03a6 \u03b5 \u03b1\u03b9 \u03b5\u03b9 \u03b1\u03c5 \u03bf\u03c5 \u03b1\u03bd \u2194\u03bd \u03b1\u039d \u2194\u039d \uf6db</td></tr><tr><td>\u96c6\u5408\u662f\u56fa\u5b9a\u4e14\u6709\u9650\u7684\uff0c\u6211\u5011\u53ef\u4ee5\u5617\u8a66\u5728\u97f3\u7d20\u7684\u5c64\u6b21\u4f86\u6bd4\u8f03\u3002\u5169\u500b\u97f3\u7d20\u7684\u767c\u97f3\u4f4d \u5b57\u8f49\u63db\u6210\u82f1\u6587\u8a5e\u4e4b\u5f8c\uff0c\u518d\u8207\u5019\u9078\u540d\u55ae\u76f8\u6bd4\u8f03\u3002\u5982\u524d\u4f8b\uff0c \u300c\u611b\u723e\u9813\u300d\u5148\u8f49\u63db\u6210\u6ce8\u97f3 (2) \u6f22\u5b57\u8f49\u6ce8\u97f3\u7b26\u865f\uff1a\u5c07\u524d\u4e00\u500b\u6b65\u9a5f\u65b7\u51fa\u4f86\u7684\u6f22\u5b57\uff0c\u7d93\u67e5\u8868\u5f8c\u5f97\u5230\u76f8\u5c0d\u61c9\u6f22 \u5c0d\u5019\u9078\u540d\u55ae\u4e2d\u7684\u97f3\u8b6f\u5b57\uff0c\u8655\u7406\u7684\u65b9\u5f0f\u4e5f\u662f\u985e\u4f3c\u3002\u6bcf\u4e00\u500b\u82f1\u6587\u8a5e\u5f59\uff0c\u6211\u5011\u67e5\u8868</td></tr><tr><td>\u7f6e\uff0c\u6216\u662f\u767c\u97f3\u65b9\u5f0f\u8d8a\u76f8\u8fd1\uff0c\u5169\u500b\u8072\u97f3\u4e5f\u6703\u8d8a\u76f8\u4f3c\u3002\u7576\u6211\u5011\u4ee5\u66f8\u5beb\u6587\u5b57\u4f86\u6bd4\u8f03\u6642\uff0c \u7b26\u865f\u300c\u311e\u2022\u3126\u2022\u3109\u3128\u3123\u300d \uff0c\u7136\u5f8c\u67e5\u8868\u5f8c\u5f97\u5230\u300ce\u2022l\u2022don\u300d \u3002\u62ff\u6389\u97f3\u7bc0\u7b26\u865f\u5f8c\u5c31\u5f97 \u5b57\u7684\u6ce8\u97f3\u7b26\u865f\u3002\u4f8b\u5982\u300c\u4e9e\u300d\u67e5\u8868\u5f8c\uff0c\u5f97\u5230\u300c\u3127\u311a\u300d \uff0c\u5728\u6b64\u6211\u5011\u5ffd\u7565\u8072\u8abf\u7b26\u865f\u3002 (CMU dict 0.6)\uff0c\u4ee5\u5f97\u5230\u8a72\u8a5e\u5f59\u7684\u767c\u97f3\u3002\u4f8b\u5982\u300cArthur\u300d\u7684\u767c\u97f3\uff0c\u7d93\u67e5\u8868\u5f8c\u5f97\u5230\u300cAA</td></tr><tr><td>\u5c31\u662f\u76f4\u63a5\u6bd4\u8f03\u5f62\u7d20\u7684\u76f8\u4f3c\u5ea6\u3002\u5982\u679c\u66f8\u5beb\u6587\u5b57\u7cfb\u7d71\u4e0d\u540c\uff0c\u4f8b\u5982\u4e2d\u6587\u7684\u65b9\u584a\u5b57\uff0c\u8207\u82f1 \u5230\u300celdon\u300d \uff0c\u7136\u5f8c\u518d\u505a\u914d\u5076\u914d\u5c0d(mate matching)\u3002 (3) \u6ce8\u97f3\u7b26\u865f\u8f49 IPA\uff1a\u5c07\u524d\u4e00\u500b\u6b65\u9a5f\u4e2d\u7684\u6ce8\u97f3\u7b26\u865f\uff0c\u7d93\u67e5\u8868\u4e8c\u5f8c\uff0c\u5f97\u5230\u76f8\u5c0d R TH ER\u300d(\u5ffd\u7565\u91cd\u97f3\u6a19\u793a)\u3002</td></tr><tr><td>\u6587\u7684\u7f85\u99ac\u62fc\u97f3\u6587\u5b57\uff0c\u5c31\u5fc5\u9808\u5148\u8f49\u63db\u5230\u76f8\u540c\u7684\u5b57\u6bcd\u96c6\u5408\uff0c\u624d\u80fd\u9032\u884c\u6bd4\u5c0d\u3002 \u8868\u4e00\u5217\u51fa\u4e0a\u8ff0\u4f8b\u5b50\u7684\u8a13\u7df4\u7d50\u679c\u3002\u8ddf\u5176\u4ed6\u7f85\u99ac\u62fc\u97f3\u7cfb\u7d71\u4f86\u6bd4\u8f03\uff0c\u6211\u5011\u53ef\u4ee5\u767c \u61c9\u6ce8\u97f3\u7b26\u865f\u7684 IPA\u3002\u8868\u4e8c\u5217\u51fa\u6bcd\u97f3\u548c\u5b50\u97f3\u8207 IPA \u5c0d\u7167\uff0c\u9019\u90e8\u4efd\u53c3\u8003\u8b1d\u570b\u5e73(1998)\uff0c \u300c\u8a9e\u97f3\u6bd4\u8f03\u300d\u662f\u6574\u500b\u6d41\u7a0b\u6700\u91cd\u8981\u7684\u90e8\u5206\uff0c\u7576\u6211\u5011\u62ff\u5230\u5169\u4e32 IPA \u6642\uff0c\u5982\u4f55\u6bd4\u8f03</td></tr><tr><td>\u5728\u5f62\u7d20\u4e0a\u7684\u6bd4\u8f03\uff0cOdell \u8207 Russell \u7684 Soundex \u7cfb\u7d71(Knuth, 1973)\uff0c\u662f\u5c6c\u65bc\u540c \u73fe\uff1a\u7531\u9019\u500b\u7cfb\u7d71\u6240\u7522\u751f\u7684\u5c0d\u61c9\uff0c\u5728\u5f62\u7d20\u4e0a\u6bd4\u5176\u4ed6\u62fc\u97f3\u7cfb\u7d71\u4f86\u5f97\u66f4\u63a5\u8fd1\u5be6\u969b\u60c5\u5f62\u3002 \u4e26\u7565\u4f5c\u4fee\u6b63\u3002\u4f8b\u5982\u300c\u3127\u300d\u67e5\u8868\u5f8c\uff0c\u5f97\u5230\u300c\u03b9\u300d \u3002\u4e00\u822c IPA \u7684\u8868\u793a\u5fc5\u9808\u914d\u5408\u7279\u6b8a\u5b57\u9ad4\uff0c \u9019\u5169\u500b IPA \u5b57\u4e32\u7684\u76f8\u4f3c\u5ea6\u5462\uff1f\u9996\u5148\u4f86\u770b\u4ee5\u4e0b\u4e09\u500b\u5b9a\u7fa9\uff0c\u7531\u5b57\u6bcd\u5c0d\u9f4a\u76f8\u4f3c\u5ea6\u3001\u5230\u5b57</td></tr><tr><td>\u8a9e\u8a00\u7684\u7f85\u99ac\u62fc\u97f3\u5b57\u6bcd\uff0c\u5229\u7528\u5b50\u97f3\u4f86\u6355\u6349\u8a5e\u5f59\u767c\u97f3\u7684\u7279\u6027\u3002\u7576\u5169\u500b\u8a5e\u5f59\u7684\u5b50\u97f3\u4f4d\u7f6e \u50cf\u662f\u82f1-\u4e2d\u97f3\u8b6f\u5b57\u4e2d\u7684\u3126\uff0c\u4f8b\u5982\u8c9d\u723e(Bell)\u4e2d\u7684\u300c\u723e\u300d\u5b57\uff0c\u5982\u679c\u63a1\u7528\u5176\u4ed6\u62fc\u97f3\u7cfb</td></tr></table>",
+                "num": null,
+                "html": null,
+                "text": "CMU dict) (ftp://ftp.cs.cmu.edu/project/fgdata/dict/)\u63a1\u7528 ASCII \u4f86\u8868\u793a\uff0c\u9644\u9304\u5217\u51fa CMU dict",
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-1006.json

@@ -0,0 +1,1011 @@
+{
+    "paper_id": "O00-1006",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:04.639131Z"
+    },
+    "title": "Clustering Similar Query Sessions Toward Interactive Web Search",
+    "authors": [
+        {
+            "first": "Chien-Kang",
+            "middle": [],
+            "last": "Huang",
+            "suffix": "",
+            "affiliation": {},
+            "email": "ckhuang@mars.csie.ntu.edu.tw"
+        },
+        {
+            "first": "Lee-Feng",
+            "middle": [],
+            "last": "Chien",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Academic Sinica",
+                "location": {
+                    "country": "Taiwan"
+                }
+            },
+            "email": "*lfchien@iis.sinica.edu.tw"
+        },
+        {
+            "first": "Yen-Jen",
+            "middle": [],
+            "last": "Oyang",
+            "suffix": "",
+            "affiliation": {},
+            "email": "yjoyang@csie.ntu.edu.tw"
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "A new effective log-based approach for interactive Web search is presented in this paper. The most important feature of the proposed approach is that the suggested terms corresponding to the user's query are extracted from similar query sessions, rather than from the contents of the retrieved documents. The experiment results demonstrate that this approach has a great potential in developing more effective web search utilities and may inspire more studies on advanced log mining mechanisms.",
+    "pdf_parse": {
+        "paper_id": "O00-1006",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "A new effective log-based approach for interactive Web search is presented in this paper. The most important feature of the proposed approach is that the suggested terms corresponding to the user's query are extracted from similar query sessions, rather than from the contents of the retrieved documents. The experiment results demonstrate that this approach has a great potential in developing more effective web search utilities and may inspire more studies on advanced log mining mechanisms.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "Users' queries for Web search are usually short. For example, the average length of TREC topic description for conventional text retrieval is 15 tokens [11, 12] , while analyses of web search engine logs reveal that the average query length for Web search is about 2.3 tokens [6, 9] . Short queries means that the information about the user's intention provided to the search engine is very limited. To deal with the short query problem, interactive search techniques [2, 7] which attempt to identify the user's intentions and suggest more precise query terms are therefore commonly incorporated in Web search engine design.",
+                "cite_spans": [
+                    {
+                        "start": 152,
+                        "end": 156,
+                        "text": "[11,",
+                        "ref_id": "BIBREF7"
+                    },
+                    {
+                        "start": 157,
+                        "end": 160,
+                        "text": "12]",
+                        "ref_id": "BIBREF8"
+                    },
+                    {
+                        "start": 276,
+                        "end": 279,
+                        "text": "[6,",
+                        "ref_id": "BIBREF3"
+                    },
+                    {
+                        "start": 280,
+                        "end": 282,
+                        "text": "9]",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 468,
+                        "end": 471,
+                        "text": "[2,",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 472,
+                        "end": 474,
+                        "text": "7]",
+                        "ref_id": "BIBREF4"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "To determine more relevant query terms for each given query, the conventional interactive search processes often rely on the key terms in the retrieved documents [2, 7, 10] . The key term set is extracted either statically from the documents during preprocessing or dynamically on-the-fly. Since the precision rates of the retrieved documents are usually not high enough, the extracted key terms are often found not relevant and not very helpful in practical Web search services.",
+                "cite_spans": [
+                    {
+                        "start": 162,
+                        "end": 165,
+                        "text": "[2,",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 166,
+                        "end": 168,
+                        "text": "7,",
+                        "ref_id": "BIBREF4"
+                    },
+                    {
+                        "start": 169,
+                        "end": 172,
+                        "text": "10]",
+                        "ref_id": "BIBREF6"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In fact, extraction of relevant terms can be carried out by analyzing users' logs. In recent years, mining search engine logs has been obtaining more attention. Silverstein et al. [9] performed a second-order analysis on a log with a huge number of Web query terms. The results are then used to facilitate phrase recognition and query expansion [3] .",
+                "cite_spans": [
+                    {
+                        "start": 180,
+                        "end": 183,
+                        "text": "[9]",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 345,
+                        "end": 348,
+                        "text": "[3]",
+                        "ref_id": "BIBREF2"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we propose a new approach based on log analysis for developing more effective interactive Web search engines. The most important feature of the proposed approach is that the suggested terms are extracted from similar query sessions, rather than from the contents of the retrieved documents. A query session is defined as a sequence of search requests issued by a user for a certain information need. The basis of the proposed approach is that two users with the same information need will issue common or related query terms. For example, in search for a subject regarding \"search engine technology\", a user may submit query terms such as \"search engine\", \"Web search\", \"Google\", \"Web search and multimedia\", while another user may submit \"Web search\", \"Lycos\". Therefore, if similar query sessions could be identified, query terms for the same information need can be extracted and applied to improve the effectiveness of search engines.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "The remainder of the paper will be organized as follows. Section 2 is a brief introduction to the idea of interactive search based on similar query sessions. The method proposed for segmenting query sessions from proxy logs will be described in Section 3. Then, how query sessions are clustered is addressed in Section 4.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Section 5 will present some experiment results and a conclusion is given in Section 6. Fig.1 is an abstract diagram showing our idea for interactive search. Before introducing the basic idea of the proposed approach, the concept of query session is presented and defined below:",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 87,
+                        "end": 92,
+                        "text": "Fig.1",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "The proposed approach is assumed that the query space of users is formed by clusters of users' query sessions, and a set of query sessions grouped in the same clusters contain similar information needs. For each input query session with a sequence of i query terms, the interactive search process is then designed to retrieve the most similar cluster of query sessions from the query space, and then extract relevant terms in the cluster as suggested terms for next search. Once the i+1th query term is selected, it forms a new query session with i+1 terms and the interactive process will perform again.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query session = (ID, R 1 ,\u2026, R m ) where ID means the identifier of a user submitting a sequence of requests to a search engine in a certain period of time. Each request R i = (t i , q i ) means user ID sends a query term q to the search engine at time t",
+                "sec_num": null
+            },
+            {
+                "text": "Terms",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Suggested",
+                "sec_num": null
+            },
+            {
+                "text": "R i-1 , R i , R i+1",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Suggested",
+                "sec_num": null
+            },
+            {
+                "text": "Popular Visited Pages with the Subject Based on the above definition and idea, the problem to be dealt with is then formulated.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Input Query Session",
+                "sec_num": null
+            },
+            {
+                "text": "For a set of query sessions from a query session log, the considering problem is to cluster these query sessions into different groups based on estimated similarity between query sessions. Each cluster can be defined as {S i | f(S i , S j ) > threshold}, in which f() is the similarity estimation function between query sessions.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Query Session Clustering Problem",
+                "sec_num": null
+            },
+            {
+                "text": "The proposed approach, as shown in Fig. 2 , is composed of three processing modules: query session segmentation module, query session clustering module and relevant term extraction module. In the stage of query session segmentation, each query session will be segmented and extracted from a proxy log, according to the time gap between successive search requests. All of the extracted query sessions will form as a query session log. In the session clustering stage, the sessions with similar queries will be clustered and the cluster names extracted from composed high frequency terms. In the relevant term extraction stage, the relevance between the recorded query terms will be calculated and sets of relevant terms will be extracted for term suggestion applications in a search engine. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 35,
+                        "end": 41,
+                        "text": "Fig. 2",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Overview of the Proposed Approach",
+                "sec_num": null
+            },
+            {
+                "text": "A common proxy server might easily have thousands of clients accessing the web through it. Not only the general HTTP requests could pass through the proxy server, all of search HTTP requests are same. Compared with common search engine logs, a proxy server's log records more rigid information for users' information access and, more importantly, the recorded search requests are not limited to certain search engines. However, a proxy log might record too much information and only some of them are useful in terms of search engine applications [13] . In our application it is sufficient to only use the following fields of logging information:",
+                "cite_spans": [
+                    {
+                        "start": 546,
+                        "end": 550,
+                        "text": "[13]",
+                        "ref_id": "BIBREF9"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": "A timestamp that indicates when a search request was submitted.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": "A client address that indicates the IP address of the requesting instance.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": "A URL string that contains the request content.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": "Since the experiments are just performing, the testing log is from NTU proxy servers and is still small. Some statistics of the testing proxy log are listed in Table 1 that queries for a single information need come clustered in time, and then there is a gap before the user returns to the search engine.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 160,
+                        "end": 167,
+                        "text": "Table 1",
+                        "ref_id": "TABREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": "The method for query session segmentation is then proposed as follows:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": "The Method for Query Session Segmentation:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": "For a proxy log, it will segment the whole log L = {T i | where",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": "T i = (ID k , t i , q i )} into a set of query sessions {S i | S i = (ID k , R 1 ,\u2026, R m )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": ", where R i = (t i , q i ), and t it i-1 < threshold}, where t i is the timestamp when the query q i issued.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Segmentation",
+                "sec_num": "3."
+            },
+            {
+                "text": "To realize the performance of the above method, several experiments have been performed. Table 2 : Percentages of the extracted singleton and non-singleton query sessions, when the time threshold is set as 5 minutes.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 89,
+                        "end": 96,
+                        "text": "Table 2",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Analysis of Segmented Query Sessions",
+                "sec_num": null
+            },
+            {
+                "text": "As the definition of the session clustering problem in Section 2, the similarity estimation function is necessary and formulated below: ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Clustering",
+                "sec_num": "4."
+            },
+            {
+                "text": "Similarity",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Clustering",
+                "sec_num": "4."
+            },
+            {
+                "text": "( ) ( ) \u2211 < < < < = n j m i j i mn R R sim S S sim 1 , 1 2 1 2 1 / , ,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Clustering",
+                "sec_num": "4."
+            },
+            {
+                "text": "The similarity between two composed query terms will be further described below.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Clustering",
+                "sec_num": "4."
+            },
+            {
+                "text": "Development of an effective relevance estimation function is important. Since our research is just in the beginning, only two kinds of relevance estimation functions were developed and tested. In the first method, the relevance between two query terms is simply calculated by the co-occurrence frequency value of the query terms in the segmented query sessions. In the second method, the relevance is calculated by the cosine value of the query terms' feature vectors.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Query Session Clustering",
+                "sec_num": "4."
+            },
+            {
+                "text": "In the first method, we define the relevance estimation function below.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Method I for Similarity Estimation of Relevant Terms",
+                "sec_num": null
+            },
+            {
+                "text": "Before calculating the relevance between query terms, a set of query sessions has been segmented and extracted from the testing proxy log. After preprocessing the query session log, we calculate the co-occurrence frequency between each unique query term and its associated terms occurring together in the same query sessions.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "f(x, y i ) = co-occurrence(x, y i )",
+                "sec_num": null
+            },
+            {
+                "text": "We explain the calculation process with a simple example below. After segmenting the proxy log, it is assumed that we got five query sessions S1-5 and each contains several query terms from A to F, e.g., In this case, f(B, C) will be 3, because B and C occur together in three sessions, i.e., S1, S2 and S3. Although the above method looks straightforward, its obtained performance is really out of our expectation.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "f(x, y i ) = co-occurrence(x, y i )",
+                "sec_num": null
+            },
+            {
+                "text": "In the first method, the relevance of two query terms needs a strong support of their co-existence in a certain number of query sessions. Using a VSM-like technique it can release such a constraint. The second method is based on vector space model, and it can be formalized as below.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Method II for Similarity Estimation of Relevant Terms",
+                "sec_num": null
+            },
+            {
+                "text": "Assuming there are two terms T 1 and T 2 :",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "f(x, y i ) = cos(FV(x), FV(y I )), FV(x) means feature vector of term x, FV(T i ) = {S j :N ij | Sj and Ti are coexist in query sessions, Nij is the count of their co-occurrence}",
+                "sec_num": null
+            },
+            {
+                "text": "T 1 {S 1 :N 11 , S 2 :N 12 , S 4 :N 14 , S 5 :N 15 } T 2 {S 1 :N 21 , S 2 :N 22 , S 3 :N 23 , S 7 :N 27 }",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "f(x, y i ) = cos(FV(x), FV(y I )), FV(x) means feature vector of term x, FV(T i ) = {S j :N ij | Sj and Ti are coexist in query sessions, Nij is the count of their co-occurrence}",
+                "sec_num": null
+            },
+            {
+                "text": "The relevance value of T1 and T2 is the obtained cosine or say the inner product value of these two vectors.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "f(x, y i ) = cos(FV(x), FV(y I )), FV(x) means feature vector of term x, FV(T i ) = {S j :N ij | Sj and Ti are coexist in query sessions, Nij is the count of their co-occurrence}",
+                "sec_num": null
+            },
+            {
+                "text": "( ) ( ) ( ) ( ) ( ) \u2211 \u2211 \u2211 \u2022 \u2022 = \u2022 = k k j j i i i N N N N T FV T FV T T f 2 1 2 1 2 1 2 1 2 1 cos ,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "f(x, y i ) = cos(FV(x), FV(y I )), FV(x) means feature vector of term x, FV(T i ) = {S j :N ij | Sj and Ti are coexist in query sessions, Nij is the count of their co-occurrence}",
+                "sec_num": null
+            },
+            {
+                "text": "A issues to be dealt with in the clustering process, that is, what each cluster means and how to name these clusters. In order to find out the representative meaning of each cluster and avoid the difficulty in classifying short sessions, the clustering process is being developed as shown in Fig. 4 , which is designed as an incremental adaptive procedure.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 292,
+                        "end": 298,
+                        "text": "Fig. 4",
+                        "ref_id": "FIGREF5"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "The Clustering Process",
+                "sec_num": null
+            },
+            {
+                "text": "This procedure consists of 4 processing steps:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Clustering Process",
+                "sec_num": null
+            },
+            {
+                "text": "1. For each incoming query session, check whether there are certain common query terms between the session and existing clusters. If the common query terms exist, assign the session to these clusters.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Clustering Process",
+                "sec_num": null
+            },
+            {
+                "text": "clusters, calculate the similarity between the session and existing clusters. If the estimated similarity is higher than a predefined threshold, the session will be assigned to the cluster.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "If the incoming session doesn't have sufficient common query term with existing",
+                "sec_num": "2."
+            },
+            {
+                "text": "3. If the incoming session isn't assigned to any cluster, it will be sent to the delay queue for further processing. In this step, the incoming session will compare with other sessions in delay queue to check whether there are common query terms in the sessions that could be combined. 4. A standalone module will dynamically merge or split the clusters according to the new requirements or the new incoming sessions. When the similarity of two clusters are higher than another pre-defined threshold, merge will happen; when the cluster grows larger, split will happen. Merging and splitting are strategies for maintaining the similarity of query sessions in a cluster. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "If the incoming session doesn't have sufficient common query term with existing",
+                "sec_num": "2."
+            },
+            {
+                "text": "The above clustering process was just implemented. The sessions grouped by Step 1 are set that should contain at least two common query terms, and each obtained cluster is then named by the pair of common query terms with the highest frequency.",
+                "cite_spans": [
+                    {
+                        "start": 75,
+                        "end": 81,
+                        "text": "Step 1",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Performance of Query Session Clustering",
+                "sec_num": null
+            },
+            {
+                "text": "Currently, there are about 700 initial clusters have been obtained from the query session log shown in Section 2. Table 3 Table 3 . An example of the obtained session clusters.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 114,
+                        "end": 121,
+                        "text": "Table 3",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 122,
+                        "end": 129,
+                        "text": "Table 3",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Performance of Query Session Clustering",
+                "sec_num": null
+            },
+            {
+                "text": "It is worthy to note that clusters with similar names (that with shared query terms as the names of the clusters) usually contain similar information needs. Table 4 . An example which contains a number of obtained clusters with information needs related to \u5716\u7247(picture)",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 157,
+                        "end": 164,
+                        "text": "Table 4",
+                        "ref_id": "TABREF6"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Performance of Query Session Clustering",
+                "sec_num": null
+            },
+            {
+                "text": "In fact, the proposed approach is also useful in relevant term extraction. We evaluate the proposed estimation methods with a testing set of query terms that were randomly selected from the testing proxy log. For Method I, the relevant terms are whose co-occurrence frequency large than 1, and for Method II the relevant terms are whose cosine value large than 0.25. The obtained preliminary result is shown in Table 5 .",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 411,
+                        "end": 418,
+                        "text": "Table 5",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Performance of Relevant Term Extraction",
+                "sec_num": null
+            },
+            {
+                "text": "The first column \"rank\" in Table 5 is the order of the testing terms in the extracted term set, which is sorted by their occurrences. The real query terms are listed in the \"term\" column, and their English translations are listed in the next column. The data in the \"freq\" column represents the occurrence of each query term. The \"total\" column indicates the numbers of all different co-occurred terms, and the \"related\" column the numbers of relevant terms among co-occurred terms that were checked manually. The next nine columns are the obtained statistics of the proposed methods.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 27,
+                        "end": 34,
+                        "text": "Table 5",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Performance of Relevant Term Extraction",
+                "sec_num": null
+            },
+            {
+                "text": "Each method consists of three columns, the first is the number of extracted relevant terms, the second is the number of correct relevant terms, and the third is the obtained accuracy. Note that the third method is the result merged with the proposed two methods. -\u66ff\u4ee3\u5f79 20",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Performance of Relevant Term Extraction",
+                "sec_num": null
+            },
+            {
+                "text": "\u2022 \u570b\u9632\u90e8:5 \u793e\u6703\u5f79:3 \u5175\u5f79:3 \u5357\u6295\u7e23\u653f\u5e9c:2 \u5167\u653f\u90e8:2 \u570b\u9632\u5f79:2",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Performance of Relevant Term Extraction",
+                "sec_num": null
+            },
+            {
+                "text": "In this paper, a new approach based on log analysis is proposed for implementing",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "6."
+            },
+            {
+                "text": "interactive Web search. The most important feature of the proposed approach is that the suggested terms corresponding to a user query are extracted from similar query sessions, rather than from the contents of the retrieved documents. Furthermore, the estimation of term relevance is also based on co-occurrence analysis of the query terms in query sessions. The experiment results presented in this paper are based on analysis of the proxy server logs. The results obtained so far demonstrate that the proposed approach is quite promising in respect to improving the effectiveness of interactive web search engines.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "6."
+            },
+            {
+                "text": "The results presented in this paper is just a beginning of mining log data toward developing more effective web search engines. Since this approach already demonstrates quite promising results, further investigation on mining log data deserves more of our attention. Further study may result in more advanced mining mechanism that can give us more comprehensive information about term relevance and allow us to identify users' information need more effectively. For example, some sort of thesaurus information may be derived from mining log data.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "6."
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "The Paraphrase Search Assistant: Terminology Feedback for Iterative Information Seeking",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [
+                            "G"
+                        ],
+                        "last": "Anick",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Tipirneni",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of 22nd International Conference on Research and Development in Information Retrieval (SIGIR-99)",
+                "volume": "",
+                "issue": "",
+                "pages": "153--159",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "P.G. Anick and S. Tipirneni, \"The Paraphrase Search Assistant: Terminology Feedback for Iterative Information Seeking,\" in Proceedings of 22nd International Conference on Research and Development in Information Retrieval (SIGIR-99), pages 153-159, 1999.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Phrase recognition and expansion for short precision-biased queries based on a query log",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "De Lima",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "O"
+                        ],
+                        "last": "Pedersen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of 22nd International Conference on Research and Development in Information Retrieval (SIGIR-99)",
+                "volume": "",
+                "issue": "",
+                "pages": "145--152",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "E.F. de Lima and J.O. Pedersen, \"Phrase recognition and expansion for short precision-biased queries based on a query log,\" in Proceedings of 22nd International Conference on Research and Development in Information Retrieval (SIGIR-99), pages 145-152, 1999.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Real life information retrieval: A study of user queries on the web",
+                "authors": [
+                    {
+                        "first": "B",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Jansen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Spink",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Bateman",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Saracevic",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "SIGIR FORUM",
+                "volume": "32",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "B.J. Jansen, A. Spink, J. Bateman, and T. Saracevic, \"Real life information retrieval: A study of user queries on the web,\" SIGIR FORUM, 32(1), 1998.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Phrasier: a System for Interactive Document Retrieval Using Keyphrases",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Jones",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Staveley",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of 22nd International Conference on Research and Development in Information Retrieval (SIGIR-99)",
+                "volume": "",
+                "issue": "",
+                "pages": "160--167",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "S. Jones and M.S. Staveley, \"Phrasier: a System for Interactive Document Retrieval Using Keyphrases,\" in Proceedings of 22nd International Conference on Research and Development in Information Retrieval (SIGIR-99), pages 160-167, 1999.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Analysis of a very large AltaVista query log",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Silverstein",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Henzinger",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Marais",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Morics",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C. Silverstein, M. Henzinger, H. Marais, and M. Morics., \"Analysis of a very large AltaVista query log,\" Technical Report 1998-014, Digital Systems Research Center, 1998.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Fast and Effective Query Refinement",
+                "authors": [
+                    {
+                        "first": "B",
+                        "middle": [],
+                        "last": "Velez",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Weiss",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Sheldon",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Gifford",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of 20th International Conference on Research and Development in Information Retrieval (SIGIR-97)",
+                "volume": "",
+                "issue": "",
+                "pages": "6--15",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "B. Velez, R. Weiss, M.A. Sheldon and D.K. Gifford, \"Fast and Effective Query Refinement,\" in Proceedings of 20th International Conference on Research and Development in Information Retrieval (SIGIR-97), pages 6-15, 1997.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Overviw of the sixth text retrieval conference TREC-5",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Voorhees",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Harman",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the Fifth Text REtrieval Conference (TREC-5)",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "E. Voorhees and D.K. Harman, \"Overviw of the sixth text retrieval conference TREC-5,\" in Proceedings of the Fifth Text REtrieval Conference (TREC-5), 1997",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Overviw of the sixth text retrieval conference TREC-6",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Voorhees",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Harman",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "Proceedings of the Sixth Text REtrieval Conference",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "E. Voorhees and D.K. Harman, \"Overviw of the sixth text retrieval conference TREC-6,\" in Proceedings of the Sixth Text REtrieval Conference (TREC-6),",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "SQUID Frequently Asked Questions. Section 6. Squid Log Files",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Wessels",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "D. Wessels, SQUID Frequently Asked Questions. Section 6. Squid Log Files. http://www.squid-cache.org/Doc/FAQ/FAQ-6.html",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Yam",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Yam. http://www.yam.com.tw.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "text": "An abstract diagram showing our idea for interactive search.",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF1": {
+                "text": "An overview of the proposed approach",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF2": {
+                "text": "shows the relationship between the time thresholds and the numbers of segmented query sessions. The time thresholds determine the maximum time gap between two successive requests from the same client. The values of the time thresholds were tuned from 0 seconds to 360 seconds. In the research of Silverstein et al, 5 minutes as suggested is a proper threshold value. With the same threshold value, the number of segmented query sessions is shown in Table 2. The percentages of the segmented singleton and non-singleton query sessions are found similar to those reported by Silverstein et al.",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF3": {
+                "text": "The numbers of segmented query sessions (that with more than 1 queries), regarding to the change of increasing time thresholds.",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF4": {
+                "text": "S1: {A, B} S2: {C, D, B} S3: {A, B, C} S4: {A, E} S5: {B, C, E, F}",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF5": {
+                "text": "The work flow of the session clustering process.",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "TABREF1": {
+                "type_str": "table",
+                "content": "<table><tr><td>.</td></tr></table>",
+                "html": null,
+                "text": "Some statistics of the testing proxy log.",
+                "num": null
+            },
+            "TABREF3": {
+                "type_str": "table",
+                "content": "<table><tr><td>Given two sessions, S 1 = (ID K , R 11 ,\u2026,R 1m ) and S 1 = (ID L , R 21 ,\u2026,R 2n ), in which R ij</td></tr><tr><td>is the j-th query term occurred in session S i which is issued by a client. The</td></tr><tr><td>similarity estimation function is defined as:</td></tr></table>",
+                "html": null,
+                "text": "",
+                "num": null
+            },
+            "TABREF5": {
+                "type_str": "table",
+                "content": "<table><tr><td>Cluster Name</td><td colspan=\"2\">Sessions in Cluster</td><td/></tr><tr><td colspan=\"2\">\u53f0\u5927__\u53f0\u7063\u5927\u5b78 10: \u53f0\u5927</td><td>\u53f0\u7063\u5927\u5b78</td><td/></tr><tr><td/><td>1: +\u53f0\u5927</td><td colspan=\"2\">\u53f0\u5927 \u53f0\u7063\u5927\u5b78</td></tr><tr><td/><td>1: \u5716\u66f8\u9928</td><td colspan=\"3\">\u53f0\u5927\u5716\u66f8\u9928 \u53f0\u5927 \u53f0\u7063\u5927\u5b78</td><td>\u53f0\u7063\u5927\u5b78\u5716\u66f8\u9928</td></tr><tr><td/><td colspan=\"4\">1: \u53f0\u5927\u4e2d\u570b\u6587\u5b78\u7cfb \u53f0\u5927 \u53f0\u7063\u5927\u5b78</td></tr><tr><td/><td>1: \u53f0\u5927</td><td>\u53f0\u7063\u5927\u5b78</td><td>\u5175\u5b78</td></tr><tr><td/><td colspan=\"3\">1: \u81fa\u5927\u7269\u7406 \u53f0\u5927 \u53f0\u7063\u5927\u5b78</td></tr><tr><td/><td colspan=\"4\">1: \u53f0\u7063\u5927\u5b78 \u53f0\u5927 \u53f0\u7063\u5927\u5b78^\u62db\u6a19</td><td>\u53f0\u7063\u5927\u5b78^\u5de5\u7a0b</td><td>\u62db\u6a19^\u53f0\u5927</td></tr><tr><td/><td colspan=\"3\">1: \u53f0\u5927\u96fb\u6a5f \u53f0\u5927 \u53f0\u7063\u5927\u5b78</td></tr><tr><td/><td>1: \u53f0\u5927</td><td>\u53f0\u7063\u5927\u5b78</td><td colspan=\"2\">\u570b\u7acb\u53f0\u7063\u5927\u5b78</td></tr><tr><td/><td>1: \u793e\u6703\u5b78</td><td colspan=\"3\">\u53f0\u5927\u5716\u66f8\u9928 \u53f0\u5927 \u53f0\u7063\u5927\u5b78</td><td>\u5e02\u7acb\u5716\u66f8\u9928</td></tr><tr><td/><td>1: \u53f0\u5927</td><td colspan=\"3\">\u53f0\u5927\u91ab\u5b78\u9662 \u53f0\u5927\u91ab\u5b78\u7cfb \u53f0\u7063\u5927\u5b78</td></tr><tr><td/><td colspan=\"4\">1: \u6642\u5831\u80b2\u6a02 \u6642\u5831\u80b2\u6a02\u80a1\u4efd\u6709\u9650\u516c\u53f8</td><td>\u6642\u5831 \u53f0\u5927 \u53f0\u7063\u5927\u5b78</td><td>\u5927\u5b78\u806f\u62db\u653e\u699c</td><td>\u699c\u55ae</td></tr><tr><td/><td colspan=\"3\">1: \u53f0\u7063\u5927\u5b78 \u53f0\u5927 \u6210\u529f\u5927\u5b78</td></tr><tr><td/><td>1: \u5716\u66f8</td><td>\u53f0\u5927\u5716\u66f8</td><td colspan=\"2\">\u53f0\u5927 \u53f0\u7063\u5927\u5b78</td></tr><tr><td/><td>1: \u6930\u6797</td><td colspan=\"2\">\u53f0\u5927 \u53f0\u7063\u5927\u5b78</td><td>\u53f0\u7063\u5927\u5b78\u8a08\u7b97\u6a5f\u4e2d\u5fc3</td></tr><tr><td/><td>1: \u53f0\u5927</td><td>\u53f0\u7063\u5b78\u5927</td><td colspan=\"2\">\u53f0\u7063\u5927\u5b78</td><td>\u6bd4\u8cfd</td></tr></table>",
+                "html": null,
+                "text": "illustrates an example of the clusters. The numbers ahead in each row of Column 2 are frequency values of the corresponding sessions. Based on our initial observations, the relevance of the clustered query sessions is often high. It is obviously higher than that obtained with document-based approach in our experiences.",
+                "num": null
+            },
+            "TABREF6": {
+                "type_str": "table",
+                "content": "<table><tr><td>Cluster Name</td><td>Translation</td></tr><tr><td>\u5361\u901a\u5716\u7247__kitty</td><td>cartoon picture __ kitty</td></tr><tr><td>\u53ef\u611b\u5716\u7247__\u5361\u901a</td><td>lovable picture __ cartoon</td></tr><tr><td>\u6bcd\u89aa\u7bc0\u5716\u7247__\u6bcd\u89aa</td><td>mother's day picture __ mother</td></tr><tr><td>\u6bcd\u89aa\u7bc0\u5716\u7247__\u6bcd\u89aa\u7bc0</td><td>mother's day picture __ mother's day</td></tr><tr><td colspan=\"2\">\u6bcd\u89aa\u7bc0\u5716\u7247__\u6bcd\u89aa\u7bc0\u5361\u7247 mother's day picture __ mother's day greeting card</td></tr><tr><td>\u6bcd\u89aa\u7bc0\u5716\u7247__\u8db4\u8db4\u718a</td><td>mother's day picture __ bear</td></tr><tr><td>\u6bcd\u89aa\u7bc0\u5716\u7247__\u5eb7\u4e43\u99a8</td><td>mother's day picture __ carnation</td></tr><tr><td>\u5716\u7247__kitty</td><td>picture __ kitty</td></tr><tr><td>\u5716\u7247__\u5361\u901a</td><td>picture __ cartoon</td></tr><tr><td>\u5716\u7247__\u5361\u901a\u5716\u7247</td><td>picture __ cartoon picture</td></tr><tr><td>\u5716\u7247__\u5e03\u4e01\u72d7</td><td>picture __ pudding dog</td></tr><tr><td>\u5716\u7247__\u6bcd\u89aa\u7bc0</td><td>picture __ mother's day</td></tr><tr><td>\u5716\u7247__\u6bcd\u89aa\u7bc0\u5361\u7247</td><td>picture __ mother's day greeting card</td></tr><tr><td>\u5716\u7247__\u6bcd\u89aa\u7bc0\u5716\u7247</td><td>picture __ mother's day picture</td></tr><tr><td>\u5716\u7247__\u76ae\u5361\u4e18</td><td>picture __ picachu</td></tr><tr><td>\u5716\u7247__\u6709\u8da3</td><td>picture __ funny</td></tr><tr><td>\u5716\u7247__\u72d7</td><td>picture __ dog</td></tr><tr><td>\u5716\u7247__\u8db4\u8db4\u718a</td><td>picture __ bear</td></tr><tr><td>\u5716\u7247__\u684c\u9762</td><td>picture __ theme</td></tr><tr><td>\u5716\u7247__\u684c\u9762\u738b</td><td>picture __ themeking</td></tr><tr><td>\u5716\u7247__\u795e\u5947\u5bf6\u8c9d</td><td>picture __ pokemon</td></tr><tr><td>\u5716\u7247__\u52d5\u7269</td><td>picture __ animal</td></tr><tr><td>\u5716\u7247__\u52d5\u756b</td><td>picture __ animation</td></tr><tr><td>\u5716\u7247__\u5eb7\u4e43\u99a8</td><td>picture __ carnation</td></tr><tr><td>\u5716\u7247__\u904a\u6232</td><td>picture __ game</td></tr><tr><td>\u5716\u7247__\u904a\u6232\u4e0b\u8f09</td><td>picture __ game download</td></tr><tr><td>\u5716\u7247__\u5716</td><td>picture __ graph</td></tr><tr><td>\u5716\u7247__\u5716\u7247\u4e0b\u8f09</td><td>picture __ picture download</td></tr><tr><td>\u5716\u7247__\u5716\u5eab</td><td>picture __ picture bank</td></tr><tr><td>\u5716\u7247__\u5716\u6a94</td><td>picture __ picture file</td></tr><tr><td>\u5716\u7247__\u6f2b\u756b</td><td>picture __ comic</td></tr></table>",
+                "html": null,
+                "text": "is an example which contains a number of clusters with information needs related to \u5716\u7247 (picture). In these clusters, \u5716\u7247(picture) will relate to several different kinds of search subjects, including characters in cartoon (e.g. kitty and pokemon), downloading, online picture banks, greeting cards for some festivals, and etc. These similar clusters could be further taken as sub-clusters of the information needs. The obtained information would be very useful in performing term suggestion in interactive search process.",
+                "num": null
+            },
+            "TABREF8": {
+                "type_str": "table",
+                "content": "<table><tr><td>University includes \u53f0 \u5927 \u91ab \u5b78 \u9662 (medical school), \u53f0 \u7063 \u5927 \u5b78 \u8a08 \u7b97 \u6a5f \u4e2d \u5fc3</td></tr><tr><td>(computing center), \u53f0\u5927\u5716\u66f8\u9928 (library), \u53f0\u5927\u96fb\u6a5f (department of electrical</td></tr><tr><td>engineering),</td></tr></table>",
+                "html": null,
+                "text": "The performance obtained with the proposed methods.AnalyzingTable 5., we can find that Method I favors high frequency terms (e.g., term frequency > 50). It is really suited in applications that need not many but accurate relevant terms. However, for the query terms with not high frequency, we might rely on Method II. On the other hand, for those low frequency terms (term frequency < 10), Method II can not maintain a consistent performance. The effectiveness of this method is not reliable. In order to realize the effectiveness of the obtained result, we list an example of the extracted relevant query terms inTable 6.The query term is \u53f0\u7063\u5927\u5b78 (Taiwan University). The obtained relevant terms can be classified into 4 major categories. The first category is abbreviations including \u53f0 \u5927, +\u53f0\u5927 (\"+\" is the query syntax). The second is synonyms with different character forms like \u81fa\u7063\u5927\u5b78, with additional prefix like \u570b\u7acb\u53f0\u7063\u5927\u5b78, or nick name in semantics like \u6930\u6797 (Palm trees). The third is the sub divisions of Taiwan \u53f0\u5927\u4e2d\u570b\u6587\u5b78\u7cfb (department of Chinese literature), \u53f0\u5927\u91ab\u5b78\u7cfb (department of medicine). The final category is the events happened in Taiwan University, like \u62db\u6a19, \u5de5\u7a0b, \u5927\u5b78\u806f\u62db\u653e\u699c and \u699c\u55ae. An example of relevant terms extracted with the proposed methods.For more references, there are several examples that were not used in the testing are also illustrated below: \u5716\u7247:4 \u52d5\u756b:3 \u5716\u6a94:3 \u7db2\u9801\u88fd\u4f5c:2 \u5716:2 \u7db2\u9801\u5716\u5eab:2 \u904a\u6232\u4e0b\u8f09:2 \u4e16\u754c\u5730 \u5716:2 \u5730\u5716:2",
+                "num": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-1007.json

@@ -0,0 +1,403 @@
+{
+    "paper_id": "O00-1007",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:12.158722Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-1007",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "\u5176\u4e2d 1 pf \u8207 2 pf \u5206\u5225\u662f 1 d \u8207 2 d \u7684\u4e3b\u8981\u7279\u5fb5\uff0c 1 sf \u8207 2 sf \u5206\u5225\u662f 1 d \u8207 2 d \u7684\u6b21\u8981\u7279\u5fb5\uff0c PF Sim \u8207 SF Sim \u5206\u5225\u662f 1 d \u548c 2 d \u7684\u4e3b\u8981\u7279\u5fb5\u8207\u6b21\u8981\u7279\u5fb5\u7684\u76f8",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Sim IS IS ( , ) ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "Sim T T Sim T T T T Sim T T T T T Sim T T T T T Sim T T Sim T T T = = \uff0c \uff0c 2, 1 2 ),",
+                        "eq_num": "( , )"
+                    }
+                ],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "word Sim T T \u8868\u793a\u5169\u500b\u55ae\u7bc0\u9ede IS \u6790\u6a39\u9593\u7684\u76f8 \u5ea6\uff0c 1,i T \u548c 2, j T \u5206\u5225\u8868\u793a 1 T \u548c 2 T \u7684\u5b50 \u6a39\uff0c 1 T \u548c 2 T \u5206\u5225\u8868\u793a 1 T \u548c 2 T \u5b50\u6a39\u7684\u500b\u6578\uff0c subtree Sim \u8868\u793a\u5169\u500b\u975e\u55ae\u7bc0\u9ede IS \u6790\u6a39\u9593\u7684\u76f8 \u5ea6\uff0c \u5176\u5b9a\u7fa9\u5982\u4e0b\uff1a , , 1 1 2 ( , ( )) ( , ) max A T tree A k A k k subtree g A Sim T g T Sim T T T = = \u2211 (11) \u5176\u4e2d g \u662f\u4e00\u500b\u5f9e A T \u5230 B T \u7684\u4e00\u5c0d\u4e00\u51fd\u6578\uff0c , A k T \u8868\u793a A T \u7684\u4e00\u500b\u5b50\u6a39\uff0c A T \u8868\u793a A T \u5b50\u6a39\u7684\u500b\u6578\u3002 \u7531\u65bc g \u70ba\u4e00\u5c0d\u4e00\u51fd\u6578\uff0c\u6240\u4ee5 A B T T \uff0c\u56e0\u6b64\u9700\u8981\u7279\u5225 \u610f\uff1a\u82e5 1 2 T T \uff0c\u5247\u8a2d\u5b9a 1 A T T = \u4e14 2 B T T = \uff0c \u5426\u5247\u8a2d\u5b9a 2 A T T = \u4e14 1 B T T = \u3002 \u7576 1 T \u548c 2 T \u90fd\u662f\u5916\u90e8\u7bc0\u9ede\u7684\u6642\u5019\uff0c\u8868\u793a\u6b64\u4e8c\u8005\u7686\u70ba\u8a5e\uff0c\u5c0d\u65bc\u5169\u500b\u8a5e\u7684\u76f8 \u5ea6\uff0c\u5c31\u5229\u7528\u516c\u5f0f (1)\u6240\u63cf\u8ff0\u7684\u8a5e\u610f\u76f8 \u5ea6\u4f86\u91cf \u3002\u7576 1 T \u6216 2 T \u5176\u4e2d\u4e4b\u4e00\u70ba\u5916\u90e8\u7bc0\u9ede\u6642\uff0c\u8868\u793a\u5176\u4e2d\u4e00\u500b\u70ba\u8a5e\u53e6\u4e00\u500b \u5247\u70ba\u4e00\u500b\u7247\u8a9e\uff0c\u6b64\u6642\u5247\u905e\u8ff4\u5411\u4e0b \u51fa\u8a72\u7247\u8a9e\u4e2d\u8207\u8a72\u8a5e\u6700\u76f8 \u7684\u8a5e\u3002\u7576 1 T \u548c 2 T \u90fd\u4e0d\u70ba\u5916\u90e8\u7bc0\u9ede \u6642\uff0c\u5c31\u8868\u793a 1 T \u548c 2 T \u90fd\u542b\u6709\u5404\u81ea\u7684\u5b50\u6a39\u3002\u6b64\u6642\uff0c\u53ef\u4ee5\u5f9e\u4e09\u500b\u65b9\u5411\u4f86\u601d\u8003\uff1a\u6700\u57fa\u672c\u7684\u60f3\u6cd5\uff0c\u82e5\u5169 \u6a39\u7684\u6240\u6709\u5b50\u6a39\u90fd\u975e\u5e38\u76f8 \uff0c\u5247\u9019\u5169 \u6a39\u53ef\u80fd\u662f\u975e\u5e38\u76f8 \u7684\uff0c\u56e0\u6b64\u8003\u616e 1 2 ( , ) subtree Sim T T \u4f5c\u70ba 1 T \u548c 2 T \u7684\u76f8 \u5ea6\uff1b\u53e6\u5916\uff0c\u5982\u679c 1 T \u76f8 \u65bc 2 T \u7684\u4e00\u500b\u5b50\u6a39\uff0c\u6216\u662f 2 T \u76f8 \u65bc 1 T \u7684\u4e00\u500b\u5b50\u6a39\uff0c\u5247\u6839\u64da\u5206 \u7684\u591a\u5be1\u4f86\u6c7a\u5b9a\u8a72\u76f8 \u5ea6\u4e4b\u6b0a\u91cd\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "A Discourse Analysis of Questions in Mandarin Conversion",
+                "authors": [
+                    {
+                        "first": "Chung-Yin",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "16--81",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chang, Chung-Yin, \"A Discourse Analysis of Questions in Mandarin Conversion,\" M.A. Thesis, National Taiwan University Graduate Institute of Linguistics, June 1997, pp. 16-81.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Head-driven Statistical Models for Natural Language Parsing",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Collins",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Collins, M. J., \"Head-driven Statistical Models for Natural Language Parsing,\" Ph.D. Thesis, University of Pennsylvania, Philadelphia, 1999.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "An Efficient Context-free Parsing Algorithm",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Earley",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1970,
+                "venue": "Communications of the ACM",
+                "volume": "6",
+                "issue": "8",
+                "pages": "451--455",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Earley, J., \"An Efficient Context-free Parsing Algorithm,\" Communications of the ACM, vol. 6, no. 8, 1970, pp. 451-455.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Semantic Similarity Based on Corpus Statistics and Lexical Taxonomy",
+                "authors": [
+                    {
+                        "first": "Jay",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Jiang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "David",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Conrath",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the ROCLING X",
+                "volume": "",
+                "issue": "",
+                "pages": "19--33",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Jiang, Jay J. and David W. Conrath, \"Semantic Similarity Based on Corpus Statistics and Lexical Taxonomy,\" Proceedings of the ROCLING X, 1997, pp. 19-33.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Mandarin Chinese: A functional reference grammar",
+                "authors": [
+                    {
+                        "first": "Charles",
+                        "middle": [],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Sandra",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Thompson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1981,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Li, Charles and Sandra A. Thompson, \"Mandarin Chinese: A functional reference grammar,\" Berkeley and Los Angeles: University of California Press, 1981.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "An Information-Theoretic Definition of Similarity",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Lin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the International Conference on Machine Learning",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lin, D., \"An Information-Theoretic Definition of Similarity,\" Proceedings of the International Conference on Machine Learning, July 1998.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Foundations of Statistical Natural Language Processing",
+                "authors": [
+                    {
+                        "first": "Christopher",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Manning",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Hinrich",
+                        "middle": [],
+                        "last": "Sch Tze",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "296--303",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Manning, Christopher D. and Hinrich Sch tze, \"Foundations of Statistical Natural Language Processing,\" The MIT Press, 1999, pp. 296-303.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Structural Alignment During Similarity Comparisons",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [
+                            "B"
+                        ],
+                        "last": "Markman",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Gentner",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Cognitive Psychology",
+                "volume": "25",
+                "issue": "",
+                "pages": "431--467",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Markman, A. B. and D. Gentner, \"Structural Alignment During Similarity Comparisons,\" Cognitive Psychology, vol. 25, 1993, pp. 431-467.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "Using Information Content to Evaluate Semantic Similarity in a Taxonomy",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Resnik",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Proceedings of the 14 th International Joint Conference on Artificial Intelligence",
+                "volume": "1",
+                "issue": "",
+                "pages": "448--453",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Resnik, P., \"Using Information Content to Evaluate Semantic Similarity in a Taxonomy,\" Proceedings of the 14 th International Joint Conference on Artificial Intelligence, vol. 1, August 1995, pp. 448-453.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "A First Course in Probability",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Ross",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ross, S., \"A First Course in Probability,\" Macmillan, 1994.",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "An Efficient Probabilistic Context-Free Parsing Algorithm that Computes Prefix Probabilities",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Stolcke",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Computational Linguistics",
+                "volume": "21",
+                "issue": "2",
+                "pages": "165--202",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Stolcke, A., \"An Efficient Probabilistic Context-Free Parsing Algorithm that Computes Prefix Probabilities,\" Computational Linguistics, vol. 21, no. 2, 1995, pp. 165-202.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "\u73fe\u4ee3\u6f22\u8a9e\u4e2d\u7684\u6cd5\u76f8\u8a5e",
+                "authors": [],
+                "year": 1993,
+                "venue": "CKIP Technical Report",
+                "volume": "",
+                "issue": "93-06",
+                "pages": "1--16",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": ", \"\u73fe\u4ee3\u6f22\u8a9e\u4e2d\u7684\u6cd5\u76f8\u8a5e\", CKIP Technical Report, no.93-06, June 1993, pp. 1-16.",
+                "links": null
+            },
+            "BIBREF16": {
+                "ref_id": "b16",
+                "title": "\u77e5\u7db2",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "[17] , , \"\u77e5\u7db2\", http://how-net.com.",
+                "links": null
+            },
+            "BIBREF17": {
+                "ref_id": "b17",
+                "title": "The Hows of Why and the Whys of How",
+                "authors": [],
+                "year": 2000,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "1--27",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": ", \"The Hows of Why and the Whys of How\", \u8a9e\u8a00\u5b78\u7684 \u9020\u529b\u5b78\u8853\u7814\u8a0e\u6703, 2000, pp.1-27.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF0": {
+                "html": null,
+                "content": "<table><tr><td>\u7db2 \u7db2\u8def FAQ \u6aa2 \u4e2d\u610f\u5716\u8403\u53d6\u8207\u8a9e\u610f\u6bd4\u5c0d\u4e4b\u7814\u7a76 \u5f0f\u624d\u80fd\u5f97\u5230\u60f3\u8981\u7684\u7d50\u679c\u3002(2)\u7576\u4f7f\u7528\u8005\u60f3\u8981\u67e5\u8a62\u7684\u8cc7\u6599\u4e0d\u5b58\u5728\u95dc \u8a5e\uff0c\u6216\u8005\u4f7f\u7528\u8005\u7121\u6cd5 \u5230\u9069 \u6240\u63d0\u51fa\u7684\u65b9\u6cd5 \u662f\u5e0c\u671b\u80fd\u6709\u6548\u5730\u8403\u53d6\u51fa\u8a62\u554f\u53e5\u4e2d\u6240\u5305\u542b\u7684\u610f\u5716\uff0c\u4e26\u4e14\u85c9\u7531\u610f\u5716\u4f86 \u52a9\u6211\u5011\u5206 \u6240\u8868\u793a\u6210\u7684\u5411\u91cf\u505a\u6bd4\u5c0d\uff0c \u51fa\u8207 KS \u6700\u76f8\u95dc\u7684 \u3002 \u4fc2\u554f\u53e5\u53ca\u8868\u610f\u554f\u53e5\u56db\u5927\u985e\u3002\u9019\u4e9b\u529f\u80fd\u6210\u4e00\u7dda\u6027\u5206\u4f48\uff0c\u5f9e\u8aaa \u8005\u7684 \u5b9a\u5ea6\u4f86 \uff0c\u5206\u5225\u8868\u793a\u8aaa \u7528\u5728\u7db2\u8def\u4e0a\u7684\u554f\u53e5\u985e\u578b\u9032\u884c\u5206\u6790\uff0c\u7814\u7a76\u554f\u53e5\u5728\u5404\u7a2e\u53e5\u578b\u7d50\u69cb\u4e0b\u7684\u610f\u5716\u3002 \u8868 4 \u4e0d\u542b\u6cd5\u76f8\u526f\u8a5e\u4e4b\u53e5 \u8a9e\u52a9\u8a5e\u70ba \u7684\u554f\u53e5\u53ca\u5176\u5c0d\u61c9\u7684\u610f\u5716 \u6bb5(IS) \u7d93\u7531 AutoTag \u7684 \u52a9\uff0c\u53ef\u4ee5\u5c07\u4e00\u500b\u53e5\u5b50\u4f9d\u7167\u5206\u6790\u7684\u7d50\u679c\u8f49\u63db\u6210\u4e00\u500b \u6709\u8a5e\u6027\u7684\u8a5e\u5e8f\u5217\u3002 \u8868 8 How-net \u5b9a\u7fa9\u7bc4\u4f8b</td></tr><tr><td>\u80b2 \u3001 \u570b\u7acb\u6210\u529f\u5927\u5b78\u8cc7\u8a0a\u5de5\u7a0b\u7814\u7a76\u6240 \u3001 {laiys, leekl, chwu}@csie.ncku.edu.tw Fax: +886-6-2747076 \u6458\u8981 \u672c\u8ad6\u6587\u4e4b\u4e3b\u8981\u76ee\u7684\u662f\u5e0c\u671b\u80fd\u5229\u7528\u81ea\u7136\u8a9e\u8a00\u67e5\u8a62\u4f86\u505a\u70ba FAQ \u6aa2 \u7684\u65b9\u5f0f\u3002\u4e00\u500b\u5b8c\u6574\u7684 FAQ \u6a23\u672c\u5fc5\u5b9a\u542b\u6709\u4e00\u500b\u554f\u984c\u8207\u8a72\u554f\u984c\u7684 \u3002\u85c9\u7531\u6bd4\u8f03\u4f7f\u7528\u8005\u7684\u8a62\u554f\u53e5\u4ee5\u53ca FAQ \u6a23\u672c\u7684\u554f\u53e5\uff0c\u5982 \u7576\u7684\u95dc \u8a5e\uff0c\u5247 \u81f3\u7121\u6cd5 \u5230\u6240\u9700\u7684\u8cc7\u6599\u3002 \u76f8\u8f03\u65bc\u95dc \u8a5e\u67e5\u8a62\uff0c\u4f7f\u7528\u81ea\u7136\u8a9e\u8a00\u67e5\u8a62\u662f\u6700\u80fd \u6e05 \u8868 \u4f7f\u7528\u8005\u610f\u5716\u7684\u65b9\u5f0f\uff0c\u4e5f\u662f\u6700\u81ea \u7136\u7684\u65b9\u5f0f\u3002 \u8457\u7db2\u8def\u7684 \u767c\u5c55\u4ee5\u53ca\u81ea\u7136\u8a9e\u8a00\u8655\u7406\u6280\u8853\u7684\u63d0 \uff0c\u4ee5\u81ea\u7136\u8a9e\u8a00\u70ba\u4e3b\u7684\u8cc7\u8a0a\u6aa2 \u662f\u4e00\u500b\u6b63\u5728\u8208\u8d77\u7684\u7814\u7a76\u65b9\u5411\u3002\u76ee\u524d\u5df2\u6709\u5e7e\u500b\u7db2 \u63d0\u4f9b\u81ea\u7136\u8a9e\u8a00\u67e5\u8a62\u7684 \uff1a\u5728\u570b\u5916\u6709 Ask Jeeves \u7db2 [1]\u4ee5\u53ca FAQ Finder \u7cfb\u7d71[7]\uff0c\u570b \u6709 \u4f86\u8b49 \u7684 E \u535a\u58eb[5]\u3002\u4f46\u662f\u7531\u65bc\u76ee\u524d\u96fb \u6280 \u8853\u9084\u4e0d\u80fd\u505a\u5230\u5b8c\u5168\u7406\u89e3\u81ea\u7136\u8a9e\u8a00\u7684\u610f\u7fa9\uff0c\u4ee5\u81f4\u4f7f\u7528\u81ea\u7136\u8a9e\u8a00\u4f86\u505a\u8cc7\u8a0a\u6aa2 \u7684\u7814\u7a76\u5c1a\u672a\u6210 \uff0c \u4f46\u662f\u9019\u537b\u662f\u672a\u4f86\u8cc7\u8a0a\u6aa2 \u5fc5\u5b9a\u8981\u767c\u5c55\u7684\u65b9\u5411\u3002\u82e5\u80fd\u4f7f\u4e4b\u7d50\u5408\u524d \u7684\u8a9e\u97f3\u8fa8\u8b58\uff0c\u76f4\u63a5\u5229\u7528\u8a9e\u97f3 \u67e5\u8a62\uff0c\u5c07\u662f\u66f4\u52a0 \u5229\u4e14\u4eba\u6027\u5316\u7684\u4e00\u7a2e\u65b9\u5f0f\u3002 1-2. \u7814\u7a76\u52d5\u6a5f\u8207\u76ee\u7684 \u8fa8\u5169\u500b\u53e5\u5b50\u7684\u8a9e\u610f\u3002\u900f\u904e\u8a9e\u610f\u6587\u6cd5(semantic grammar)\u4ee5\u53ca \u7528\u8a5e(stopping words)\u7684\u7be9\u9078\uff0c\u6211 \u5011\u5c07\u554f\u53e5\u5206\u6210\u5169\u500b\u90e8\u5206\uff1a \u610f\u5716 \u6bb5(intention segment, IS) \u548c \u95dc \u8a5e \u6bb5(keyword segment, KS) \uff0c\u554f\u53e5\u53e5\u610f\u7684\u6bd4\u5c0d\u5c07\u5efa\u7acb\u5728\u9019\u5169\u90e8\u5206\u5404\u81ea\u7684\u8a9e\u610f\u6bd4\u5c0d\u4e0a\u3002\u6b64\u5916\uff0c\u5728\u95dc \u8a5e\u7684\u6bd4\u5c0d\u4e0a\uff0c\u6211 \u5011\u4f9d \u4fdd \u76ee\u524d\u88ab \u4f7f\u7528\u7684\u95dc \u8a5e\u67e5\u8a62\u70ba\u57fa\u790e\u7684\u8cc7\u8a0a\u6aa2 \u6280\u8853 \u5411\u91cf \u9593\u6a21\u578b(vector space model, VSM)\uff0c\u7528\u4f86\u6bd4\u8f03\u8a62\u554f\u53e5\u4e2d\u7684\u95dc \u8a5e\u8207 FAQ \u6a23\u672c\u7684 \u3002 2. \u7cfb\u7d71\u67b6\u69cb \u5982\u5716 1 \u6240\u793a\uff0c\u672c\u8ad6\u6587\u6240\u63d0\u51fa\u4e4b\u7cfb\u7d71\u67b6\u69cb\u4e3b\u8981\u5206\u70ba\u4e09\u5927\u90e8\u5206\uff1a \u8a9e\u610f\u5206\u6790\u5668 \u3001 \u554f\u53e5\u6bd4\u5c0d\u5668 \u53ca \u6587\u6bd4\u5c0d\u5668 \u3002\u4ee5\u4e0b\u91dd\u5c0d\u9019\u4e09\u500b\u90e8\u5206\u505a\u4e00\u500b\u7c21\u55ae\u7684\u4ecb\u7d39\u3002 \u9664\u4e86\u4e0a\u8ff0\u7684\u4e09\u5927\u6a5f\u5236\u5916\uff0cRanking Strategy \u5c07\u554f\u53e5\u6bd4\u5c0d\u5668\u53ca \u6587\u6bd4\u5c0d\u5668\u6240\u5f97\u5230\u7684\u7d50\u679c\uff0c\u5728 \u6b64\u505a\u4e00\u6574\u5408\uff0c\u6700\u5f8c\u5c07 \u540d\u5f8c\u7684\u7db2 \u8d85\u9023\u7d50\u8f38\u51fa\u3002 \u8005\u4e0d\u78ba\u5b9a\u6027\u9ad8\u7684\u5230\u4e0d\u78ba\u5b9a\u6027\u4f4e\u7684\uff1b\u5f9e\u8a0a \u7684\u89d2\u5ea6\u4f86 \uff0c\u5247\u8868\u793a\u8aaa \u8005\u5728 \u6c42\u8a0a \u7684\u5230\u50b3\u905e\u8a0a \u7684\u7591\u554f\u53e5\u3002\u540c\u6642\uff0c\u7591\u554f\u53e5\u4ea6\u986f\u73fe\u51fa\u5f9e \u6c42\u8f03 \u89c0\u3001\u6307\u793a\u6027\u7684\u8a0a \uff0c\u81f3\u50b3\u905e\u8f03\u4e3b\u89c0\u3001\u4ee5\u8aaa \u8005\u70ba\u51fa\u767c\u9ede\u7684 \u5ea6\u548c \u6cd5\u7684\u5206\u4f48\u3002\u56e0\u6b64\u9019\u8aaa\u660e\u5373\u4f7f\u5728\u53e5\u69cb\u5c64\u6b21\u610f\u7fa9\u7684\u4e3b\u89c0\u5316\u6216\u8aaa \u8005\u4ecb\u5165\u7a0b \u5ea6\u7684\u8868 \uff0c \u7a2e\u6a5f\u5236\u7684 \u4f5c\u4ea6\u660e\u986f\u53ef\u898b\u3002 \u5176\u7814\u7a76\u7d50\u679c\u986f\u793a\uff0c\u7591\u554f\u53e5\u7684\u8a9e\u6cd5\u5f62\u5f0f\u8207 \u901a\u529f\u80fd\u96d6\u662f\u591a\u5c0d\u591a\u7684\u95dc\u4fc2\uff0c\u5176\u4e2d\u537b\u4ecd\u5b58\u6709\u67d0\u7a2e \u7279\u5b9a\u7684\u5c0d\u61c9\u95dc\u4fc2\u3002\u8aaa \u8005 \u5411\u65bc\u4f7f\u7528 \u7591\u554f\u8a5e\u554f\u53e5 \u3001 \u662f\u975e\u554f\u53e5 \u53ca \u53e5\u5c3e\u8a9e\u52a9\u8a5e\u554f\u53e5\u70ba \u7684\u554f\u53e5 \u4f86 \u6c42\u81ea \u4e0d \u89e3 \u7684\u5916\u5728\u8a0a \u3002\u5728\u7db2 \u7db2\u8def\u4e0a\u7684\u554f\u984c\u4e5f\u591a\u4ee5\u9019\u4e09\u7a2e\u5f62\u5f0f\u5b58\u5728\uff0c \u56e0\u6b64\uff0c\u672c\u8ad6\u6587\u5373\u91dd\u5c0d\u6b64\u4e09\u7a2e\u985e\u578b\u7684\u554f\u53e5\u4f86\u505a\u5206\u6790\u3002 3-2. \u610f\u5716 \u6bb5(Intention Segment)\u7684\u5b9a\u7fa9 \u554f\u53e5 IS \u4e00\u822c\u5728\u505a\u95dc \u8a5e\u67e5\u8a62\u6642\uff0c\u591a \u7528\u7684\u662f\u540d\u8a5e\u6216\u52d5\u8a5e\uff0c\u6240\u4ee5\u65b7\u8a5e\u5f8c\uff0c\u6211\u5011\u5148\u5f9e\u53e5\u5b50\u4e2d \u51fa\u540d W_C G_C DEF 3-3-1. \u7591\u554f\u8a5e\u554f\u53e5 \u7591\u554f\u8a5e\u554f\u53e5\u76f8\u5c0d\u65bc\u82f1\u6587\u7684 WH \u554f\u53e5\u6709\u76f8\u7576\u63a5\u8fd1\u7684\u5730\u4f4d\uff0c\u7591\u554f\u8a5e\u901a\u5e38\u51fa\u73fe\u5728\u8207\u4e0d \u7591\u554f\u8a0a \u8a5e\u76f8\u540c\u6587\u6cd5\u529f\u80fd\u7684\u4f4d \u4e0a[3]\u3002\u4e2d\u6587\u5b58\u5728\u6709\u8a31\u591a\u7591\u554f\u8a5e\uff0c\u4f8b\u5982\uff1a \u3001 \u3001 \u6a23 \u3001 \u70ba \u3001 \u591a\u5c11 \u3001 \u88e1 \u3001 \u4f8b\u5982\u554f\u53e5\u4e2d\u5982\u679c\u554f\u5230 \u70ba \u8868 2 \u7591\u554f\u8a5e \u7684\u610f\u5716\u56e0\u8a9e\u6cd5\u4f4d \u7684\u4e0d\u540c\u800c\u6709\u6240\u4e0d\u540c \u5f8c\u52d5\u8a5e\u7247\u8a9e\u3002\u8868 5 \u5217\u8209\u51fa\u90e8\u5206\u662f\u975e\u554f\u53e5\u53ca\u5176\u5c0d\u61c9\u7684 IS\u3002 \u8868 7 \u554f\u53e5\u53ca\u5176\u76f8\u5c0d\u61c9\u95dc \u8a5e \u6bb5(KS)\u4e4b\u7bc4\u4f8b \u505a\u70ba\u8a62\u554f\u505a\u67d0\u4ef6\u4e8b\u7684\u65b9\u6cd5[16]\u3002 \u4e0d\u540c\u53e5\u578b\u7684\u554f\u53e5\u4e5f\u6703\u5177\u6709\u76f8\u540c\u7684 IS\u3002\u5c0d\u662f\u975e\u554f\u53e5\u800c\u8a00\uff0c\u6211\u5011\u8a8d\u70ba\u610f\u5716\u70ba\u63a5\u5728 A-not-A \u8a5e\u7d44\u4e4b \u8868 7 \u5217\u8209\u51fa\u90e8\u5206\u554f\u53e5\u53ca\u5176\u5c0d\u61c9\u7684 KS\u3002 \u9019\u500b\u7591\u554f\u8a5e\uff0c\u82e5\u51fa\u73fe\u5728\u526f\u8a5e\u4e4b\u524d\u53ef\u505a\u70ba\u8a62\u554f\u67d0\u4ef6\u4e8b\u60c5\u6216\u73fe\u8c61\u7684\u539f\u56e0\uff0c\u4f46\u82e5\u51fa\u73fe\u5728\u52d5\u8a5e\u4e4b\u524d\u537b \u52a9\u52d5\u8a5e\u958b\u982d\u7684\u554f\u53e5\uff0c\u9019\u5169\u985e\u554f\u53e5\u5728\u7d50\u69cb\u4e0a\u662f\u53ef\u4ee5\u4e92\u63db\u7684\u3002\u540c\u6a23\u5730\uff0c\u8868\u73fe\u5728 IS \u4e0a\u9762\uff0c\u76f8\u540c\u8a9e\u610f \u8a5e (stopping word dictionary)\uff0c\u7576\u4e00\u500b\u8a5e\u51fa\u73fe\u5728 \u7528\u8a5e\u8a5e \u4e2d\uff0c \u5c07\u4e4b\u5f9e\u95dc \u8a5e\u7d44\u88e1\u53bb\u9664\u3002 \u6709\u4e9b\u7591\u554f\u8a5e\u6703 \u8457\u5728\u53e5\u5b50\u4e2d\u7684\u76f8\u5c0d\u8a9e\u6cd5\u4f4d \u4e0d\u540c\uff0c\u5176\u610f\u7fa9\u4e5f\u4e0d\u76e1\u76f8\u540c\u3002\u5982\u8868 2 \u6240\u793a\uff0c \u4e0d\u53ef\u4ee5 \u3001 \u662f\u5426 \u3002\u662f\u975e\u554f\u53e5\u548c\u53e5 \u8a9e\u52a9\u8a5e\u70ba \u7684\u554f\u53e5\uff0c\u76f8\u5c0d\u65bc\u82f1\u6587 \u662f\u7531 be \u52d5\u8a5e\u6216\u662f \u8996\u70ba\u975e\u95dc \u8a5e\u3002\u7d93\u7531\u7d71\u8a08\u8a9e\u6599\u5eab\u53ef\u5f97\u5230\u4e00\u4e9b\u8a5e\u983b\uff0c\u5c07\u9ad8\u983b\u7684\u8a5e\u7d93\u904e\u4eba\u5de5\u7be9\u9078\u5efa\u7acb\u4e00\u500b \u7528\u8a5e \uff0c\u5e7e\u4e4e\u53ef\u4ee5\u60f3\u898b\u7684\u8a72\u53e5\u5c31\u662f\u5728\u554f\u67d0\u4ef6\u4e8b\u60c5\u6216\u73fe\u8c61\u7684\u539f\u56e0\uff1b\u4f46\u662f\uff0c \u662f\u975e\u554f\u53e5\u662f\u6307\u5305\u542b\u5177\u6709 A-not-AB \u6216\u662f A-not-A \u7279\u6027\u4e4b\u8a5e\u7d44\u7684\u554f\u53e5\uff0c\u4f8b\u5982\uff1a \u662f\u4e0d\u662f \u3001 \u53ef \u53e6\u5916\uff0c\u6709\u4e9b\u8a5e\u96d6\u7136\u7b26\u5408\u4ee5\u4e0a\u898f\u5247\uff0c\u4f46\u662f\u51fa\u73fe\u983b\u7387\u537b\u76f8\u7576\u9ad8\uff1b\u76f8\u5c0d\u800c\u8a00\uff0c\u5176\u91cd\u8981\u6027 \u4f4e\uff0c \u3001 \u70ba\u4f55 \u3002\u901a\u5e38\u7591\u554f\u8a5e\u53ef\u4ee5 \u52a9\u5224\u65b7\u554f\u53e5\u7684\u610f\u5716\uff0c 3-3-3. \u662f\u975e\u554f\u53e5 \u6211\u5011 \u9019\u4e9b\u8a5e\u985e\u7684\u8a5e\u8996\u70ba\u975e\u95dc \u8a5e\u3002 \u3001 \u6027 C \u578b \u53ef \u53ef \u5b50 \u5207\u7247\u7684\u7d50\u679c\u6b63\u78ba \u5bdf N human|\u4eba, police| \u8a5e\u53ca\u52d5\u8a5e\u7684\u90e8\u5206\u3002\u4f46\u662f AutoTag \u6240\u6a19\u8a18\u7684\u8a5e\u6027\u5206\u985e\u76f8\u7576 \uff0c\u5373\u4f7f\u662f\u540d\u8a5e\u985e\u4ecd\u6709\u8a31\u591a \u5206\uff0c\u800c \u4eba N human|\u4eba, *SufferFrom| , $cure| , #medical| , undesired| \u6b63\u78ba \u90e8\u5206\u985e\u5225\u96d6\u5c6c\u65bc\u540d\u8a5e\u537b\u4e0d\u505a\u95dc \u8a5e\u7528\uff0c\u5982\u5b9a\u8a5e(Ne)\u3001\u91cf\u8a5e(Nf)\u3001\u65b9\u4f4d\u8a5e(Ng)\u4ee5\u53ca\u4ee3\u540d\u8a5e(Nh)\uff0c N FlowerGrass|</td></tr><tr><td>\u679c\u5169\u8005\u7684\u8a9e\u610f\u76f8\u7576\u63a5\u8fd1\uff0c\u5247\u8a72 FAQ \u6a23\u672c\u7684 FAQ \u6a23\u672c\u7684 \u5728\u4ee5\u81ea\u7136\u8a9e\u8a00\u67e5\u8a62\u70ba\u4e3b\u7684\u8cc7\u8a0a\u6aa2 \u61c9\u7528\u4e2d\uff0cFAQ (Frequently Asked Questions)\u6aa2 \u662f\u4e00\u500b \u4e5f\u5c31\u53ef\u80fd\u5305\u542b\u4f7f\u7528\u8005\u60f3\u8981\u7684\u8cc7\u8a0a\u3002\u6b64\u5916\uff0c\u4e00\u500b 2-1. \u8a9e\u610f\u5206\u6790\u5668 \u5c0d\u4e00\u500b\u81ea\u7136\u8a9e\u8a00\u554f\u53e5\u800c\u8a00\uff0c\u6211\u5011\u8a8d\u70ba\u9664\u4e86\u95dc \u8a5e\u4e4b\u5916\uff0c\u4ecd\u6709\u5176\u4ed6\u56e0\u7d20\u53ef\u7528\u4f86\u5206\u8fa8\u554f\u53e5\u9593 \u554f\u53e5 \u610f\u5716 \u554f\u53e5 KS \u8868 5 \u90e8\u5206\u662f\u975e\u554f\u53e5\u53ca\u5176\u5c0d\u61c9\u7684\u610f\u5716 \u6bb5(IS) \u4e2d \u5982\u4f55 \u4e2d (Na)\u3001 (VC)\u3001 (Na) \u4e5f\u53ef\u80fd\u5305\u542b\u5176\u4ed6 \u5916\u7684\u8cc7\u8a0a\u3002\u56e0\u6b64\uff0c\u9664\u4e86\u5169\u500b\u7591\u554f\u53e5\u7684\u6bd4\u5c0d\u4e4b\u5916\uff0c\u4f7f\u7528\u8005\u6240 \u9700\u7684\u8cc7\u8a0a\u4e5f\u53ef\u4ee5\u900f\u904e\u6bd4\u5c0d\u8a62\u554f\u53e5\u8207 FAQ \u6a23\u672c\u7684 \u800c\u5f97\u5230\u3002 \u900f\u904e\u8a9e\u610f\u6587\u6cd5\u4ee5\u53ca \u7528\u8a5e\u7684\u7be9\u9078\uff0c\u6211\u5011\u5c07\u554f\u53e5\u5206\u6210\u5169\u500b\u90e8\u5206\uff1a \u610f\u5716 \u6bb5 \u548c \u95dc \u8a5e \u6bb5 \u3002\u610f\u5716 \u6bb5\u50b3 \u4f7f\u7528\u8005\u4e3b\u8981\u7684\u610f\u5716\uff0c\u95dc \u8a5e \u6bb5\u5305\u542b\u554f\u53e5\u4e2d\u6240\u6709\u7684\u95dc \u8a5e\uff0c\u554f\u53e5\u53e5\u610f\u7684\u6bd4 \u5c0d\u5c07\u5efa\u7acb\u5728\u9019\u5169\u90e8\u5206\u5404\u81ea\u7684\u8a9e\u610f\u6bd4\u5c0d\u4e0a\u3002\u6b64\u5916\uff0c\u6211\u5011 \u7528\u5411\u91cf \u9593\u6a21\u578b\u4f86\u6bd4\u8f03\u8a62\u554f\u53e5\u4e2d\u7684\u95dc \u8a5e\u8207 FAQ \u6a23\u672c\u7684 \u3002 \u7d93\u5be6\u9a57\u9a57\u8b49\uff0c\u672c\u8ad6\u6587\u6240\u63d0\u51fa\u7684\u65b9\u6cd5\u78ba\u5be6\u6bd4\u55ae\u7d14\u4f7f\u7528\u95dc \u8a5e\u67e5\u8a62\u4f86\u5f97\u6e96\u78ba\uff0c\u4f7f\u5e73\u5747\u6b63\u78ba \u7684 1. \u7dd2\u8ad6 1-1. \u8aaa\u660e \u76ee\u524d\u8cc7\u8a0a\u6aa2 (information retrieval)\u7684\u6280\u8853\u5df2\u7d93 \u4f7f\u7528\u5728\u6211\u5011\u65e5\u5e38\u751f \u4e2d\u3002\u8209 \u4e0a\u5716\u66f8 \u9928\u501f\u66f8\u3001\u7db2\u8def\u641c \u8cc7\u6599\uff0c\u6211\u5011\u5e38\u6703\u9700\u8981\u4e00\u4e9b\u8cc7\u8a0a\u6aa2 \u7684\u5de5\u5177 \u52a9\u6211\u5011 \u51fa\u60f3\u8981\u7684\u8cc7\u6599\u3002\u4ee5\u76ee \u524d\u7684\u6280\u8853\uff0c\u8cc7\u8a0a\u6aa2 \u7684\u61c9\u7528\u5927\u591a\u53ea\u63d0\u4f9b\u7531\u95dc \u8a5e\u9032\u884c\u67e5\u8a62\uff0c\u85c9\u7531\u95dc \u8a5e\u7684\u6bd4\u5c0d\u4ee5 \u51fa\u76f8\u95dc\u7684 \u6587 \u6216\u8cc7\u6599\u3002\u4f46\u662f\uff0c\u53ea\u5229\u7528\u95dc \u8a5e\u67e5\u8a62\u6709\u5169\u500b\u7f3a\u9ede\uff1a(1)\u95dc \u8a5e\u4e0d\u80fd\u6e05 \u4e14\u5b8c\u6574\u5730\u8868 \u4f7f\u7528\u8005 \u7684\u610f\u5716\uff0c\u4ee5\u81f4\u76f8\u95dc\u7684\u641c \u7d50\u679c\u904e\u591a\uff0c\u4f7f\u7528\u8005\u5f80\u5f80\u9700\u8981\u7d93\u904e\u597d\u5e7e\u6b21\u7684\u4f86\u56de\u4fee\u6539\u95dc \u8a5e\u6216\u67e5\u8a62\u65b9 \u4e0d\u932f\u7684\u65b9\u5411\u3002\u8a31\u591a\u7db2 \u901a\u5e38\u6703\u91dd\u5c0d\u8a72 \u57df\u4e2d\u5e38\u88ab\u554f\u5230\u7684\u554f\u984c\uff0c\u7d93\u7531\u4eba\u5de5\u6574\u7406\u9019\u4e9b\u554f\u984c\u53ca \uff0c \u63d0\u4f9b \u9032\u5165\u8a72\u7db2 \u7684\u4f7f\u7528\u8005\u76f4\u63a5 \u89bd\uff0c\u4ee5\u7bc0 \u8a62\u554f\u8207\u56de \u91cd\u8907\u6216\u76f8\u95dc\u6027\u554f\u984c\u7684\u6642\u9593\u3002\u4f46\u662f \u8457\u91cf\u7684\u589e\u52a0\uff0c\u4f7f\u7528\u8005\u4e5f \u4f86 \u96e3\u85c9\u7531\u76f4\u63a5 \u89bd \u5230\u6240\u9700\u7684 \uff0c\u56e0\u6b64\uff0c\u73fe \u8a31\u591a\u7db2 \u4e5f\u63d0\u4f9b FAQ \u6aa2 \u7684 \uff0c \u4f7f\u7528\u8005\u641c \u6240\u9700\u7684\u8cc7\u8a0a\u3002\u672c\u8ad6\u6587\u4e4b\u4e3b\u8981\u76ee\u7684 \u662f\u5e0c\u671b\u80fd\u5229\u7528\u81ea\u7136\u8a9e\u8a00\u67e5 \u8a62\u4f86\u505a\u70ba FAQ \u6aa2 \u7684\u65b9\u5f0f\u3002 1-3. \u7814\u7a76\u65b9\u6cd5\u7c21\u4ecb \u4e00\u500b\u5b8c\u6574\u7684 FAQ \u6a23\u672c\u5fc5\u5b9a\u542b\u6709\u4e00\u500b\u554f\u984c\u8207\u8a72\u554f\u984c\u7684 \u3002\u85c9\u7531\u6bd4\u8f03\u4f7f\u7528\u8005\u7684\u8a62\u554f\u53e5\u4ee5\u53ca FAQ \u6a23\u672c\u7684\u554f\u53e5\uff0c\u5982\u679c\u5169\u8005\u7684\u8a9e\u610f\u76f8\u7576\u63a5\u8fd1\uff0c\u5247\u8a72 FAQ \u6a23\u672c\u7684 \u4e5f\u5c31\u53ef\u80fd\u5305\u542b\u4f7f\u7528\u8005\u60f3\u8981 \u7684\u8cc7\u8a0a\u3002\u6b64\u5916\uff0c\u4e00\u500b FAQ \u6a23\u672c\u7684 \u4e5f\u53ef\u80fd\u5305\u542b\u5176\u4ed6 \u5916\u7684\u8cc7\u8a0a\u3002\u56e0\u6b64\uff0c\u9664\u4e86\u5169\u500b\u7591\u554f\u53e5\u7684 \u6bd4\u5c0d\u4e4b\u5916\uff0c\u4f7f\u7528\u8005\u6240\u9700\u7684\u8cc7\u8a0a\u4e5f\u53ef\u4ee5\u900f\u904e\u6bd4\u5c0d\u8a62\u554f\u53e5\u8207 FAQ \u6a23\u672c\u7684 \u800c\u5f97\u5230\u3002 FAQ Finder \u7cfb\u7d71\u5229\u7528 Word-Net \u4f86\u8861\u91cf\u82f1\u6587\u8a5e\u8207\u8a5e\u7684\u8a9e\u610f\u76f8 \u5ea6\uff0c\u70ba\u6574\u500b\u7cfb\u7d71\u767c\u5c55\u8a9e\u610f\u76f8 \u5ea6\u7684\u57fa\u790e\u3002\u4f46\u662f\u5728\u554f\u53e5\u7684\u76f8 \u5ea6\u90e8\u5206\uff0c\u5247\u662f\u55ae\u7d14\u5730\u6bd4\u5c0d\u5169\u500b\u554f\u53e5\u4e2d\u6240\u5305\u542b\u7684\u8a5e\u7d44\uff0c\u6211\u5011\u8a8d \u70ba \u662f\u6bd4\u8f03\u8a5e\u7d44\u4e26\u4e0d \u4ee5\u4ee3\u8868\u6574\u500b\u53e5\u610f\u3001\u6709 \uff0c\u800c\u4e14\u4e5f\u6709\u660e\u986f\u7684\u7f3a \u3002\u4f8b\u5982\uff1a \u6703 \u4e0d\u6703\u5c0e\u81f4 \u5316 \u3001 \u5316\u6703\u4e0d\u6703\u5c0e\u81f4 \uff0c\u6b64\u4e8c\u53e5\u6709\u5b8c\u5168\u76f8\u540c\u7684\u8a5e\u7d44\uff0c\u4f46\u662f\u5728\u610f\u7fa9\u4e0a \u537b\u662f\u5b8c\u5168\u4e0d\u540c\u3002 \u6bcf\u500b\u554f\u53e5\u90fd\u6709\u5176\u610f\u5716(intention)\uff0c\u8a72\u610f\u5716 \u4e00\u800c\u4e14\u5728\u53e5\u5b50\u88e1\u626e\u6f14\u76f8\u7576\u91cd\u8981\u7684\u89d2\u8272\u3002\u672c\u7814\u7a76 \u900f\u904e\u8a9e\u610f\u5206\u6790\u5668\uff0c\u6211\u5011\u53ef\u4ee5\u5f9e\u554f\u53e5\u4e2d\u8403\u53d6\u51fa IS \u53ca KS\uff0c\u505a\u70ba\u5f8c\u7e8c\u554f\u53e5\u6bd4\u5c0d\u4ee5\u53ca \u6587\u6bd4\u5c0d \u4e4b\u7528\u3002\u8a9e\u610f\u5206\u6790\u5668\u7531\u4e0b\u9762\u5e7e\u500b\u5b50\u90e8\u5206\u6240\u7d44\u6210\uff1a(1) AutoTag\uff0c\u4e2d\u7814\u9662 CKIP \u5c0f\u7d44\u767c\u5c55\u7684\u8a5e\u6027\u6a19\u8a18 \u7cfb\u7d71\uff0c\u505a\u70ba\u672c\u7cfb\u7d71\u7684\u524d\u8655\u7406\u5668\uff0c\u5c07\u4e00\u500b\u53e5\u5b50\u65b7\u8a5e\u4e26\u6a19\u793a\u8a5e\u6027\u3002(2) \u95dc \u8a5e\u8403\u53d6\uff0c\u7531\u8a5e\u6027\u7684\u5224 \u65b7\u4ee5\u53ca \u7528\u8a5e\u7684\u7be9\u9078\uff0c\u5f9e\u65b7\u8a5e\u5f8c\u7684\u53e5\u5b50\u4e2d \u51fa\u5176 KS\u3002(3) \u610f\u5716\u8403\u53d6\uff0c\u7d93\u7531\u6574\u7406\u6b78 \u7684\u8a9e\u610f\u6587 \u6cd5\uff0c\u5f9e\u554f\u53e5\u4e2d \u51fa\u5176 IS\u3002 2-2. \u554f\u53e5\u6bd4\u5c0d\u5668 \u7684\u95dc \u8a5e\u76f8 \u5ea6\uff0c\u914d\u5408\u4e00\u5c0d\u4e00\u51fd\u6578\u7684\u6700\u4f73\u5316\uff0c\u6c42\u53d6\u5169\u8005\u7684\u6700\u5927\u76f8 \u5ea6\u3002 2-3. \u5716 1 \u7cfb\u7d71\u67b6\u69cb\u5716 3. \u554f\u53e5\u7684\u8a9e\u610f\u5206\u6790\u8207\u8655\u7406 \u5728\u5927\u90e8\u5206\u7684\u60c5\u6cc1\u4e0b\uff0c\u95dc \u8a5e\u6709\u52a9\u65bc\u6aa2 \u51fa\u6211\u5011\u60f3\u8981\u7684 \uff0c\u4f46\u662f\u5728\u7b26\u5408\u95dc \u8a5e\u6bd4\u5c0d\u7684\u7d50 \u679c\u4e2d\uff0c\u5f80\u5f80\u542b\u6709\u5927\u91cf\u4e0d\u662f\u539f\u4f86\u6240 \u671b\u7372\u5f97\u7684 \uff0c\u800c\u5176\u4e3b\u8981\u539f\u56e0\u5728\u65bc\u95dc \u8a5e\u6c92\u6709 \u6cd5\u6b63\u78ba\u5730 \u50b3 \u4f7f\u7528\u8005\u7684\u610f\u5716\u3002\u56e0\u6b64\uff0c\u6211\u5011\u5e0c\u671b\u900f\u904e\u5c0d\u65bc\u554f\u53e5\u7684\u8a9e\u610f\u5206\u6790\uff0c\u80fd\u7522\u751f\u51fa\u554f\u53e5\u7684\u8a9e\u610f\u6587\u6cd5\uff0c \u9032\u800c\u8403\u53d6\u51fa\u5305\u542b\u5728\u554f\u53e5\u4e2d\u7684\u4f7f\u7528\u8005\u610f\u5716\u3002 3-1. \u7591\u554f\u53e5\u5206\u985e \u7684\u5dee\u7570\u3002\u89c0\u5bdf\u4e0b\u9762\u4e09\u500b\u554f\u53e5\uff1a \u3001 \u70ba \u8981 \u3001 \u8981 \u7684\u65b9\u6cd5 \u554f\u53e5 IS \u70ba \u6709\u96dc\u97f3 (Na)\u3001 (VC)\u3001\u96dc\u97f3(Na) \u7684\u65b9\u6cd5 \u6709 \u4e9b \u3002\u5982\u679c\u53ea\u8003\u616e\u95dc \u8a5e\uff0c\u5247 \u548c \u4f60 \u6703(\u80fd/\u53ef\u4ee5) \u958b \u958b\u7684\u539f\u56e0 B \u578b \u5f8c\u6703\u4e0d\u6703\u81ea\u52d5 \u6703\u81ea\u52d5 \u90fd\u70ba\u4ee5\u4e0a\u4e09\u53e5\u7684\u95dc \u8a5e\u3002\u5982\u6b64\u4e00\u4f86\uff0c \u6211\u5011\u5c31\u7121\u6cd5\u5f9e\u95dc \u8a5e\u4f86\u5224\u65b7\u7b2c\u4e00\u548c\u7b2c\u4e09\u53e5\u61c9\u8a72\u8f03\u63a5\u8fd1\uff0c\u56e0\u70ba\u6b64\u4e8c\u53e5\u7686 \u5728\u8a62\u554f \u7684\u65b9 \u6cd5\uff0c\u800c\u7b2c\u4e8c\u53e5\u5247\u662f\u5728\u8a62\u554f\u4e4b\u6240\u4ee5\u8981 3-3-2. \u53e5 \u8a9e\u52a9\u8a5e\u70ba \u7684 \u53ef\u4e0d\u53ef\u4ee5 \u7528 \u53ef\u4ee5 \u7528 4. \u8a5e\u610f\u6bd4\u5c0d \u7684\u554f\u53e5 \u7d93\u7531\u8a9e\u8a00\u5b78\u4e0a\u7684\u4e00\u4e9b\u7814\u7a76\u7d50\u679c\uff0c\u4ee5\u53ca\u5f9e \u96c6\u5230\u7684\u554f\u53e5\u4e2d\u6574\u7406\u6b78 \uff0c\u6211\u5011\u5b9a\u7fa9\u4e00\u5957\u7d50\u5408\u8a9e \u7684\u539f\u56e0\u3002 \u56e0\u6b64\uff0c\u4e00\u500b\u81ea\u7136\u8a9e\u8a00\u554f\u53e5\u4e2d\u7684 \u610f\u5716 \u6bb5 \uff0c\u6211\u5011\u5c07\u5176\u5b9a\u7fa9\u70ba\uff1a \u554f\u53e5\u4e2d\u6240\u50b3 \u6700\u76f4\u63a5\u60f3 \u7372\u5f97\u7684 \uff0c\u4e0d\u9700\u5305\u542b\u524d\u63d0\uff1bIS \u53ef\u4ee5\u662f\u554f\u53e5\u4e4b\u5b50\u53e5\u6216\u7247\u8a9e\uff0c \u81f3\u7d50\u5408\u5176\u4ed6\u7279\u5b9a\u7247\u8a9e\u800c\u6210\u3002 \u53e5 \u8a9e\u52a9\u8a5e\u554f\u53e5\u6307\u53e5\u5b50 \u6709\u4e00\u500b\u8a9e\u52a9\u8a5e\u50cf\u662f \u3001 \u3001 \u3001 \u672c\u8ad6\u6587\u4e2d\uff0c\u8a5e\u610f\u6bd4\u5c0d\u662f\u6240\u6709\u8a9e\u610f\u6bd4\u5c0d\u65b9\u6cd5\u7684\u57fa\u790e\uff0c\u50b3\u7d71\u8a9e\u8a00\u5b78\u8a8d\u70ba\u8a5e\u662f\u69cb\u6210\u8a9e\u610f\u7684\u6700\u5c0f \u7b49\u3002\u7576\u8a9e\u52a9 \u8a5e\u70ba \u6642\uff0c\u8a72\u554f\u53e5\u5c0d\u65bc \u76f8\u7576\u4e0d \u5b9a\uff0c\u800c\u9700\u8981\u8f03\u591a\u7684\u5916\u5728\u8a0a \u4e88\u89e3 \u3002\u9019\u985e\u578b\u7684\u554f \u6cd5\u898f\u5247\u8207\u8a9e\u610f\u7684\u8a9e\u610f\u6587\u6cd5\uff0c\u7576\u554f\u53e5\u7b26\u5408\u8a9e\u610f\u6587\u6cd5\u4e2d\u67d0\u4e00\u5247\u6642\uff0c\u5176\u76f8\u5c0d\u61c9\u7684 IS \u4e4b\u8403\u53d6\u65b9\u5f0f\u4e5f\u6e05 \u55ae\u5143[19]\uff0c\u800c\u76ee\u524d\u8a08\u7b97\u8a9e\u8a00\u5b78\u7684\u8da8\u5411\u662f \u8a5e\u8996\u70ba\u8a31\u591a \u8a9e\u610f\u6210\u5206 (semantic features)\u7684\u7d44\u5408\u3002 \u5716 2 How-net \u4e3b\u8981\u7279\u5fb5\u968e\u5c64\u5716 \u7684\u88ab\u898f\u7bc4\u8457\u3002\u8868 6 \u5217\u8209\u90e8\u5206\u8a9e\u610f\u6587\u6cd5\u53ca\u5176 IS \u8403\u53d6\u65b9\u5f0f\uff0c\u4e26\u8209\u4f8b\u8aaa\u660e\u4e4b\u3002 \u57fa\u65bc\u5f8c\u8005\uff0c\u6211\u5011\u5229\u7528\u77e5\u7db2(How-net)[17]\u4f5c\u70ba\u8a5e\u610f\u6bd4\u5c0d\u7684\u77e5\u8b58\u5eab\u3002 \u53e5\u5728\u53e5\u5b50\u4e2d\u901a\u5e38\u6703\u5305\u542b\u4e00\u500b \u6cd5\u76f8(modality)\u526f\u8a5e [16]\uff0c\u5982 \u6703 \u3001 \u53ef\u80fd \u3001 \u61c9\u8a72 \u3002 \u6cd5\u76f8 \u8868 6 \u90e8\u4efd\u8a9e\u610f\u6587\u6cd5\u53ca\u5176\u4f8b\u53e5 4-2. \u8a5e\u610f\u76f8 \u5ea6\u7684\u91cf \u900f\u904e\u5c0d\u65bc\u554f\u53e5\u7684\u5206\u6790\uff0c\u610f\u7fa9\u76f8\u540c\u537b\u4ee5\u4e0d\u540c\u53e5\u578b\u8868\u73fe\u7684\u554f\u53e5\uff0c\u6240\u8403\u53d6\u51fa\u4f86\u7684 IS \u61c9\u8a72\u80fd \u4fdd \u76f8 \u540c\u3002\u5982\u8868 1 \u6240\u793a\uff0c\u900f\u904e\u7684 KS \u53ca IS \u7684\u8403\u53d6\uff0c\u6211\u5011\u53ef\u4ee5 \u6613\u5730\u5206\u8fa8\u4e0a\u8ff0\u4f8b\u53e5\u7684\u7570\u540c\u3002 \u8868 1 \u4e09\u500b\u76f8 \u554f\u53e5\u6240\u5c0d\u61c9\u4e4b\u95dc \u8a5e \u6bb5(KS)\u53ca\u610f\u5716 \u6bb5(IS) \u554f\u53e5 KS IS \u3001 \u7684\u65b9\u6cd5 \u70ba \u8981 \u3001 \u7684\u539f\u56e0 \u7684\u65b9\u6cd5\u6709 \u4e9b \u3001 \u7684\u65b9\u6cd5 \u7684\u5b9a\u7fa9\u662f \u8aaa \u8005\u7684\u5c0d\u4e00\u500b\u53ef\u80fd\u4e8b\u4ef6\u7684 \u6cd5\u6216 \u5ea6 \uff0c\u6cd5\u76f8\u526f\u8a5e\u7684\u5b9a\u7fa9\u7531\u8a9e\u610f\u898f\u5b9a\uff0c\u5176\u6240\u5305\u542b \u554f\u53e5\u985e\u578b \u554f\u53e5 \u8a9e\u610f\u6587\u6cd5 IS 4-1. \u77e5\u7db2\u6982\u8ff0 \u57fa\u65bc\u5c0d\u77e5\u7db2\u7684\u7814\u7a76\uff0c\u6211\u5011\u5229\u7528\u77e5\u7db2\u5c0d\u65bc\u6bcf\u500b\u8a5e\u5f59\u5b8c\u6574\u7684\u5b9a\u7fa9\uff0c\u91cf \u5169\u500b\u8a5e\u5f59\u5728\u8a9e\u610f\u4e0a\u7684 \u7684\u8a5e\u6027\u542b\u6709\u4ee5\u5f80\u8a9e\u8a00\u5b78\u5206\u985e\u4e2d\u7684\u5927\u591a\u6578\u52a9\u52d5\u8a5e\u3001\u90e8\u5206\u52d5\u8a5e\u53ca\u52d5\u8a5e\uff0c\u4f46\u4ed6\u5011\u537b\u6709\u8a31\u591a\u5171\u540c\u7684\u8a9e \u7591\u554f\u8a5e\u554f\u53e5 \u70ba \u7522\u5f8c\u5fc5\u9808 \u7528\u751f\u5316 QW 1 NP Dba VP IS=VP \u7684\u539f\u56e0 \u7528\u751f\u5316 \u7684\u539f\u56e0 \u77e5\u7db2\u662f\u91dd\u5c0d\u96fb \u8a2d\u8a08\u7684\u96d9\u8a9e\u5e38\u8b58\u77e5\u8b58\u5eab\uff0c\u70ba \u5efa\u4eba \u5148\u751f\u7814\u7a76\u5341\u5e7e\u5e74\u7684\u91cd\u8981\u6210\u679c\uff0c \u76f8 \u5ea6\u3002\u540c\u4e00\u500b\u8a5e\u5f59\u901a\u5e38\u53ef\u8868\u793a\u4e00\u500b\u4ee5\u4e0a\u7684\u6982 \uff0c\u6240\u4ee5\u5169\u500b\u8a5e\u5f59\u7684\u76f8 \u5ea6\u53ef\u4ee5\u7531\u500b\u5225\u7684\u6982 \u6cd5\u7279\u8272\u3002\u800c\u6cd5\u76f8\u526f\u8a5e\u4e4b\u5f8c\u6240\u63a5\u7684\u662f\u52d5\u8a5e\u7247\u8a9e\uff0c\u6211\u5011\u8a8d\u70ba\u6b64\u52d5\u8a5e\u7247\u8a9e\u5373\u70ba\u5176\u610f\u5716\u6240\u5728\u3002\u8868 3 \u4e2d \u5217\u8209\u51fa\u90e8\u5206\u53e5 \u8a9e\u52a9\u8a5e\u70ba \u7684\u554f\u53e5\u53ca\u5176\u5c0d\u61c9\u7684 IS\u3002\u6b64\u5916\uff0c\u5982\u679c\u9019\u985e\u578b\u554f\u53e5\u4e0d\u542b\u6709\u4efb\u4f55\u6cd5 \u76f8\u526f\u8a5e\uff0c\u5247\u4ee5\u4e3b\u8981\u52d5\u8a5e\u7247\u8a9e\u4f5c\u70ba IS\uff0c\u5982\u8868 4 \u6240\u793a\u3002 \u8868 3 \u53e5 \u8a9e\u52a9\u8a5e\u70ba \u7684\u554f\u53e5\u53ca\u5176\u5c0d\u61c9\u7684\u610f\u5716 \u6bb5(IS) \u53e5 \u8a9e\u52a9\u8a5e\u554f\u53e5 \u4eba\u61c9 NP Dba VP \u61c9 \u63d0\u4f9b\u4e86\u8a2d\u8a08\u4eba\u5de5 \u9ad4\u6240\u9700\u7684\u77e5\u8b58\u3002\u77e5\u7db2\u5171 \u4e86 50220 \u500b\u4e2d\u6587\u8a5e\u8a9e\uff0c\u6240\u6db5 \u7684\u6982 \u7e3d\u91cf \u76f8 \u5ea6\u6c42\u5f97\uff0c\u800c\u6982 \u76f8 \u5ea6\u5247\u662f\u900f\u904e\u7279\u5fb5\u7684\u6bd4\u5c0d\u800c\u4f86\u3002\u5982\u516c\u5f0f(1)\u6240\u793a\uff0c\u4efb\u5169\u500b\u8a5e\u7684\u8a9e\u610f\u76f8 IS=VP \u4e2d\u7684 \u53ef\u4e0d\u53ef\u4ee5 P Dba1 not Dba2 VP \u53ef\u4ee5 \u7528 62174 \u500b\uff0c\u76ee\u524d\u4ecd\u5728 \u7576\u4e2d\u3002\u505a\u70ba\u4e00\u500b\u63d0\u4f9b\u4e2d\u6587\u8a08\u7b97\u9700\u6c42\u7684\u77e5\u8b58\u5eab\uff0c\u77e5\u7db2 \u76e1\u5730\u63cf\u8ff0\u4e86 \u5ea6( word Sim )\u88ab\u5b9a\u7fa9\u6210\u9019\u5169\u500b\u8a5e\u6240\u6709\u53ef\u80fd\u6982 \u5b9a\u7fa9\u4e4b\u9593\u76f8 \u5ea6( def Sim )\u7684\u6700\u5927\u503c\u3002 \u662f\u975e\u554f\u53e5 \u7528 IS=Dba2 VP \u6982 \u4e4b\u9593\u7684\u95dc\u4fc2\uff0c\u6982 \u6240\u5177\u6709\u7684\u5c6c\u6027\u4e4b\u9593\u7684\u95dc\u4fc2\uff0c\u4ee5\u53ca\u6982 \u8207\u6240\u5177\u6709\u7684\u5c6c\u6027\u4e4b\u9593\u7684\u95dc\u4fc2\u3002 1 1 2 2 1 2 1 2 ( ), ( ) ( , ) max ( , ) word def d def w d def w Sim w w Sim d d \u2208 = (1) \u2208 \u5c0d\u4e00\u500b\u8a5e\u800c\u8a00\uff0c\u5728\u4e0d\u540c\u60c5\u6cc1\u4e0b\u53ef\u80fd\u4ee3\u8868\u4e0d\u540c\u7684\u6982 \u3002\u77e5\u7db2\u5c07\u4e00\u500b\u6982 \u7684\u5b9a\u7fa9\u8868\u793a\u6210\u7279\u5fb5 \u6587\u6bd4\u5c0d\u5668 \u672c\u8ad6\u6587 \u7528\u5411\u91cf \u9593\u6a21\u578b\uff0c\u900f\u904e\u6bd4\u5c0d KS \u8207 FAQ \uff0c \u51fa\u6700\u9069\u5408\u56de \u8a72\u8a62\u554f\u53e5\u7684 \u3002 \u6839\u64da [3]\u7684\u5206\u6790\uff0c\u5c31\u8a9e\u6cd5\u5f62\u5f0f\u800c\u8a00\uff0c\u7591\u554f\u53e5\u53ef\u5206\u6210\u53e5\u5b50\u548c\u975e\u53e5\u5b50\u5169\u5927\u985e\uff0c\u518d\u6b78\u6210 \u7591 3-3. \u610f\u5716\u7684\u8403\u53d6 \u554f\u53e5 IS \u80fd \u65b7\u51fa\u6240\u6709 3-4. \u95dc \u8a5e\u7684\u8403\u53d6 \u7531\u65bc\u6211\u5011 \u7528 AutoTag \u505a\u70ba\u8a5e\u6027\u6a19\u8a18\u5de5\u5177\uff0c\u6240\u4ee5\u53ef \u9664\u90e8\u4efd\u8a5e\u7fa9 \u7684\u60c5\u5f62\u3002\u4f8b\u5982\uff1a\u7576 1 w \u53ca\u6a19\u8b58\u7b26 \u7684\u7d44\u5408\u3002\u8868 8 \u5217\u8209\u5e7e\u500b\u6982 \u5728\u77e5\u7db2\u4e2d\u4e4b\u5b9a\u7fa9\uff0c\u5176\u4e2d W_C \u70ba\u4e00\u6982 \uff0cG_C \u8868\u793a\u8a72 \u80fd \u65b7\u51fa\u6240\u6709 \u4eba\u61c9 \u61c9 \u76f8\u5c0d\u65bc\u610f\u5716\u7684\u8403\u53d6\uff0c\u95dc \u8a5e\u7684\u8403\u53d6\u4e5f\u662f\u4e00\u500b\u4e0d\u53ef \u7565\u7684\u90e8\u5206\uff0c\u85c9\u7531\u95dc \u8a5e\u8403\u53d6\u6211\u5011\u53ef\u5f9e \u6982 \u7684\u8a5e\u985e\uff0cDEF \u5247\u70ba\u5176\u5b9a\u7fa9\u3002\u5728\u5b9a\u7fa9\u4e2d\uff0c\u7279\u5fb5\u9593\u4ee5 \u540c\u6642\u5305\u542b\u540d\u8a5e\u548c\u52d5\u8a5e\u7684\u6982 \u6642\uff0c\u82e5\u5176\u8a5e\u6027\u6a19\u8a18\u70ba\u52d5\u8a5e\uff0c\u5247\u5176\u4ed6\u540d\u8a5e\u985e\u7684\u6982 \u5c07\u4e0d\u4e88\u8003\u616e\u3002\u7531 \u9694\uff0c\u7b2c\u4e00\u500b\u7279\u5fb5\u7a31\u70ba\u4e3b\u8981\u7279\u5fb5\uff0c \u554f\u8a5e\u554f\u53e5 \u3001 \u9078 \u554f\u53e5 \u3001 \u53e5\u5c3e\u8a9e\u52a9\u8a5e\u554f\u53e5 \u3001 \u7368\u7acb\u8a9e\u52a9\u8a5e\u554f\u53e5 \u3001 \u662f\u975e\u554f\u53e5 \u3001 \u9644\u52a0\u554f\u53e5 \u554f\u53e5 \u51fa\u5176 KS\u3002\u5c0d\u4e2d\u6587\u800c\u8a00\uff0c\u65b7\u8a5e\u4ee5\u53ca\u8a5e\u6027\u6a19\u8a18\u7684\u554f\u984c\u4e00\u76f4 \u570b \u8a08\u7b97\u8a9e\u8a00\u5b78\u7684\u767c\u5c55\u3002\u672c \u65bc\u6982 \u7684\u5b9a\u7fa9\u662f\u7531\u4e3b\u8981\u7279\u5fb5\u53ca\u6b21\u8981\u7279\u5fb5\u6240\u5171\u540c\u63cf\u8ff0\uff0c\u6240\u4ee5\u4efb\u5169\u500b\u6982 \u7684\u76f8 \u5ea6\u53ef\u5b9a\u7fa9\u5982\u4e0b\uff1a \u8868\u793a\u6982 \u7684\u985e\u5225\u5c6c\u6027\uff0c\u5177\u6709\u4e0a\u4e0b\u4f4d\u95dc\u4fc2\uff0c\u5982\u5716 2 \u6240\u793a\uff1b\u5f8c\u9762\u6240\u63a5\u7684\u7279\u5fb5\u5247\u70ba\u6b21\u8981\u7279\u5fb5\uff0c\u7528\u4f86 \u53ca \u76f4\u8ff0\u554f\u53e5 \u7b49\u4e03\u500b\u985e\u578b\u3002\u5c31 \u901a\u529f\u80fd\u800c\u8a00\uff0c\u7591\u554f\u53e5\u53ef\u5206\u70ba\u5916\u5728\u8a0a \u554f\u53e5\u3001\u8a00 \u554f\u53e5\u3001\u95dc \u7814\u7a76\u4ee5 AutoTag \u505a\u70ba\u65b7\u8a5e\u53ca\u8a5e\u6027\u6a19\u8a18\u7684\u5de5\u5177\uff0c\u6b64 \u9ad4\u70ba\u4e2d\u7814\u9662\u8cc7\u8a0a\u6240 CKIP \u5c0f\u7d44\u6240\u7814\u767c\u7684\uff0c \u898f\u7bc4\u6982 \u7684\u5c6c\u6027\u3002 1 2 1 2 1 2 ( , ) ( , ) (1 ) ( , ) def PF SF Sim d d Sim pf pf Sim sf sf = + \u2212 (2)</td></tr></table>",
+                "type_str": "table",
+                "num": null,
+                "text": "\u540d\u5f9e\u7b2c 12.04 \u540d\u63d0\u5347\u5230\u7b2c 2.91 \u540d\uff0c\u4e14\u4f7f\u5f97\u524d\u5341\u540d\u7684\u53ec\u56de\u7387\u7531 78.06%\u63d0\u5347\u5230 95.11%\u3002"
+            },
+            "TABREF3": {
+                "html": null,
+                "content": "<table><tr><td colspan=\"2\">\u8868 9 \u57fa\u7dda\u7cfb\u7d71\u4e4b\u524d N \u540d\u53ec\u56de\u7387 \u4e8c\u3001\u5728\u8a9e\u610f\u76f8 \u5ea6\u65b9\u9762\uff0c\u6211\u5011 \u7528\u77e5\u7db2\u505a\u70ba\u8a5e\u610f\u76f8 \u5ea6\u91cf \u7684\u77e5\u8b58\u5eab\uff0c\u4f46\u662f\u77e5\u7db2\u4e2d\u6c92\u6709 6. \u5be6\u9a57\u7d50\u679c\u8207\u8a0e\u8ad6 Top N 1 2 3 4 5 6 7 8 9 \u8868 10 \u986f\u793a\u4f7f\u7528 Dice coefficient \u4e4b\u7d50\u679c\u70ba\u6700\u4f73\uff0c\u6240\u4ee5\u5728\u63a5\u4e0b\u4f86\u7684\u5be6\u9a57\u90fd \u7528 Dice coefficient \u5b9a\u7fa9\u7684\u8a5e\uff0c\u5247\u7121\u6cd5\u85c9\u7531 \u4f86\u91cf \u8a5e\u610f\u76f8 \u5ea6\u3002\u89e3\u6c7a\u7684\u65b9\u6cd5\u6709\u4e8c\uff1a\u4e00\u662f\u589e\u52a0\u672a\u5b9a\u7fa9\u8a5e 10 \u672c\u7814\u7a76\u4e2d\uff0c\u6211\u5011\u5be6\u9a57\u4f7f\u7528\u7684\u6a5f\u5668\u70ba Pentium III 450 \u500b\u4eba\u96fb \uff0c128 MB RAM\uff0c\u958b\u767c\u7528\u7684 \u4f86\u4f5c\u70ba\u6b21\u8981\u7279\u5fb5\u76f8 \u5ea6\u4e4b\u91cf \u65b9\u6cd5\u3002 \u5230\u77e5\u8b58\u5eab\u4e2d\uff0c\u53e6\u4e00\u500b\u662f \u51fa\u81ea\u52d5\u5efa\u7acb\u77e5\u8b58\u5eab\u7684\u65b9\u6cd5\u3002</td></tr><tr><td>\u7a0b\u5f0f\u8a9e\u8a00\u662f Microsoft Visual C++ 6.0\u3002\u9664\u4e86\u5be6\u9a57</td><td>\u4e4b\u5916\uff0c\u4e5f\u900f\u904e IIS 4.0 \u67b6\u8a2d\u4e86\u4e00\u500b\u7db2 \uff0c</td></tr><tr><td colspan=\"2\">\u958b\u653e \u7db2\u8def\u4e0a\u7684\u4f7f\u7528\u8005\u67e5\u8a62\uff0c\u7db2 \u5728 http://chinese.csie.ncku.edu.tw/faq/\u3002\u5728\u8a9e\u6599\u5eab\u7684 \u96c6\u65b9 \u9762\uff0c\u6211\u5011\u4ee5\u4eba\u5de5\u5728\u7db2\u8def\u4e0a \u96c6\u4e86 1,022 \u5247 FAQ\uff0c \u5bb9\u4e3b\u8981\u5305\u62ec \u4e09\u3001\u5728 \u5efa \u7acb \u610f \u5716 \u6bb5 \u6790 \u6a39 \u65b9 \u9762 \uff0c \u5c0d \u65bc \u6790 \u6642 \u5230 \u8a5e \u6027 \u4e0d \u660e \u78ba \u7684 \u554f \u984c 6-3-3. \u4e3b\u8981\u7279\u5fb5\u8207\u6b21\u8981\u7279\u5fb5\u4e4b\u7d50\u5408\u4fc2\u6578\u5be6\u9a57 6-3. \u8a5e\u610f\u76f8 \u5ea6\u4e4b\u5be6\u9a57 (ambiguity)\uff0c\u4ecd\u6709\u56f0\u96e3\u7121\u6cd5 \u3002\u8003\u616e\u73fe\u6709\u8cc7 \uff0c\u53ef\u4ee5\u5148\u5efa\u7acb\u6a5f\u7387 \u6790\u5668[4][15]\uff0c\u9032 \u6982 \u5b9a\u7fa9\u7684\u76f8 \u5ea6\u7531\u4e3b\u8981\u7279\u5fb5\u76f8 \u5ea6\u53ca\u6b21\u8981\u7279\u5fb5\u76f8 \u5ea6\u7d50\u5408\u800c\u4f86\uff0c\u56e0\u6b64\u672c\u5be6\u9a57\u7684\u5e0c\u671b\u5f97 \u4ee5\u53ca \u8cc7\u7406 \u76f8\u95dc\u4e4b FAQ\u3002 5-1-2. \u95dc \u8a5e \u6bb5\u76f8 \u5ea6 6-3-1. \u4e3b\u8981\u7279\u5fb5\u76f8 \u5ea6\u4e4b \u5ea6\u5f71\u97ff\u4fc2\u6578\u5be6\u9a57 \u5230\u7279\u5fb5\u7d50\u5408\u4fc2\u6578 \u5c0d\u7cfb\u7d71\u6548\u80fd\u7684\u5f71\u97ff\uff0c\u540c\u6a23\u5730\uff0c\u6211\u5011\u56fa\u5b9a\u4fc2\u6578 0 = \u8207 1 \u800c\u5efa\u7acb\u5305\u542b\u8a9e\u610f\u4e4b \u6790\u5668\u3002 = \u3002\u5982\u5716 5 \u6240\u793a\uff0c\u8a72 \u5728\u7cfb\u7d71\u8a55\u4f30\u65b9\u9762\uff0c\u6211\u5011 10 \u4f4d\u975e\u7cfb\u7d71\u958b\u767c\u4eba \uff0c\u4e26 \u77e5\u672c\u7db2 \u6240\u63d0\u4f9b\u8cc7\u8a0a\u7684 \u5bb9\u7bc4\u570d\uff0c \u5728\u91cf \u5169\u500b KS \u7684\u76f8 \u5ea6\u4e0a\uff0c\u6211\u5011\u505a\u4e86\u4e00\u500b\u5047\u8a2d\uff1a\u5c0d\u4efb\u4e00\u500b\u95dc \u8a5e\u800c\u8a00\uff0c\u4e0d\u6703\u6709\u5169\u500b\u6216 \u4ee5\u4eba\u5de5\u7684\u65b9\u5f0f\u5efa\u7acb 185 \u5247\u554f\u53e5\u4e26\u6a19\u8a18\u8207\u5176\u76f8\u95dc\u4e4b FAQ\u3002\u6709\u5225\u65bc\u95dc \u8a5e\u8cc7\u8a0a\u6aa2 \uff0c\u81ea\u7136\u8a9e\u8a00\u554f \u5be6\u9a57\u7d50\u679c\u986f\u793a\uff0c 0.3 = \u4f7f\u5f97\u5e73\u5747\u6b63\u78ba \u540d \u5230 5.89 \u70ba\u6700\u5c0f\u3002 \u56db\u3001\u5728\u81ea\u7136\u8a9e\u8a00\u7406\u89e3\u65b9\u9762\uff0c\u76ee\u524d\u7684\u7cfb\u7d71\u4e26\u672a\u5177 \u7406\u80fd\u529b\uff0c\u5728\u8a31\u591a\u60c5\u6cc1\u4e0b\uff0c\u8a5e\u8a9e\u7684\u7d44\u5408 \u5f9e\u516c\u5f0f(5)\u4e2d\u5f97\u77e5\uff0c\u4fc2\u6578 \u6c7a\u5b9a \u5ea6\u5c0d\u65bc\u4efb\u5169\u76f8\u9130\u7bc0\u9ede\u9593\u8ddd ( Cost )\u7684\u5f71\u97ff\u7a0b\u5ea6\uff0c\u70ba\u4e86 \u5169\u500b\u4ee5\u4e0a\u7684\u95dc \u8a5e\u8207 \u5c0d\u61c9\u3002\u800c\u9019\u7a2e\u5c0d\u61c9\u95dc\u4fc2 \u53ef\u4ee5\u4e00\u5c0d\u4e00\u5c0d\u61c9\u51fd\u6578\u8868\u793a\u4e4b\uff0c\u6240\u4ee5\u6211\u5011\u63d0\u51fa \u516c\u5f0f(12)\u4f86\u91cf \u5169\u500b KSs 1 1 2 { , , , } m K w w w = L \u548c 2 1 2 { , , , } n K t t t = \u7684\u76f8 \u5ea6\u3002 \u53e5\u4e4b\u610f\u5716\u8f03\u660e\u78ba\uff0c\u56e0\u6b64\u6bcf\u5247\u554f\u53e5\u6240\u5c0d\u61c9\u7684 \u76f8\u7576\u5c11\uff0c\u5e73\u5747\u53ea\u6709 1.36 \u5247\u3002\u56e0\u6b64\uff0c\u6211\u5011\u4e0d\u4f7f\u7528 \u51fa \u4e4b\u6700\u4f73\u503c\uff0c\u6211\u5011\u56fa\u5b9a\u516c\u5f0f(2)\u4e2d\u7684\u4fc2\u6578 1 \u53ef\u80fd \u53e6\u5916\u7684\u610f\u7fa9\u3002\u9019\u4e9b\u6703 \u5230\u4f46\u4ecd\u7121\u6cd5\u89e3\u6c7a\u7684\u554f\u984c\uff0c\u6709 \u672a\u4f86 \u7e8c\u5730\u7814\u7a76\u3002 = \uff0c\u4e5f\u5c31\u662f\u5b8c\u5168\u4ee5\u4e3b\u8981\u7279\u5fb5\u76f8 \u5ea6\u505a\u70ba\u8a5e\u610f\u76f8 \u5716 6 \u610f\u5716-\u95dc \u8a5e\u7d50\u5408\u4fc2\u6578 \u4e4b\u65bc\u5e73\u5747\u6b63\u78ba \u540d\u6bd4\u8f03\u5716 L 1 1 2 ( , ( )) ( , ) max A word i i i KS f Sim a f a Sim K K A = = \u2211 (12) \u5176\u4e2d f \u662f\u4e00\u500b\u5f9e A \u5230 B \u7684\u4e00\u5c0d\u4e00\u51fd\u6578\uff0c i a \u662f A \u4e2d\u7684\u4e00\u500b\u5143\u7d20\uff0c ( , ( )) word i i Sim a f a \u8868\u793a\u95dc \u8a5e i a \u8207\u5176\u5c0d\u61c9\u7684\u95dc \u8a5e\u7684\u8a5e\u610f\u76f8 \u5ea6\u3002\u5982\u540c\u524d\u4e00\u5c0f\u7bc0\uff0c\u9700\u7279\u5225 \u610f\uff1a\u82e5 m n \uff0c\u5247\u8a2d\u5b9a 1 A K = \u4e14 2 B K = \uff1b\u53cd\u4e4b\uff0c\u5247\u8a2d\u5b9a 2 A K = \u4e14 1 B K = \u3002 5-2. \u6587\u6bd4\u5c0d \u9664\u4e86\u554f\u53e5\u6bd4\u5c0d\u5916\uff0c\u6211\u5011\u4e5f\u5229\u7528\u554f\u53e5\u8207 FAQ \u7684\u6bd4\u5c0d\u4f86 \u52a9 \u51fa\u6240\u9700\u7684 \uff0c\u4f7f\u7528\u7684\u65b9 \u6cd5\u5247\u662f\u76ee\u524d\u88ab \u4f7f\u7528\u5728\u8cc7\u8a0a\u6aa2 \u61c9\u7528\u7684 vector space model (VSM)\u3002VSM \u4e3b\u8981\u5206\u6210\u5169\u500b\u6b65 \uff1a(1) \u8403\u53d6\u7279\u5fb5\u4e26\u4ee5\u5411\u91cf\u4f86\u63cf\u8ff0\u4e4b\uff0c(2)\u6bd4\u8f03\u5169\u500b\u7279\u5fb5\u5411\u91cf\u5728\u5411\u91cf \u9593\u4e2d\u7684 \u89d2\u3002\u672c\u7814\u7a76\u4e2d\uff0c \u7279\u5fb5\u5411\u91cf\u662f\u7531\u6bcf\u500b\u95dc \u8a5e\u7684 TF\u00d7IDF \u6b0a\u91cd\u6240\u69cb\u6210\u3002\u91dd\u5c0d\u554f\u53e5\u53ca FAQ \u6c42\u53d6\u500b\u5225\u7684\u7279\u5fb5\u5411\u91cf 1 2 { , , , } N u a a a = L \u548c 1 2 { , , , } N v b b b = L \uff1b\u7136\u5f8c\u5229\u7528\u9918 \u516c\u5f0f\u8a08\u7b97\u5176 \u89d2\uff0c \u89d2 \u5c0f\u8868\u793a\u5169\u5411\u91cf \u63a5\u8fd1\uff0c\u4ee5\u6b64\u505a\u70ba\u8a72\u554f\u53e5\u8207 FAQ \u7684\u76f8\u95dc\u7a0b\u5ea6\uff0c\u5982\u516c\u5f0f(13)\u6240\u793a\u3002 1 cos 2 2 1 1 ( , ) ( , ) N i i i content N N i i i i a b Sim u v u v a b = = = = = \u2211 \u2211 \u2211 (13) \u5176\u4e2d N \u8868\u793a\u7279\u5fb5\u5411\u91cf\u7684 \u5ea6\uff0c\u4e5f\u5c31\u662f\u8a5e\u5f59\u91cf\u3002 \u78ba\u7387(precision rate)\u4f86\u8861\u91cf\u7cfb\u7d71\u7684\u6548\u80fd\uff0c\u56e0\u70ba\u5373\u4f7f\u7b2c\u4e00\u540d\u5c31\u662f\u6b63\u78ba \uff0c \u78ba\u7387\u4ecd\u6703 \u8457\u540d \u6b21\u589e\u52a0\u800c\u905e\u6e1b\u3002\u6211\u5011\u63d0\u51fa\u4e00\u500b\u8f03 \u7576\u7684\u8a55\u4f30\u65b9\u5f0f \u5e73\u5747\u6b63\u78ba \u540d\uff0c\u5176\u5b9a\u7fa9\u5982\u4e0b\uff1a \u2211 ) ( AvgRank (14) 6-1. \u610f\u5716 \u6bb5\u8403\u53d6\u5be6\u9a57 \u6839\u64da\u8a9e\u6599\u5eab\u4e2d\u554f\u53e5\u7684\u8a9e\u6cd5\u578b \uff0c\u8a02\u5b9a\u4e86 85 \u689d\u8a9e\u610f\u6587\u6cd5\u3002\u70ba\u4e86 \u6839\u64da\u8a72\u8a9e\u610f\u6587\u6cd5\u6240\u8403\u53d6 \u51fa\u4f86\u7684 IS \u7684\u6b63\u78ba\u6027\uff0c\u4ee5\u4eba\u5de5\u5efa\u7acb 185 \u5247\u554f\u53e5\u4f86\u505a \uff0c\u4e26\u4ee5\u4eba\u5de5\u6aa2\u9a57\u662f\u5426\u7b26\u5408\u539f\u672c\u9810 \u7684\u7d50 \u679c\u3002\u6aa2\u9a57\u6642\uff0c\u82e5\u5176\u8aa4\u5dee\u4e0d\u5f71\u97ff\u610f\u5716\u7684\u8fa8\u5225\uff0c\u5247\u8996\u70ba\u6b63\u78ba\u8403\u53d6\uff0c\u7d93\u7d71\u8a08\u53ef \u5230 91.89%\u7684\u6b63\u78ba\u8403 \u53d6\u7387\uff0c\u5176\u4e2d\u7121\u6cd5\u6b63\u78ba\u8403\u53d6\u7684\u60c5\u6cc1\u53ef\u5206\u70ba\u4ee5\u4e0b\u5e7e\u7a2e\uff1a \u4e00\u3001 \u5c6c\u65bc\u7591\u554f\u8a5e\u554f\u53e5\u3001\u662f\u975e\u8a9e\u53e5\u3001\u53e5 \u8a9e\u52a9\u8a5e\u70ba \u4e4b\u5916\u7684\u554f\u53e5\uff0c\u7531\u65bc\u4e26\u672a\u5728\u8a9e\u610f\u6587 \u6cd5\u4e2d\u5b9a\u7fa9\u5176\u8403\u53d6\u65b9\u5f0f\uff0c\u6240\u4ee5\u5c6c\u65bc \u8d85\u8d8a\u6587\u6cd5\u7bc4\u570d (out-of-grammar) \u800c\u7121\u6cd5\u8403\u53d6\u3002 \u4e8c\u3001 \u554f\u53e5\u7d50\u69cb\u904e\u65bc\u8907\u96dc \u81f3 \u6709\u5169\u500b\u7591\u554f\u5b50\u53e5\uff0c\u5c0d\u65bc\u9019\u985e\u578b\u554f\u53e5\u76ee\u524d\u4ecd\u7121\u6cd5\u8655\u7406\u3002 \u4e09\u3001 \u5728 AutoTag \u65b7\u8a5e\u53ca\u6a19\u793a\u8a5e\u6027\u6642\u5df2\u7d93\u51fa\u932f\uff0c\u5c0e\u81f4\u5f8c\u9762\u610f\u5716\u8403\u53d6\u7121\u6cd5\u6b63\u78ba\u5224\u65b7\u3002 6-2. \u57fa\u6e96\u7cfb\u7d71 \u672c\u5be6\u9a57\u4ee5\u95dc \u8a5e\u67e5\u8a62\u70ba\u57fa\u6e96(baseline)\uff0c\u8207\u81ea\u7136\u8a9e\u8a00\u67e5\u8a62\u505a\u6bd4\u8f03\u3002\u56e0\u6b64\u6211\u5011 \u516c\u5f0f(8)\u4e2d\u7684 \u4fc2\u6578 0 = \uff0c\u4f7f\u5f97 \u7531 \u5bb9\u6bd4\u5c0d\u4f86\u6c7a\u5b9a\u6574\u9ad4\u4e4b\u76f8 \u5ea6\u3002\u7d93\u7531\u7d71\u8a08\u6bcf\u4e00\u689d \u53e5\u4e4b \u540d\uff0c\u7d50 \u679c\u7372\u5f97\u5e73\u5747\u6b63\u78ba \u540d\u70ba 12.04 \u540d\uff0c\u4e26\u5f97\u5230\u524d N \u540d\u7684\u53ec\u56de\u7387(recall rate)\u8868\u5217\u5982\u4e0b\uff1a \u5ea6\uff1b\u516c\u5f0f(9)\u4e2d\u7684\u4fc2\u6578 =0\uff0c\u8868\u793a\u4e0d\u8003\u616e IS \u5c0d\u554f\u53e5\u76f8 \u5ea6\u4e4b\u5f71\u97ff\uff1b\u516c\u5f0f(8)\u4e2d\u7684\u4fc2\u6578 =1\uff0c\u8868 \u793a\u5b8c\u5168\u4ee5\u554f\u53e5\u76f8 \u5ea6\u4f5c\u70ba\u6aa2 \u7684\u4f9d\u64da\uff0c\u7136\u5f8c\u6839\u64da\u5e73\u5747\u6b63\u78ba \u540d\u4f86\u6c7a\u5b9a \u4e4b\u6700\u4f73\u503c\u3002 \u5716 4 \u4fc2\u6578 \u76f8\u5c0d\u65bc\u5e73\u5747\u6b63\u78ba \u540d\u4e4b\u6bd4\u8f03\u5716 \u5982\u5716 4 \u6240\u793a\uff0c\u7576 4.0 = \u6642\uff0c\u5176\u5e73\u5747\u6b63\u78ba \u540d 13.54 \u70ba\u6700\u4f73\u7d50\u679c\uff0c\u6b64\u7d50\u679c\u8207 \u5ea6\u8d8a \u5247\u7bc0\u9ede\u9593\u8ddd \u8d8a\u77ed \u7684\u89c0\u9ede\u76f8\u7b26\u3002 6-3-2. \u6b21\u8981\u7279\u5fb5\u76f8 \u5ea6\u8a08\u7b97\u65b9\u5f0f\u5be6\u9a57 \u672c\u5be6\u9a57\u6bd4\u8f03\u56db\u7a2e\u4e8c\u5143\u5411\u91cf\u76f8 \u5ea6\u91cf \u65b9\u5f0f\u5c0d\u7cfb\u7d71\u6548\u80fd\u7684\u5f71\u97ff\u3002\u6211\u5011\u56fa\u5b9a 0 = \uff0c\u4e5f\u5c31\u662f\u5b8c \u5168\u4ee5\u6b21\u8981\u7279\u5fb5\u76f8 \u5ea6\u70ba\u4e3b\uff0c 0 = \u5373\u4e0d\u8003\u616e IS\uff0c 1 = \u5b8c\u5168\u4ee5\u554f\u53e5\u6bd4\u5c0d\u4f86\u8a55\u4f30\uff0c\u7d50\u679c\u8868\u5217\u5982\u4e0b\uff1a \u8868 10 \u6bd4\u8f03\u5404\u7a2e\u4e8c\u5143\u5411\u91cf\u76f8 \u5ea6\u91cf \u4fc2\u6578\u5c0d\u7cfb\u7d71\u5e73\u5747\u6b63\u78ba \u540d\u4e4b\u5f71\u97ff Dice coefficient Jaccard coefficient Overlap coefficient Cosine \u5716 5 \u7279\u5fb5\u7d50\u5408\u4fc2\u6578\u6b63\u78ba \u540d\u6bd4\u8f03\u5716 6-4. \u53e5\u610f\u76f8 \u5ea6\u5be6\u9a57 \u7531\u516c\u5f0f(9)\uff0c\u672c\u5be6\u9a57\u60f3\u4e86\u89e3\u610f\u5716-\u95dc \u8a5e\u7d50\u5408\u4fc2\u6578 \u5c0d\u7cfb\u7d71\u6548\u80fd\u7684\u5f71\u97ff\uff0c\u56e0\u6b64\u56fa\u5b9a\u5be6\u9a57\u503c 4 = \u8207 0.3 = \u4ee5\u53ca\u5c1a\u672a\u5be6\u9a57\u7684 1 = \u3002\u7531\u5716 6 \u5f97\u77e5\uff0c\u7576 0.3 = \u6642\uff0c\u5176\u5e73\u5747\u6b63\u78ba \u540d 3.59 \u70ba\u6700\u4f73\u7d50\u679c\u3002\u6b64\u5916\uff0c\u7576 \u8f03\u5927\u6642\uff0c\u66f2\u7dda\u8fc5\u901f\u4e0a \uff0c\u8868\u793a\u7576 IS \u76f8 \u5ea6\u7684\u6bd4\u91cd\u904e\u5927\u6642\uff0c\u5176\u7d50\u679c\u4e26 \u4e0d\u7406\u60f3\u3002\u9019\u662f\u56e0\u70ba IS \u5305\u542b\u554f\u984c\u7684\u610f\u5716\uff0c\u4e26\u672a\u5c07\u524d\u63d0\u5305\u542b\u9032\u4f86\uff1b\u56e0\u6b64\uff0cIS \u4e26\u4e0d\u80fd\u5b8c\u5168\u53d6\u4ee3 KS\uff0c\u800c\u662f\u76f8\u8f14\u76f8\u6210\u3002 6-5. \u554f\u53e5\u76f8 \u5ea6\u8207 \u6587\u76f8 \u5ea6\u4e4b\u7d50\u5408\u4fc2\u6578\u5be6\u9a57 \u7531\u516c\u5f0f(8)\uff0c\u554f\u53e5\u8207 FAQ \u6a23\u672c\u7684\u6bd4\u5c0d\u7531\u554f\u53e5\u7684\u76f8 \u5ea6\u8207 \u6587\u76f8 \u5ea6\u5171\u540c\u6c7a\u5b9a\uff0c\u56e0\u6b64\u672c\u5c0f\u7bc0 \u5be6\u9a57\u5176\u4fc2\u6578 \u3002\u5be6\u9a57\u7d50\u679c\u986f\u793a\uff0c 0.5 = \u6642\uff0c\u5176\u5e73\u5747\u6b63\u78ba \u540d\u843d\u5728 2.91 \u70ba\u6700\u4f73\u7d50\u679c\u3002\u89c0\u5bdf \u5716 7\uff0c \u5728\u7bc4\u570d[0.2, 1.0]\u4e2d\u6642\uff0c\u5c0d\u7cfb\u7d71\u6548\u80fd\u7684\u5f71\u97ff\u4e26\u4e0d\u5927\uff1b\u53ef\u5f97\u77e5\uff0c\u76f8\u8f03\u65bc \u6587\u76f8 \u5ea6\uff0c\u554f\u53e5 \u76f8 \u5ea6\u5c0d\u7cfb\u7d71\u6548\u80fd\u7684\u5f71\u97ff\u8f03\u5927\u3002 \u5716 7 \u6bd4\u5c0d\u7d50\u5408\u53c3\u6578 \u4e4b\u65bc\u5e73\u5747\u6b63\u78ba \u540d\u6bd4\u8f03\u5716 6-6. \u5be6\u9a57\u7e3d\u7d50 \u6700\u5f8c\uff0c\u85c9\u7531 \u5236\u53c3\u6578\uff0c\u5c07\u5404\u500b\u65b9\u6cd5\u4ee5\u5e73\u5747\u6b63\u78ba \u540d\u8207\u53ec\u56de\u7387\u505a\u4e00\u500b\u6bd4\u8f03\u3002\u7531\u5716(8)\u548c \u5716(9)\u53ef\u4ee5\u767c\u73fe\uff0c\u7121\u8ad6\u5f9e\u5e73\u5747\u6b63\u78ba \u540d\u6216\u662f\u524d N \u540d\u7684\u53ec\u56de\u7387\u4f86 \uff0c\u672c\u8ad6\u6587\u6240\u63d0\u51fa\u7684\u65b9\u6cd5\u660e \u986f\u5730\u6539 \u4e86\u6548\u80fd\u3002\u76f8\u8f03\u65bc\u57fa\u6e96\u7cfb\u7d71\uff0c\u5e73\u5747\u6b63\u78ba \u554f\u53e5\u7684\u95dc \u8a5e\uff0cIS \u8868\u793a\u53ea\u6bd4\u8f03\u554f\u53e5\u7684\u610f\u5716 \u5716 9 \u7cfb\u7d71\u53ec\u56de\u7387\u6bd4\u8f03\u5716\uff0c\u5176\u4e2d baseline \u8868\u793a\u53ea\u6bd4\u8f03 \u6587\u7684\u95dc \u8a5e\uff0cKS \u8868\u793a\u53ea\u6bd4\u8f03\u554f\u53e5\u7684\u95dc \u8a5e\uff0cIS \u8868\u793a\u53ea\u6bd4\u8f03\u554f\u53e5\u7684\u610f\u5716 7. \u7d50\u8ad6\u8207\u672a\u4f86\u5c55\u671b \u672c\u8ad6\u6587\u63d0\u51fa\u4ee5\u554f\u53e5\u610f\u5716\u8403\u53d6\u4ee5\u53ca\u8a9e\u610f\u6bd4\u5c0d\u7684\u65b9\u6cd5\uff0c\u61c9\u7528\u5230\u81ea\u7136\u8a9e\u8a00 FAQ \u6aa2 \u4e0a\u3002\u7d93\u5be6\u9a57 \u9a57\u8b49\uff0c\u8a72\u65b9\u6cd5\u78ba\u5be6\u6bd4\u55ae\u7d14\u4f7f\u7528\u95dc \u8a5e\u67e5\u8a62\u4f86\u5f97\u6e96\u78ba\uff0c\u4f7f\u5e73\u5747\u6b63\u78ba \u7684 \u540d\u5f9e\u7b2c 12.04 \u540d\u63d0 \u5347\u5230\u7b2c 2.91 \u540d\uff0c\u4e14\u4f7f\u5f97\u524d\u5341\u540d\u7684\u53ec\u56de\u7387\u7531 78.06%\u63d0\u5347\u5230 95.11%\uff0c\u4f46\u662f\u5176\u4e2d\u4ecd\u5b58\u5728\u4e00\u4e9b \u6539 \u9032\u4e4b\u8655\uff1a \u4e00\u3001\u610f\u5716\u8403\u53d6\u65b9\u9762\uff0c\u96d6\u7136\u6211\u5011\u80fd\u8655\u7406 92%\u7684\u8a9e\u6599\uff0c\u4ecd\u6709\u8a31\u591a\u554f\u53e5\u7684\u578b \u4e0d\u5728 \u96c6\u7684\u7bc4\u570d \u540d \u9032\u6b65\u4e86 9 \u5716 8 \u7cfb\u7d71\u5e73\u5747\u6b63\u78ba \u540d\u6bd4\u8f03\u5716\uff0c\u5176\u4e2d baseline \u8868\u793a\u53ea\u6bd4\u8f03 \u6587\u7684\u95dc \u8a5e\uff0cKS \u8868\u793a\u53ea\u6bd4\u8f03 \u53c3\u8003\u6587\u737b</td></tr><tr><td colspan=\"2\">\u5e73\u5747\u6b63\u78ba \uff0c\u4ee5\u53ca\u5c0d\u65bc\u8f03\u8907\u96dc\u8a9e\u6cd5\u554f\u53e5\u7684\u8aa4\u5224\uff0c\u53ef\u4ee5\u85c9\u7531\u6539 \u8a9e\u610f\u6587\u6cd5\u4e0a\u4f86\u89e3\u6c7a\u3002 \u540d 6.28 6.29 7.61 6.51</td></tr></table>",
+                "type_str": "table",
+                "num": null,
+                "text": "\u53ec\u56de\u7387 (%) 36.06 48.56 56.00 60.22 63.89 66.56 73.56 73.56 76.72 78.06 \u500b\u540d\u6b21\u3002 \u7b2c\u4e00\u540d\u7684\u53ec\u56de\u7387 \u5f9e 36.06%\u63d0\u5347\u5230 64.67%\uff0c \u63d0 \u4e86 80%\uff1b\u800c\u524d\u5341\u540d\u7684\u53ec\u56de\u7387\u4e5f\u5f9e 78.06%\u63d0\u5347\u5230 95.11%\u3002"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-1008.json

@@ -0,0 +1,55 @@
+{
+    "paper_id": "O00-1008",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:14.689944Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-1008",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "\u672c\u6587\u4e3b\u8981\u6839\u64da\u7b46\u8005\u570b\u79d1\u6703\u8a08\u756b\u6210\u679c\u5831\u544a(\"\u53f0\u7063\u570b\u8a9e\u3001\u95a9\u5357\u8a9e\u548c\u5ba2\u5bb6\u8a71\u5404\u985e\u60c5\u614b\u8a5e\u642d\u914d\u95dc\u4fc2\u7684",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u5176\u4e2d\u300c\u53ef\u80fd\u300d\u548c\u300c\u5fc5\u8981\u300d\u88ab\u300c\u6c92(\u6709)\u300d\u5426\u5b9a\u6642\uff0c\u8a5e\u985e\u4e0a\u7a76\u61c9\u5206\u6790\u70ba\u540d\u8a5e\u6216\u52a9\u52d5\u8a5e\u6709\u5f85\u9032\u4e00\u6b65\u7814 \u7a76\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u5e73\u8861\u8a9e\u6599\u5eab\u7684\u53d6\u6a23\u5076\u6709\u91cd\u8907\u7684\u60c5\u5f62\uff0c\u5982\u4f8b(22)\u5373\u91cd\u8907\u51fa\u73fe\uff0c\u5be6\u969b\u4e0a\u662f 11 \u4f8b\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {},
+        "ref_entries": {
+            "TABREF0": {
+                "type_str": "table",
+                "num": null,
+                "html": null,
+                "text": "\u5f9e\u8a9e\u6599\u5eab\u770b\u6f22\u8a9e\u52a9\u52d5\u8a5e\u7684\u8a9e\u6cd5\u7279\u9ede\u912d\u7e08 \u975c\u5b9c\u5927\u5b78\u4e2d\u6587\u7cfb E-mail: ycheng@pu.edu.tw",
+                "content": "<table><tr><td>4)\u4e0d\u80fd\u5728\u4e3b\u8a9e\u4e4b\u524d \u5b9a\u5f0f\u3001\u52a9\u52d5\u8a5e\u7684\u4f4d\u7f6e\u548c\u6b63\u53cd\u554f\u53e5\u7b49\u56db\u5c0f\u7bc0\uff0c\u4ee5\u5be6\u969b\u8a9e\u6599\u6aa2\u9a57 17 \u500b\u8a5e\u7684\u5206\u5e03\u3002\u7d50 \u8a9e\u6599\u4e2d\uff0c \u300c\u5f88+\u6703\u300d8 \u6b21\u90fd\u662f\u80fd\u529b\u7fa9\uff0c\u4f46\u300c\u66f4+\u6703\u300d\u6709 9 \u6b21\uff0c\u8868\u9810\u65b7\u8005\u591a\u65bc\u80fd\u529b\u7fa9 (6)\u6620\u6f14\u696d\u8005\u9664\u4e86\u8981\u52a0\u5f37\u516c\u5171\u5b89\u5168\u8a2d\u65bd\u5916\uff0c\u66f4\u61c9\u52a0\u5f37\u57f7\u884c\u96fb\u5f71\u7684\u4e09\u7d1a\u5236\u7ba1 \u5728\u9019\u793e\u6703 \u8868\u6bd4\u8f03\u7684\u7a0b\u5ea6\u526f\u8a5e\u300c\u66f4\u300d\u7684\u4fee\u98fe\u3002 \u800c\u300c\u53ef\u4ee5\u300d(\u51fa\u73fe 11 \u4f8b)\u5247\u4e0d\u4e00\u5b9a\uff1a \u62ec\u6f14\u8b1b\u7a3f\u3001\u5287\u5834\u53f0\u8a5e\u3001\u6703\u8a71\u53ca\u6703\u8b70\u8a18\u9304)\u50c5\u4f54\u5168\u90e8\u7684\u5341\u5206\u4e4b\u4e00 12 \uff0c\u5f85\u5c07\u4f86\u8a9e\u6599\u5eab\u9054 \u7528\u7684\u983b\u7387\u4e0a\uff0c\u975e\u60c5\u614b\u7528\u6cd5\u9ad8\u65bc\u60c5\u614b\u7528\u6cd5\u8005\u3002\u5982\u300c\u61c9\u8a72\u300d \u3001 \u300c\u61c9\u7576\u300d \u3001 \u300c\u61c9\u300d\u6216\u300c\u8a72\u300d \u4f46\u4ee5\u5e73\u8861\u8a9e\u6599\u5eab\u4e2d\u5404\u52a9\u52d5\u8a5e\u7684\u7528\u4f8b\u4f86\u770b\uff0c\u52a9\u52d5\u8a5e\u4f7f\u7528\u6b63\u53cd\u554f\u53e5\u7684\u983b\u7387\u4e0d\u50c5\u5f88\u4f4e\uff0c</td></tr><tr><td>5) \u52a9\u52d5\u8a5e\u5fc5\u9700\u548c\u52d5\u8a5e\u4e00\u8d77\u51fa\u73fe\uff0c\u9664\u975e\u56e0\u4e0a\u4e0b\u6587\u800c\u7701\u7565\u52d5\u8a5e \u679c\u986f\u793a\u591a\u6578\u52a9\u52d5\u8a5e\u4e26\u672a\u5b8c\u5168\u7b26\u5408\u5b78\u8005\u6240\u63d0\u7684\u8a9e\u6cd5\u7279\u9ede\uff0c\u7b2c\u4e94\u5c0f\u7bc0\u9032\u4e00\u6b65\u6307\u51fa\u5b78\u8005 a.\u300c\u5f88+\u6703\u300d \u7406 (2)\u662f\u81ea\u5df1\u7684\u8ca0\u9762\uff0c\u81ea\u5df1\u62cb\u68c4\u4e86\u624d\u8ddf\u6c23\uff0c\u6c92\u80fd\u4fee\u6210\u6b63\u679c\uff1b\u800c\u5728\u4eba\u6211\u4e4b\u9593\u7528 (13) \u9ce5\u7c60\u5c31\u5728\u9435\u76ae\u9644\u8fd1\uff0c\u6211\u5011\u90fd\u71b1\uff0c\u5c0f\u9ce5\u66f4\u4e0d\u5fc5\u8aaa\u4e86\uff0c\u7c21\u76f4\u5feb\u8b8a\u6210\u70e4 2.3.2 \u52a9\u52d5\u8a5e\u51fa\u73fe\u65bc\u53e5\u4e2d (10) \u7406\u8ad6\u4e0a\u53ef\u4ee5\uff0c\u800c\u5be6\u969b\u4e0a\u4e0d\u884c\u3002(\u5c0d\u7167\u53e5) \u53e5\u9996 \u53e5\u4e2d \u53e5\u5c3e \u51fa\u73fe\u6b21\u6578 \u983b\u7387% \u5230\u53e3\u8a9e\u8207\u66f8\u9762\u8a9e\u7684\u8cea\u548c\u91cf\u76f8\u7576\u5f8c\uff0c\u6216\u53ef\u77ad\u89e3\u6b63\u53cd\u554f\u53e5\u662f\u5426\u70ba\u52a9\u52d5\u8a5e\u7684\u4e00\u500b\u57fa\u672c\u7279 \u90fd\u53ef\u8868\u793a\u60c5\u7406\u4e0a\u5fc5\u9808\u5982\u6b64\uff0c\u6216\u4f30\u8a08\u5fc5\u7136\u5982\u6b64\uff0c\u751a\u81f3\u6709\u4e9b\u8a5e\u9084\u6709\u5176\u4ed6\u975e\u60c5\u614b\u7528 \u800c\u4e14\u51fa\u73fe\u9019\u7a2e\u53e5\u578b\u7684\u52a9\u52d5\u8a5e\u53ea\u6709\u300c\u80af\u3001\u8a72\u3001\u9858\u610f\u3001\u61c9\u8a72\u3001\u53ef\u4ee5\u3001\u6703\u3001\u61c9\u300d7 \u500b\u3002</td></tr><tr><td>6) \u4e0d\u80fd\u5e36\u6642\u8c8c\u8a5e \u7684\u7406\u8ad6\u8207\u8a9e\u6599\u4e0d\u5b8c\u5168\u4e00\u81f4\uff1a\u5982\u6709\u5c11\u6578\u60c5\u614b\u8a5e\u7d55\u5927\u591a\u6578\u7528\u70ba\u540d\u8a5e\u6216\u540d\u8a5e\u7684\u4fee\u98fe\u8a9e\uff1b (1)\u5de8\u4eba\u65cf\u88e1\u7684\u4e00\u500b\u82f1\u96c4\u3002\u4ed6\u6709\u5169\u689d\u9577\u817f\uff0c\u5f88\u6703\u8dd1\u3002 (7)\u540c\u6642\uff0c\u5728\u6c61\u6c34\u8655\u7406\u65b9\u9762\uff0c\u66f4\u61c9\u6709\u9577\u9060\u7684\u898f\u5283\uff0c\u4ee5\u514d\u5e36\u4f86\u6c34\u6c61\u67d3\u7684\u554f\u984c\u3002 \u624d\u6c23\u4f86\u58d3\u57ae \u5c0f\u9ce5\u4e86\uff0c\u6211\u5011\u53ea\u597d \u52a9\u52d5\u8a5e\u51fa\u73fe\u65bc\u53e5\u4e2d\u4e14\u5f8c\u63a5\u52d5\u8a5e\u662f\u6700\u5e38\u898b\u7684\u60c5\u5f62\uff0c\u63db\u8a00\u4e4b\uff0c\u53e5\u4e2d\u662f\u52a9\u52d5\u8a5e\u6700\u5178\u578b\u7684 (11)\u660e\u5e74\u5230\u65e5\u672c\u6216\u662f\u6b50\u6d32\u6253\u7403\u90fd\u53ef\u4ee5\uff0c\u5979\u53ea\u5728\u4e4e\u80fd\u5426\u589e\u9032\u81ea\u5df1\u7684\u7403\u6280 \u53ef\u80fd \u253c \u253c \u253c \u80af 1/151 0.66 \u8272\u3002\u82e5\u5c31\u76ee\u524d\u8a9e\u6599\u770b\u4f86\uff0c\u6211\u5011\u7121\u6cd5\u628a\u6b63\u53cd\u554f\u53e5\u505a\u70ba\u5224\u65b7\u52a9\u52d5\u8a5e\u7684\u4e00\u500b\u91cd\u8981\u7279\u5fb5\u3002 \u6cd5\u3002\u7136\u800c\u5728\u5be6\u969b\u8a9e\u8a00\u4f7f\u7528\u4e2d\uff0c\u9019\u4e9b\u8a5e\u51fa\u73fe\u7684\u6b21\u6578\u4e26\u4e0d\u76f8\u540c\uff0c\u5404\u7a2e\u7528\u6cd5\u7684\u983b\u7387\u4e0d \u518d\u5c31\u52a9\u52d5\u8a5e\u7684\u4f4d\u7f6e\u800c\u8a00\uff0c\u7d93\u5168\u9762\u6aa2\u8a0e\u52a9\u52d5\u8a5e\u53ef\u51fa\u73fe\u7684\u4f4d\u7f6e(\u53e5\u9996\u3001\u53e5\u4e2d\u751a\u6216\u53e5</td></tr><tr><td>\u6458\u8981 \u672c\u6587\u4ee5\u4e2d\u7814\u9662\u7684\u5e73\u8861\u8a9e\u6599\u5eab\u70ba\u57fa\u790e\uff0c\u91cd\u65b0\u6aa2\u8a0e\u52a9\u52d5\u8a5e\u7684\u8a9e\u6cd5\u7279\u9ede\u3002Li &amp; Thompson \u4e3b\u5f35\u52a9\u52d5\u8a5e\u4e0d\u80fd\u88ab\u7a0b\u5ea6\u526f\u8a5e(\u5f88/\u66f4)\u4fee\u98fe\uff0c\u6e6f&amp;\u6e6f\u537b\u4ee5\u4e4b\u5340\u5206\u60c5\u614b\u52d5\u8a5e\u548c\u5f62\u5bb9\u8a5e\uff0c \u7d50\u679c\u8a9e\u6599\u986f\u793a\u591a\u6578\u52a9\u52d5\u8a5e\u53ef\u8207\u300c\u5f88\u300d\u6216\u300c\u66f4\u300d\u642d\u914d\uff1b\u518d\u5247\u5c31\u52a9\u52d5\u8a5e\u7684\u4f4d\u7f6e\u800c\u8a00\uff0c 7)\u4e0d\u80fd\u540d\u8a5e\u5316 8)\u4e0d\u80fd\u5e36\u540d\u8a5e\u7d44\u8cd3\u8a9e 2 Li &amp; Thompson \u8a8d\u70ba\u5c31\u4e00\u3001\u4e8c\u9ede\u800c\u8a00\uff0c\u52a9\u52d5\u8a5e\u548c\u52d5\u8a5e\u7121\u5225\uff0c\u4f46\u4e8c\u8005\u5728\u5f8c\u516d\u9ede\u4e0a\u6709 \u6240\u4e0d\u540c\uff0c\u6545\u4e3b\u5f35\u52a9\u52d5\u8a5e\u61c9\u55ae\u7368\u6210\u4e00\u985e\u3002\u4ed6\u6839\u64da\u9019\u516b\u500b\u7279\u9ede\u5217\u51fa\u7684\u52a9\u52d5\u8a5e\u6709\uff1a \u300c\u61c9 \u8a72\u3001\u61c9\u7576\u3001\u8a72\u3001\u80fd\u5920\u3001\u6703\u3001\u53ef\u4ee5\u3001\u80fd\u3001\u6562\u3001\u80af\u3001\u5f97\u3001\u5fc5\u9808\u3001\u5fc5\u8981\u3001\u5fc5\u5f97\u300d\u7b49 13 \u4e5f\u6709\u6240\u8b02\u7684\u52a9\u52d5\u8a5e\uff0c\u5176\u975e\u60c5\u614b\u7528\u6cd5\u9ad8\u65bc\u60c5\u614b\u7528\u6cd5\uff0c\u9019\u4e9b\u90fd\u61c9\u6392\u9664\u65bc\u52a9\u52d5\u8a5e\u4e4b\u5916\u3002 2.1 \u52a9\u52d5\u8a5e\u8207\u7a0b\u5ea6\u526f\u8a5e \u300c\u61c9\u8a72\u3001\u80fd\u5920\u3001\u6703\u3001\u53ef\u4ee5\u3001\u80fd\u3001\u53ef\u4ee5\u300d\u7b49\u52a9\u52d5\u8a5e\u662f\u5426\u53ef\u88ab\u7a0b\u5ea6\u526f\u8a5e(\u5982\u300c\u5f88\u300d\u6216 \u300c\u66f4\u300d\u7b49)\u4fee\u98fe\uff0cLi &amp; Thompson (1981)\u548c\u6e6f&amp;\u6e6f(1998)\u5169\u6587\u7684\u770b\u6cd5\u76f8\u53cd\uff0c\u6b64\u8655\u5c07 (2)\u597d\u559c\u6b61\u9019\u500b\u7709\u6bdb\u90fd\u958b\u59cb\u8b8a\u767d\u7684\u7537\u4eba\uff0c\u4f60\u5f88\u6703\u8a87\u734e\u5225\u4eba\u3002 (3)\u570b\u7236\u56e0\u70ba\u5f9e\u5c0f\u53d7\u5230\u7956\u6bcd\u7684\u5f71\u97ff\uff0c\u6240\u4ee5\u5f88\u6703\u8aaa\u6545\u4e8b\uff0c\u53e3\u624d\u4e5f\u5f88\u597d\u3002 b.\u300c\u66f4+\u6703\u300d(\u6b64\u8655\u300c\u6703\u300d\u5f80\u5f80\u8868\u9810\u65b7\u7fa9) (4)\u5f35\u5b78\u53cb\u8868\u793a\uff0c\u5433\u5c0e\u6f14\u4e0d\u4f46\u6703\u6559\u6232\uff0c\u66f4\u6703\u6f14\u6232\uff0c\u4ed6\u5f97\u734e\uff0c\u5f35\u5b78\u53cb\u7121\u8a71\u53ef\u8aaa\u3002 (5)\u7279\u5225\u662f\u5728\u4e8b\u696d\u6709\u4e86\u57fa\u790e\uff0c\u5e74\u904e\u58ef\u5e74\uff0c\u66f4\u6703\u53bb\u8ffd\u554f\uff1a \u300c\u4eba\u751f\u7a76\u7adf\u662f\u4ec0\u9ebc\uff1f\u300d (8)\u5728\u5bb6\u5ead\u4e2d\u51fa\u73fe\u7684\u6a5f\u7387\u6108\u4f86\u6108\u591a\uff0c\u6c11\u773e\u66f4\u61c9\u5177\u5099\u6025\u6551\u5e38\u8b58\uff0c\u4ee5\u514d\u7a81\u81e8\u610f \u5916\u4e8b\u6545\u6642 (9)\u7814\u7a76\u7684\u5949\u737b\u614b\u5ea6 dedication\uff0c\u662f\u7814\u7a76\u751f\u6700\u8a72\u5177\u5099\u7684\u7279\u8cea\uff0c\u5728\u6b64\u63d0\u4f9b\u7d66 \u5927\u5bb6\u505a\u53c3\u8003\u3002\u9019 \u300c\u8a72\u300d\u53ea\u51fa\u73fe(9)\u4e00\u4f8b\u3002 (3)\u6c11\u751f\u7269\u8cc7\u5145\u88d5\uff0c\u6c11\u773e\u4e0d\u5fc5\u8981\u6050\u614c\uff0c\u4e5f\u6c92\u5fc5\u8981\u56e4\u7a4d\u3002\u856d\u842c\u9577\u8868\u793a\uff0c\u570b\u5167 \u6cb9\u50f9\u8abf\u6574\u7684 (4)\u4f9d\u7167\u885b\u751f\u7f72\u7684\u8655\u4e8b\u5fc3\u614b\uff0c\u4ed6\u5011\u4f3c\u4e4e\u6c92\u6709\u5fc5\u8981\u4e3b\u52d5\u53bb\u5e6b\u52a9\u6c11\u9593\u767c\u660e\u4eba\uff0c \u53d6\u5f97\u5408\u6cd5\u8b49\u660e (\u4e8c)\u52a9\u52d5\u8a5e\u5426\u5b9a\u5f0f\u8207\u7a0b\u5ea6\u526f\u8a5e (14) \u4e0d\u8ad6\u4ec0\u9ebc\u6d41\u6d3e\uff0c\u90fd\u4e0d\u5fc5\u5206\u5bb6\uff0c\u66f4\u4e0d\u5fc5\u62d8\u675f\u81ea\u5df1\uff0c\u624d\u80fd\u4f7f\u82b1\u85dd\u9054\u5230 \u6700\u9ad8\u5883\u754c\u3002 (15) \u4e0d\u50c5\u53ef\u4f7f\u7528\u6bcd\u570b\u8a9e\u8a00\uff0c\u66f4\u4e0d\u5fc5\u7576\u5834\u4ed8\u8cbb\uff0c\u65e5\u5f8c\u7531\u53d7\u8a71\u4eba\u6309\u4e00\u822c\u8cbb \u7387 d. \u300c\u4e0d\u5fc5\u300d\u5728\u52a9\u52d5\u8a5e\u4e4b\u524d\u6216\u5f8c \u4f4d\u7f6e\uff0c\u53ef\u505a\u70ba\u52a9\u52d5\u8a5e\u7684\u4e00\u500b\u4e3b\u8981\u7279\u5fb5\u3002\u4ee5\u300c\u5fc5\u9808\u300d\u70ba\u4f8b\uff0c\u51fa\u73fe 1113 \u7b46\u7684\u300c\u5fc5\u9808\u300d \u4e4b\u5f8c\u4e0d\u63a5\u52d5\u8a5e\u8005\u53ea\u6709\u4e0b\u5217 6 \u4f8b\uff1a (1)\u5340\u57df\u5716\u66f8\u9928\u6216\u793e\u5340\u5b78\u6821\uff0c\u5247 NREN \u662f\u5fc5\u9808\u7684\uff0c\u4f46\u5047\u5982\u53ea\u662f\u5efa\u7acb\u4e00\u500b\u5f88\u5927 \u7684\u7db2\u8def (2)\u5f71\u97ff\u53ef\u80fd\u53ea\u662f\u5176\u4e2d\u7684\u4e00\u7a2e\uff0c\u4f46\u9810\u9632\u7e3d\u662f\u5fc5\u9808\u7684\uff0c\u4ed6\u5efa\u8b70\u96fb\u529b\u516c\u53f8\u67b6\u8a2d (12)\u6211\u53ea\u662f\u53bb\u770b\u770b\u5979\uff0c\u5c11\u8aaa\u5e7e\u53e5\u4e5f\u53ef\u4ee5\uff0c (13) \u6211\u5f88\u60f3\u544a\u8a34\u65bd\u5de5\u55ae\u4f4d\uff1a\u8981\u6316\u53ef\u4ee5\uff0c\u4f46\u662f\u4e0d\u8981\u96a8\u8208\u4e82\u6316 (14)\u53ea\u8981\u6211\u559c\u6b61\uff0c\u6709\u4ec0\u9ebc\u4e0d\u53ef\u4ee5 (15)\u5149\u9760\u8070\u660e\u662f\u4e0d\u884c\u7684\uff0c\u9084\u8981\u52aa\u529b\u7528\u529f\u624d\u53ef\u4ee5\u3002 \u4ee5\u4e0a\u662f\u300c\u53ef\u4ee5\u300d\u7576\u53e5\u4e3b\u8a9e\u7684\u8b02\u8a9e\u7684\u60c5\u5f62\uff0c\u4e0b\u9762\u56db\u53e5\u5247\u662f\u52a9\u52d5\u8a5e\u5f8c\u52d5\u8a5e\u7d44\u7701\u7565\u7684\u4f8b \u5fc5\u9808 \u253c \u253c \u253c \u80fd -\u253c \u253c \u61c9\u8a72 -\u253c \u253c \u53ef\u4ee5 -\u253c \u253c \u5fc5\u8981 -\u253c \u253c \u6562 -\u253c \u253c \u8a72 2/383 0.52 \u9858\u610f 1/293 \u540c\uff0c\u5e73\u8861\u8a9e\u6599\u5eab\u540c\u6a23\u96a8\u6a5f\u53d6\u6a23\u6642\uff0c\u5404\u8a5e\u505a\u70ba\u60c5\u614b\u8a5e\u7528\u6cd5\u7684\u6bd4\u4f8b\u5dee\u8ddd\u9817\u5927\u3002 \u5c3e)\uff0c\u767c\u73fe(a)\u300c\u53ef\u80fd\u300d\u51fa\u73fe\u5728\u53e5\u9996\u8005\u6709\u4e94\u4f8b\uff0c \u300c\u5fc5\u9808\u300d\u53ea\u6709\u4e00\u4f8b\uff0c\u5176\u4ed6\u52a9\u52d5\u8a5e\u5247 0.34 \u61c9\u8a72 2/873 0.23 \u53ef\u4ee5 3/2000 0.15 \u6703 3/2000 9 0.15 \u61c9 1/1381 0.07 2.5 \u7406\u8ad6\u8207\u8a9e\u6599\u7684\u4e0d\u4e00\u81f4 \u5b78\u8005\u7684\u7406\u8ad6\u6216\u5047\u8a2d\u662f\u8a66\u5716\u95e1\u8ff0\u8aaa\u8a71\u8005\u7684\u8a9e\u8a00\u7684\u672c\u80fd(competence)\uff0c\u800c\u8a9e\u6599\u5247\u53cd \u6620\u51fa\u8a9e\u8a00\u7684\u904b\u7528(performance)\uff0c\u540c\u6642\u53ef\u9a57\u8b49\u5b78\u8005\u7684\u7406\u8ad6\u662f\u5426\u6210\u7acb\u3002\u5f9e\u4ee5\u4e0a\u7684 \u8a0e\u8ad6\u770b\u4f86\uff0c\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe\u6bd4\u6b63\u53cd\u554f\u53e5\u66f4\u61c9\u8996\u70ba\u52a9\u52d5\u8a5e\u7684\u8a9e\u6cd5\u7279\u9ede\u4e4b\u4e00\uff0c\u800c Li &amp; \u7121\uff1b(b)\u5176\u6b21\uff0c\u52a9\u52d5\u8a5e\u51fa\u73fe\u65bc\u53e5\u5c3e\u6642\u6709\u5176\u9650\u5236\u6216\u529f\u80fd\uff1a\u5982\u300c\u53ef\u300d\u5fc5\u9808\u4f34\u96a8\u300c\u5747/ \u300c\u61c9\u8a72\u300d \u3001 \u300c\u61c9\u7576\u300d \u3001 \u300c\u61c9\u300d\u548c\u300c\u8a72\u300d\u5404\u7a2e\u8a5e\u985e\u7684\u6bd4\u4f8b\u5206\u5225\u5982\u4e0b\u6240\u793a(\"X(Y)\"\u4e2d X \u8868\u793a\u51fa\u73fe\u7684\u6b21\u6578\uff0cY \u8868\u793a\u6240\u4f54\u7684\u767e\u5206\u6bd4) \u624d\u300d\u7b49\u526f\u8a5e\u624d\u80fd\u51fa\u73fe\u65bc\u53e5\u5c3e\uff0c \u300c\u53ef\u4ee5\u300d\u5247\u7121\u6b64\u9650\u5236\uff1b\u53e5\u5c3e\u53ef\u89e3\u6c7a\u52a9\u52d5\u8a5e\u7684\u6b67\u7fa9\uff0c 14 \uff1a \u61c9\u7576 \u61c9\u8a72 \u61c9 \u5982\u300c\u61c9\u8a72\u300d\u53ea\u6709\u8868\u60c5\u7406\u4e0a\u5fc5\u9808\u5982\u6b64\u6642(\u5373\u7fa9\u52d9\u7528\u6cd5)\uff0c\u624d\u51fa\u73fe\u65bc\u53e5\u5c3e\uff0c\u63a8\u6e2c\u7fa9\u7684\u300c\u61c9 \u8a72 \u8a72\u300d\u5247\u4e0d\u53ef\u3002\u540c\u6a23\u7684\uff0c\u8868\u80fd\u529b\u7684\u52d5\u8a5e\u300c\u6703\u300d\u53ef\u51fa\u73fe\u65bc\u53e5\u5c3e\uff0c\u8868\u6e2c\u7fa9\u8005\u4e0d\u53ef\u3002\u4ee5\u4e0a</td></tr><tr><td>\u52a9\u52d5\u8a5e\u5178\u578b\u7684\u4f4d\u7f6e\u662f\u5728\u53e5\u4e2d\uff0c\u5f8c\u63a5\u52d5\u8a5e\u3002\u96d6\u7136\u5b78\u8005\u90fd\u80af\u5b9a\u52a9\u52d5\u8a5e\u53ef\u51fa\u73fe\u65bc\u6b63\u53cd\u554f \u500b\u3002Li &amp; Thompson \u5728\u8a3b\u89e3 2 \u6307\u51fa\u6709\u4e9b\u300c\u6703\u300d\u7684\u7528\u6cd5\u4f3c\u4e4e\u9055\u53cd\u7b2c\u4e09\u689d\uff1a \u300c\u4ed6\u5f88\u6703 \u6839\u64da\u5be6\u969b\u8a9e\u6599\u4f86\u6aa2\u9a57\u9019\u4e9b\u52a9\u52d5\u8a5e\u8207\u7a0b\u5ea6\u526f\u8a5e\u300c\u5f88\u300d\u548c\u300c\u66f4\u300d\u7684\u642d\u914d\u60c5\u5f62\u3002\u6aa2\u9a57\u7d50 (6)\u7d93\u5e38\u65b7\u7ae0\u53d6\u7fa9\u50b3\u8ff0\u5225\u4eba\u7684\u8a71\uff0c\u66f4\u6703\u4ee5\u5927\u5e3d\u58d3\u9802\u7684\u65b9\u5f0f\u5047\u50b3\u8056\u65e8\u3002 \u7d9c\u5408\u4e0a\u8ff0\u7684\u4f8b\u5b50\uff0c\u986f\u793a\u300c\u61c9\u8a72/\u61c9/\u8a72\u300d\u53ea\u6709\u7fa9\u52d9\u7528\u6cd5\u53ef\u8207\u7a0b\u5ea6\u526f\u8a5e\u642d\u914d\uff0c\u5982\u4e0b\u8868 \u503c\u5f97\u6ce8\u610f\u7684\u662f\u52a9\u52d5\u8a5e\u7684\u5426\u5b9a\u5f0f\u5982\u300c\u4e0d\u61c9\u8a72\u300d\u53ef\u4ee5\u88ab\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe\uff0c\u9019\u4e00\u9ede\u548c\u5fc3\u7406 \u6e6f&amp;\u6e6f\u5224\u5225\u52a9\u52d5\u8a5e\u7684\u4e00\u500b\u689d\u4ef6\u662f\u300c\u5141\u8a31\u540c\u985e\u6216\u4e0d\u540c\u985e\u7684\u60c5\u614b\u52d5\u8a5e\u6216\u5f62\u5bb9\u8a5e\u9023\u7528\u300d \uff0c \u9ad8\u58d3\u96fb\u7dda \u5b50\uff1a \u6703 -\u253c -\u53ef\u80fd 0 10 0 Tompson \u4e3b\u89c0\u7684\u5c07\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe\u6392\u9664\u5728\u5916\uff0c\u6050\u6015\u4e0d\u80fd\u53cd\u6620\u51fa\u771f\u6b63\u7684\u8a9e\u8a00\u672c\u80fd\u3002 \u52a9\u52d5\u8a5e 31(100) 842(96.45) 1381(94.52) 383(19.15) \u7684\u7d50\u679c\u986f\u793a\uff0c\u8a9e\u8a00\u5b78\u5bb6\u82e5\u5ffd\u7565\u8a9e\u6599\u5448\u73fe\u7684\u8a9e\u8a00\u4e8b\u5be6\uff0c\u5c0d\u8a9e\u6cd5\u7279\u9ede\u7684\u63cf\u8ff0\u6d41\u65bc\u4e3b</td></tr><tr><td>\u53e5\u4e2d\uff0c\u4f46\u6aa2\u9a57\u8a9e\u6599\u7684\u7d50\u679c\u537b\u975e\u5982\u6b64\u3002\u8a9e\u8a00\u5b78\u7406\u8ad6\u6216\u5047\u8a2d\u8a66\u5716\u95e1\u8ff0\u8aaa\u8a71\u8005\u7684\u8a9e\u8a00\u672c \u80fd\uff0c\u4ee5\u5f80\u5b78\u8005\u7814\u7a76\u8a9e\u8a00\u73fe\u8c61\u6642\u591a\u4ee5\u500b\u4eba\u8a9e\u611f\u70ba\u6e96\uff0c\u5c0d\u8a9e\u6cd5\u7279\u9ede\u7684\u63cf\u8ff0\u6d41\u65bc\u4e3b\u89c0\uff0c \u7d93\u7531\u8a9e\u6599\u7684\u9a57\u8b49\u53ef\u4ee5\u53cd\u6620\u8a9e\u8a00\u7684\u5be6\u969b\u4f7f\u7528\u60c5\u5f62\uff0c\u4e26\u6aa2\u9a57\u7406\u8ad6\u7684\u53ef\u884c\u6027\u3002 0. \u524d\u8a00 \u4ee5\u5f80\u5b78\u8005\u8a0e\u8ad6\u52a9\u52d5\u8a5e\u7684\u8a9e\u6cd5\u7279\u9ede\u6642\uff0c\u591a\u6839\u64da\u500b\u4eba\u8a9e\u611f\u6216\u6240\u8490\u96c6\u7684\u6709\u9650\u8a9e\u6599\u52a0\u4ee5\u6b78 \u7d0d\u800c\u6210\uff0c\u7d50\u679c\u51fa\u73fe\u4e0d\u540c\u5b78\u8005\u9593\u63d0\u51fa\u7684\u8a9e\u6cd5\u7279\u9ede\u4e0d\u50c5\u6709\u6240\u51fa\u5165\uff0c\u751a\u81f3\u4e92\u76f8\u77db\u76fe\u3002\u4ee5 \u52a9\u52d5\u8a5e\u70ba\u4f8b\uff0cLi &amp; Thompson(1981)\u8a8d\u70ba\u300c\u61c9\u8a72\u3001\u80fd\u5920\u3001\u6703\u3001\u53ef\u4ee5\u3001\u80fd\u3001\u53ef\u4ee5\u300d \u7b49\u52a9\u52d5\u8a5e\u4e0d\u63a5\u53d7\u7a0b\u5ea6\u526f\u8a5e(\u5982\u300c\u5f88\u300d)\u7684\u4fee\u98fe\uff1b\u6e6f&amp;\u6e6f(1998)\u5247\u5c07\u52a9\u52d5\u8a5e\u6b78\u65bc\u52d5\u8a5e \u6216\u5f62\u5bb9\u8a5e\uff0c\u751a\u81f3\u5c07\u7a0b\u5ea6\u526f\u8a5e\u7684\u4fee\u98fe\u8996\u70ba\u5340\u5206\u60c5\u614b\u52d5\u8a5e\u8207\u5f62\u5bb9\u8a5e\u7684\u4e3b\u8981\u689d\u4ef6\uff0c\u56e0\u6b64 \u4ed6\u5011\u7684\u60c5\u614b\u5f62\u5bb9\u8a5e\u5305\u62ec\u300c\u61c9\u8a72/\u53ef\u80fd/\u53ef\u4ee5/\u80fd(\u5920)/\u9858\u610f/\u80af/\u6562/\u6703(\u8aaa\u8a71)\u300d \uff0c\u81ea\u7136 \u53ef\u63a5\u53d7\u7a0b\u5ea6\u526f\u8a5e(\u5982\u300c\u5f88\u300d)\u7684\u4fee\u98fe\u3002\u672c\u6587\u4e3b\u65e8\u5373\u91dd\u5c0d\u9019\u4e9b\u5b78\u8005\u6240\u63d0\u7684\u8a9e\u6cd5\u7279\u9ede\u4ee5 \u5be6\u969b\u8a9e\u6599\u52a0\u4ee5\u6aa2\u9a57 1 \u3002\u6b64\u8655\u8a9e\u6599\u662f\u6307\u4e2d\u7814\u9662\u8a5e\u5eab\u5c0f\u7d44\u767c\u5c55\u7684\u5e73\u8861\u8a9e\u6599\u5eab\uff0c\u6709\u95dc\u5176 \u8a73\u7d30\u5167\u5bb9\u53ca\u8aaa\u660e\u8acb\u53c3\u8a5e\u5eab\u5c0f\u7d44(1998)\u3002 \u5e95\u4e0b\u5167\u5bb9\u5305\u62ec\u4e09\u500b\u90e8\u4efd\uff1a\u7b2c\u4e00\u7bc0\u7c21\u4ecb Li &amp; Thompson(1981)\u548c\u6e6f&amp;\u6e6f(1998)\u70ba\u52a9 \u52d5\u8a5e (\u6216\u7a31\u60c5\u614b\u52d5\u8a5e/\u5f62\u5bb9\u8a5e)\u6240\u5217\u7684\u8a9e\u6cd5\u7279\u9ede\uff0c\u7b2c\u4e8c\u7bc0\u5373\u6839\u64da\u5be6\u969b\u8a9e\u6599\u4f86\u6aa2\u9a57\u9019 \u4e9b\u52a9\u52d5\u8a5e\u8207\u7a0b\u5ea6\u526f\u8a5e\u7684\u95dc\u4fc2\u53ca\u5176\u4ed6\u76f8\u95dc\u7684\u8a9e\u6cd5\u7279\u9ede\uff0c\u5305\u62ec\u52a9\u52d5\u8a5e\u8207\u7a0b\u5ea6\u526f\u8a5e\u3001\u52a9 \u52d5\u8a5e\u7684\u5426\u5b9a\u5f0f\u3001\u52a9\u52d5\u8a5e\u7684\u4f4d\u7f6e\u53ca\u6b63\u53cd\u554f\u53e5\u7b49\u3002\u7b2c\u4e09\u7bc0\u505a\u4e00\u7e3d\u7d50\u3002 1. \u52a9\u52d5\u8a5e\u7684\u8a9e\u6cd5\u7279\u9ede Li &amp; Thompson (1981)\u3001\u6731 (1982)\u3001\u6e6f(1984)\u3001\u6e6f&amp;\u6e6f(1998)\u8a0e\u8ad6\u52a9\u52d5\u8a5e\u6642\u6240\u63cf \u8ff0\u7684\u7684\u8a9e\u6cd5\u7279\u9ede\u4e0d\u76e1\u76f8\u540c\uff0c\u4e0b\u9762\u4ee5 Li &amp; Thompson (1981)\u548c\u6e6f&amp;\u6e6f(1998)\u70ba\u4f8b\u52a0 \u4ee5\u8aaa\u660e\u3002 Li &amp; Thompson \u6240\u5217\u7684\u8a9e\u6cd5\u7279\u9ede\uff1a 1) \u53ef\u4ee5\u51fa\u73fe\u65bc\u6b63\u53cd\u554f\u53e5 2)\u53ef\u52a0\u4ee5\u5426\u5b9a 3) \u4e0d\u80fd\u88ab\u300c\u5f88\u300d\u6216\u300c\u66f4\u300d\u7b49\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe \u8aaa\u8a71\u300d \u3002\u4f5c\u8005\u8a8d\u70ba\u9019\u53e5\u4e26\u975e\u53cd\u4f8b\uff0c\u56e0\u300c\u6703\u8aaa\u8a71\u300d\u662f\u6210\u8a9e\uff0c\u4e00\u822c\u800c\u8a00\u300c\u6703\u300d\u4ecd\u4e0d\u80fd \u7528\u300c\u5f88\u300d\u6216\u300c\u66f4\u300d\u4f86\u4fee\u98fe\uff1a (1)*\u4ed6\u5f88/\u66f4\u6703\u6e38\u6cf3 \u6e6f&amp;\u6e6f(1998)\u8a8d\u70ba\u60c5\u614b\u52d5\u8a5e/\u5f62\u5bb9\u8a5e\u7684\u53e5\u6cd5\u7279\u5fb5\u5982\u4e0b\u6240\u793a\uff1a 1) \u53ef\u4ee5\u51fa\u73fe\u65bc\u6b63\u53cd\u554f\u53e5 2) \u53ef\u51fa\u73fe\u65bc\u5426\u5b9a\u526f\u8a5e\u4e4b\u524d\u5f8c 3) \u60c5\u614b\u5f62\u5bb9\u8a5e\u53ef\u88ab\u7a0b\u5ea6\u526f\u8a5e(\u5982\u300c\u5f88\u300d)\u4fee\u98fe 4) \u53ef\u51fa\u73fe\u65bc\u53e5\u4e2d\u751a\u6216\u53e5\u9996\u7684\u4f4d\u7f6e\uff0c\u5982\u300c\u53ef\u80fd/\u61c9\u8a72\u300d 5) \u53ef\u4ee5\u55ae\u7368\u56de\u7b54 6) \u53ef\u5145\u7576\u5206\u88c2\u53e5\u7684\u4fe1\u606f\u7126\u9ede(\u5982\u300c\u4ed6\u662f\u53ef\u80fd\u4e2d\u734e\u7684\u300d) 7) \u5141\u8a31\u540c\u985e\u6216\u4e0d\u540c\u985e\u7684\u60c5\u614b\u52d5\u8a5e\u6216\u5f62\u5bb9\u8a5e\u9023\u7528 \u7531\u4e0a\u8ff0\u6bd4\u8f03\u53ef\u4ee5\u770b\u51fa\uff0cLi &amp; Thompson (1981)\u6240\u5217\u7684\u516b\u9ede\u548c\u6e6f&amp;\u6e6f(1998)\u7684\u4e03\u9ede \u4e2d\uff0c\u6709\u91cd\u758a\u7684\u90e8\u4efd\u53ea\u6709\u7b2c 1-4 \u9ede\uff0c\u5176\u4e2d\u53ea\u6709\u7b2c 1 \u9ede\u5169\u6587\u662f\u4e00\u81f4\u7684\uff0c\u7b2c 2 \u548c 4 \u9ede\u7a0d \u6709\u4e0d\u540c\uff0c\u800c\u5169\u8005\u5c0d\u7b2c\u4e09\u9ede\u7684\u770b\u6cd5\u4e92\u76f8\u77db\u76fe\u3002\u4e8b\u5be6\u4e0a\u7b2c\u4e09\u9ede\u727d\u6d89\u5230\u52a9\u52d5\u8a5e\u7684\u6b78\u985e\u554f \u984c\uff0c\u76ee\u524d\u5c31\u52a9\u52d5\u8a5e\u5728\u6574\u500b\u6f22\u8a9e\u8a5e\u985e\u7cfb\u7d71\u4e2d\u6240\u5360\u7684\u4f4d\u7f6e\uff0c\u5404\u5bb6\u8aaa\u6cd5\u53ef\u5206\u70ba\u4e09\u5927\u985e\uff1a \u4e00\u3001\u52d5\u8a5e\u8aaa\uff0c\u4e8c\u3001\u526f\u8a5e\u8aaa\uff0c\u4e09\u3001\u7368\u7acb\u4e00\u985e\u3002Li &amp; Thompson \u63a1\u7b2c\u4e09\u7a2e\u8aaa\u6cd5\uff0c\u4e14\u8a8d \u70ba\u300c\u61c9\u8a72\u3001\u80fd\u5920\u3001\u6703\u3001\u53ef\u4ee5\u3001\u80fd\u3001\u53ef\u4ee5\u300d\u7b49\u52a9\u52d5\u8a5e\u4e0d\u63a5\u53d7\u7a0b\u5ea6\u526f\u8a5e(\u5982\u300c\u5f88\u300d) \u7684\u4fee\u98fe\uff1b\u6e6f&amp;\u6e6f\u5247\u5c07\u52a9\u52d5\u8a5e\u6b78\u65bc\u52d5\u8a5e\u6216\u5f62\u5bb9\u8a5e\uff0c\u751a\u81f3\u5c07\u7a0b\u5ea6\u526f\u8a5e\u7684\u4fee\u98fe\u8996\u70ba\u5340\u5206 \u60c5\u614b\u52d5\u8a5e\u8207\u5f62\u5bb9\u8a5e\u7684\u4e3b\u8981\u689d\u4ef6\uff0c\u5982\u60c5\u614b\u5f62\u5bb9\u8a5e\u5305\u62ec\u300c\u61c9\u8a72/\u53ef\u80fd/\u53ef\u4ee5/\u80fd(\u5920)/ \u9858\u610f/\u80af/\u6562/\u6703(\u8aaa\u8a71)\u300d \uff1b\u63db\u8a00\u4e4b\uff0c \u300c\u61c9\u8a72\u3001\u80fd\u5920\u3001\u6703\u3001\u53ef\u4ee5\u3001\u80fd\u3001\u53ef\u4ee5\u300d\u7b49\u52a9\u52d5 \u8a5e\u662f\u5f62\u5bb9\u8a5e\u7684\u4e00\u7a2e\uff0c\u81ea\u7136\u53ef\u63a5\u53d7\u7a0b\u5ea6\u526f\u8a5e(\u5982\u300c\u5f88\u300d)\u7684\u4fee\u98fe\u3002\u4e0b\u9762\u5c07\u91dd\u5c0d Li &amp; Thompson(1981)\u53ca\u6e6f&amp;\u6e6f(1997)\u6240\u5217\u7684\u52a9\u52d5\u8a5e\u8a9e\u6cd5\u7279\u9ede\u4e2d\u91cd\u758a\u7684\u90e8\u4efd(\u5373\u524d\u56db \u9805)\uff0c\u4ee5\u5be6\u969b\u8a9e\u6599\u4f86\u6aa2\u9a57\u9019\u5169\u7bc7\u8ad6\u6587\u6240\u5217\u7684 17 \u500b\u52a9\u52d5\u8a5e\u3002 2. \u52a9\u52d5\u8a5e\u8a9e\u6cd5\u7279\u9ede\u7684\u6aa2\u9a57 \u7d9c\u5408\u4e0a\u8ff0\u5169\u5bb6\u8a0e\u8ad6\u7684\u52a9\u52d5\u8a5e\u8a9e\u6cd5\u7279\u9ede\uff0c\u672c\u7bc0\u5206\u70ba\u52a9\u52d5\u8a5e\u8207\u7a0b\u5ea6\u526f\u8a5e\u3001\u52a9\u52d5\u8a5e\u7684\u5426 \u6bd4\u8f03\"\uff0c\u7de8\u865f NSC 88-2411-H-126-005)\u7684\u90e8\u4efd\u5167\u5bb9\u4fee\u6539\u800c\u6210\u3002 2 \u5728\u6536\u5230\u7684\u96fb\u5b50\u4fe1\u4ef6\u4e2d\u767c\u73fe\u6709\u300c\u5982\u679c\u661f\u671f\u516d\u4e0d\u80fd\u8b80\u66f8\u6703\u300d(\u80fd+\u540d\u8a5e\u7d44)\u7684\u4f8b\u5916\uff0c\u9019\u6216\u8a31\u662f\u7701\u7565(\u4e0d \u80fd\u8209\u884c\u8b80\u66f8\u6703)\u6216\u6f0f\u5b57\u7684\u7d50\u679c\u3002 \u679c\u5982\u4e0b(X/Y(Z%)\u5206\u5225\u8868\u793a\u51fa\u73fe\u6b21\u6578/\u7e3d\u6b21\u6578(\u983b\u7387)\uff0c\u4ee5\u4e0b\u4f9d\u7167\u300c\u5f88+AUX\u300d\u51fa\u73fe\u7684 \u983b\u7387\u6392\u5217)\uff1a \u5f88+AUX \u66f4+AUX \u9858\u610f 9/293(3.07%) 2/293(0.68%) \u53ef\u80fd 58/2000(2.90%) 1/2000(0.05%) \u6703 8/2000(0.40%) 9/2000(0.45%) \u80fd 5/2000(0.25%) 24/2000(1.20%) \u80fd\u5920 1/463(0.22%) 1/463(0.22%) \u61c9\u7576 0 1/31(3.23%) \u61c9\u8a72 0 10/873(1.15%) \u61c9 0 12/1381(0.9%) \u53ef\u4ee5 0 10/2000(0.5%) \u6562 0 1/330(0.3%) \u5f97 dei 3 0 1/222(0.45%) \u5fc5\u9808 0 1/1113(0.09%) \u5f97 de 0 0 \u8a72 0 0 \u80af 0 0 \u5fc5\u5f97 0 0 \u5fc5\u8981 0 0 \u5c31\u4e0a\u8868\u5341\u4e03\u500b\u53ef\u80fd\u7684\u52a9\u52d5\u8a5e\u4f86\u770b\uff0c\u80fd\u642d\u914d\u7a0b\u5ea6\u526f\u8a5e\u7684\u52a9\u52d5\u8a5e\u4f54\u591a\u6578\uff0c\u4f46\u642d\u914d\u300c\u5f88\u300d \u7684\u52a9\u52d5\u8a5e\u4f4e\u65bc\u300c\u66f4\u300d \u3002\u6aa2\u9a57\u8a9e\u6599\u7684\u904e\u7a0b\u4e2d\uff0c\u6211\u5011\u767c\u73fe\u4e0a\u8ff0\u7d50\u679c\u9664\u4e86\u8aaa\u660e\u52a9\u52d5\u8a5e\u5be6 \u969b\u4e0a\u53ef\u548c\u7a0b\u5ea6\u526f\u8a5e\u642d\u914d\uff0c\u66f4\u91cd\u8981\u7684\u662f\uff0c\u52a9\u52d5\u8a5e\u642d\u914d\u7a0b\u5ea6\u526f\u8a5e\u6642\uff0c\u53ef\u80fd\u6709\u9650\u5236\uff0c\u5982 \u300c\u6703\u300d\u53ef\u8868\u80fd\u529b\u6216\u9810\u65b7\uff0c\u8a9e\u6599\u4e2d\u524d\u8005\u4e3b\u8981\u642d\u914d\u300c\u5f88\u300d\u800c\u5f8c\u8005\u53ea\u8207\u300c\u66f4\u300d\u5171\u73fe\uff1b \u300c\u61c9 \u8a72\u300d\u5247\u517c\u6709\u7fa9\u52d9\u548c\u9810\u65b7\u7528\u6cd5\uff0c\u4f46\u53ea\u898b\u7fa9\u52d9\u7528\u6cd5\u548c\u7a0b\u5ea6\u526f\u8a5e\u4e00\u8d77\u51fa\u73fe\u3002\u4e0b\u9762\u91dd\u5c0d\u9019 \u5169\u9ede\u5206\u5225\u8209\u4f8b\u8aaa\u660e\u4e4b\u3002 2.1.1\u300c\u5f88+\u6703\u300d/\u300c\u66f4+\u6703\u300d 3 \u300c\u5f97\u300d\u6709\u5169\u8b80 dei3 \u548c de2\uff0c\u524d\u8005\u8868\u662f\u5fc5\u9808\u7fa9\uff0c\u5f8c\u8005\u5141\u8a31\u7fa9\uff0c\u7686\u5c6c\u7fa9\u52d9\u985e\uff0c\u4f46\u8a9e\u6599\u986f\u793a\u5169\u8005\u8a9e\u6cd5 \u7279\u9ede\u4e0d\u76e1\u76f8\u540c\uff0c\u6545\u52a0\u4ee5\u5206\u5225\u3002 (7)\u6c92\u8fa6\u6cd5\u6642\uff0c\u90a3\u5de5\u5546\u754c\u4e0d\u53ea\u8981\u8d70\u4e0a\u8857\u982d\uff0c\u66f4\u6703\u51fa\u8d70\u6d77\u5916\u3002 \u8a9e\u6599\u4e2d\u300c\u66f4+\u6703\u300d\u51fa\u73fe 9 \u6b21\uff0c\u9664\u4e86\u4f8b(4)\u662f\u80fd\u529b\u7fa9\u5916\uff0c\u5176\u4ed6 8 \u4f8b\u70ba\u63a8\u6e2c\u7fa9\uff0c\u5982\u4e0b\u8868 \u6240\u793a\uff1a \u5f88+\u6703 \u66f4+\u6703 \u80fd\u529b\u7fa9 8 1 \u9810\u65b7\u7fa9 0 8 \u5f9e\u6b64\u8868\u770b\u4f86\uff0c \u300c\u5f88\u300d\u548c\u300c\u66f4\u300d\u96d6\u7136\u90fd\u662f\u7a0b\u5ea6\u526f\u8a5e\uff0c\u537b\u642d\u914d\u300c\u6703\u300d\u7684\u4e0d\u540c\u7528\u6cd5\u3002 2.1.2\u300c\u61c9\u8a72\u300d \u3001 \u300c\u61c9\u300d\u548c\u300c\u8a72\u300d\u7684\u6bd4\u8f03 \u52a9\u52d5\u8a5e\u591a\u534a\u6709\u517c\u7fa9\u73fe\u8c61\uff0c\u5373\u4e00\u500b\u8a5e\u517c\u6709\u5169\u7a2e\u4ee5\u4e0a\u7684\u60c5\u614b\u7528\u6cd5\uff0c\u4ee5\u300c\u61c9\u8a72\u300d\u70ba\u4f8b\u3002 \u300c\u61c9\u8a72\u300d\u6709\u5169\u7a2e\u7528\u6cd5(\u5442 1984)\uff1a a. \u8868\u793a\u60c5\u7406\u4e0a\u5fc5\u9808\u5982\u6b64\u3002 \u4f8b\uff1a\u5b78\u7fd2\u61c9\u8a72\u8a8d\u771f b. \u4f30\u8a08\u60c5\u6cc1\u5fc5\u7136\u5982\u6b64 \u4f8b\uff1a\u4ed6\u6628\u5929\u52d5\u8eab\u7684\uff0c\u4eca\u5929\u61c9\u8a72\u5230\u4e86 \u300c\u61c9\u8a72 1\u300d\u662f\u7fa9\u52d9\u985e\uff0c \u300c\u61c9\u8a72 2\u300d\u662f\u8a8d\u77e5\u985e\u3002\u4f46\u662f\u6574\u7406\u8a9e\u6599\u7684\u904e\u7a0b\u4e2d\uff0c\u767c\u73fe\u53ea\u6709\u524d \u8005\u53ef\u88ab\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe\u3002\u8209\u4f8b\u5982\u4e0b\uff1a a.\u66f4+\u61c9\u8a72 (1) \u4e26\u52c7\u65bc\u5275\u9020\u5c31\u696d\u7684\u65b0\u5712\u5730\uff0c\u66f4\u61c9\u8a72\u9ad4\u6703\u81ea\u8eab\u5275\u9020\u4e8b\u696d\u6240\u9700\u4ed8\u51fa\u7684\u4ee3 \u50f9\u3002 (2) \u56e0\u70ba\u6211\u5011\u662f\u904e\u4f86\u4eba\uff0c\u66f4\u61c9\u8a72\u611f\u540c\u8eab\u53d7\uff0c\u4e0d\u8981\u7528\u4e2d\u5e74\u7684\u773c\u5149\u4f86\u770b (3) \u300c\u53f0\u7063\u300d\u4e0d\u4f46\u662f\u4e00\u500b\u7814\u7a76\u5ba2\u9ad4\uff0c\u66f4\u61c9\u8a72\u662f\u7814\u7a76\u7684\u4e3b\u9ad4 b.\u6700+\u61c9\u8a72 (4) \u6700\u61c9\u8a72\u51fa\u8d70\u4e26\u6c23\u61a4\u53f0\u7063\u7684\u6797\u7fa9\u96c4\uff0c\u5373\u5c07\u56de\u4f86 (5) \u6700\u61c9\u8a72\u6210\u70ba\u53f0\u7063\u6700\u503c\u5f97\u77bb\u4ef0\u7684\u6797\u5bb6\u77f3\u7891 \u5c31\u8a9e\u6599\u986f\u793a\uff0c\u9019\u4e9b\u517c\u7fa9\u8a5e\u88ab\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe\u6642\uff0c\u51fa\u73fe\u7684\u4f4d\u7f6e\u6709\u5176\u9650\u5236\uff1a \u300c\u66f4/\u6700/* \u5f88\u61c9\u8a72+VP\u300d \uff0c\u5426\u5b9a\u5f0f\u4e5f\u53ef\u4ee5\u88ab\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe\uff0c\u4f46\u51fa\u73fe\u5728\u4e3b\u8981\u8b02\u8a9e\u7684\u4f4d\u7f6e\uff1a \u300c\u5b50\u53e5/ \u540d\u8a5e\u7d44+\u66f4/\u5f88/\u771f/\u592a\u4e0d\u61c9\u8a72\u300d \uff0c\u8acb\u53c3 2.2 \u7bc0\u3002 \u300c\u61c9/\u8a72\u300d\u4e00\u822c\u4e5f\u8a8d\u70ba\u662f\u300c\u61c9\u8a72\u300d\u7684\u8fd1\u7fa9\u8a5e\u6216\u540c\u7fa9\u8a5e\uff0c\u8a9e\u6599\u6240\u898b\u537b\u662f\u4ee5\u7fa9\u52d9\u7528\u6cd5 \u5a5a\u59fb\u4e2d\u751f\u6d3b\u5708\u672c\u5c31\u8f03\u7a84\u300d)\u3002 \u5b50\u7684\u7b2c\u4e8c\u500b\u300c\u80fd\u300d\u5b57\uff1a \u300c\u5c0d\u81ea\u5df1\u6709\u4e86\u4e86\u89e3\u4e4b\u5f8c\uff0c\u7576\u7136\u5e0c\u671b\u5225\u4eba\u80fd\u4e86\u89e3\u6211\uff0c\u4f46\u5225\u4eba\u80fd\u4e0d\u80fd\u4e86\u89e3\u5462\uff1f\u300d \u3002 13 \u8a73\u7d30\u8a0e\u8ad6\u8acb\u53c3\u912d(2000)\u3002 16 \u5305\u62ec\u9023\u63a5\u8a5e\u3002 \u70ba\u4e3b\uff0c\u6240\u4ee5\u4e5f\u53ea\u898b\u7fa9\u52d9\u7528\u6cd5\u7684\u300c\u61c9/\u8a72\u300d\u88ab\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe\uff1a \u6240\u793a\uff1a \u8a8d\u77e5 \u7fa9\u52d9 \u5f88+\u61c9\u8a72/\u61c9 --\u66f4+\u61c9\u8a72/\u61c9 -\u253c \u6700+\u61c9\u8a72/\u61c9 -\u253c 2.2 \u52a9\u52d5\u8a5e\u7684\u5426\u5b9a\u5f0f 17 \u500b\u53ef\u80fd\u7684\u52a9\u52d5\u8a5e\u4e2d\uff0c\u53ef\u76f4\u63a5\u4ee5\u300c\u4e0d\u300d\u5426\u5b9a\u8005\u5982\u4e0b\u6240\u5217\uff1a \u4e0d+AUX \u9858\u610f \u253c \u53ef\u80fd \u253c \u6703 \u253c \u80fd \u253c \u61c9\u7576 \u253c \u61c9\u8a72 \u253c \u61c9 \u253c \u53ef\u4ee5 \u253c \u6562 \u253c \u8a72 \u253c \u80af \u253c \u5f97 de \u253c \u5fc5\u8981 \u253c \u5fc5\u9808 \uff1f \u80fd\u5920 -\u5f97 dei -\u5fc5\u5f97 -\u4e0a\u8868\u4e2d\u6709\u4e09\u9ede\u9700\u8981\u88dc\u5145\u3002 (\u4e00)\u9664\u4e86\u5426\u5b9a\u8a5e\u300c\u4e0d\u300d\u4e4b\u5916\uff0c\u9084\u6709\u5c11\u6578\u53ef\u88ab\u300c\u6c92(\u6709)\u300d\u5426\u5b9a\uff0c\u5c31\u8a9e\u6599\u6240\u898b\u6709\u300c\u53ef \u80fd/\u80fd/\u5fc5\u8981 4 \u300d\u4e09\u8a5e\uff1a (1)\u5c31\u6c92\u6709\u6a5f\u6703\u4e0a\u96fb\u8996\u3002\u4e0a\u5831\uff0c\u5c31\u6c92\u6709\u53ef\u80fd\u53d7\u5230\u91cd\u8996\uff0c\u4e5f\u5c31\u6c92\u6709\u8cc7\u683c\u5b58\u6d3b \u52d5\u8a5e (\u5982\u300c\u5f88\u4e0d\u559c\u6b61\u300d)\u6216\u5f62\u5bb9\u8a5e(\u5982\u300c\u5f88\u4e0d\u9ad8\u8208\u300d)\u76f8\u540c\u3002\u4ee5\u300c\u4e0d\u61c9\u8a72\u300d\u70ba\u4f8b\uff0c \u5c31\u8a9e\u610f\u800c\u8a00\uff0c \u300c\u4e0d\u61c9\u8a72\u300d\u88ab\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe\u6642\u90fd\u662f\u8868\u7fa9\u52d9\u7684\u7528\u6cd5\uff1b\u5c31\u53e5\u6cd5\u800c\u8a00\uff0c\u6b64 \u6642\u300c\u5f88/\u66f4/\u592a\u4e0d\u61c9\u8a72\u300d\u505a\u70ba\u4e3b\u8981\u8b02\u8a9e\uff0c\u4ee5\u5b50\u53e5\u70ba\u4e3b\u8a9e\u3002\u5982\u4ee5\u4e0b\u4f8b\u5b50\u6240\u793a\uff1a (5) \u4ed6\u4e00\u60f3\uff0c\u9019\u500b\u4e8b\u60c5\u5f88\u4e0d\u61c9\u8a72\u3002\u6240\u4ee5\u4ed6\u5c31\u5f9e\u9019\u500b\u597d\u6f22\u7d44\u9000\u51fa\u4f86 (6) \u4e3b\u79d8\u5230\u7e23\u9577\u4e00\u5c64\u6c7a\u884c\uff0c\u5f71\u5370\u516c\u958b\uff0c\u66f4\u4e0d\u61c9\u8a72 (7) \u7e3d\u7d71\u8523\u516c\u807d\u4e86\uff0c\u8a8d\u70ba\u9019\u500b\u65e5\u672c\u6559\u5b98\u592a\u4e0d\u61c9\u8a72\uff0c\u600e\u9ebc\u53ef\u4ee5\u4e0d\u5c0a\u91cd\u5225\u4eba\u7684 \u570b\u5bb6\u3002 (8) \u65e2\u5f97\u5229\u76ca\u5c31\u8f49\u79fb\u5230\u7701\u7c4d\u554f\u984c\u5be6\u5728\u592a\u4e0d\u61c9\u8a72\u3002\u9673\u6c34\u6241\u5247\u4ee5\u903c\u9000\u4e0d\u662f\u672c\u7701 \u4eba\u6392\u65a5 (9) \u5011\u771f\u662f\u592a\u53ef\u60e1\u4e86\uff01\u6211\u89ba\u5f97\u5c0f\u670b\u53cb\u592a\u4e0d\u61c9\u8a72\u4e86\uff0c\u56de\u5230\u64cd\u5834\u4e0a\uff0c\u53c8\u770b\u5230\u5c0f \u670b\u53cb\u4e0d (\u4e09)\u300c\u5fc5\u9808\u300d\u7684\u5426\u5b9a \u300c\u5fc5\u9808\u300d\u96d6\u4e0d\u80fd\u76f4\u63a5\u52a0\u4ee5\u5426\u5b9a(\u300c\u4e0d\u5fc5\u9808\u300d)\uff0c\u4f46\u300c\u4e0d\u5fc5\u300d\u5728\u8a9e\u610f\u4e0a\u76f8\u7576\u65bc\u300c\u5fc5\u9808\u300d \u7684\u53cd\u7fa9\u8a5e\uff0c\u6216\u53ef\u8996\u70ba\u52a9\u52d5\u8a5e\u300c\u5fc5\u9808\u300d\u7684\u5426\u5b9a\u5f0f\u3002\u96d6\u7136\u5f35(1961)\u548c\u5442(1984)\u90fd\u5c07\u300c\u4e0d \u5fc5\u300d\u8996\u70ba\u526f\u8a5e\uff0c\u7136\u800c\u6839\u64da\u4e0b\u9762\u300c\u4e0d\u5fc5\u300d\u7684\u56db\u500b\u53e5\u6cd5\u7279\u9ede\uff0c\u53ef\u6b78\u5165\u52a9\u52d5\u8a5e\u4e00\u985e\u3002 a. \u300c\u4e0d\u5fc5\u300d\u53ef\u51fa\u73fe\u65bc\u53e5\u9996\u3001\u53e5\u4e2d\u53ca\u53e5\u5c3e (7)\u300c\u2026\u4eca\u5929\u7684\u60c5\u5f62\u6211\u5b8c\u5168\u77e5\u9053\uff0c\u4e0d\u5fc5\u4f60\u8ddf\u6211\u8b1b\u3002\u300d (8)\u4ed6\u56db\u4e0b\u770b\u4e86\u4e00\u770b\uff0c\u8aaa\uff1a \u300c\u4f60\u7e3d\u4e0d\u5fc5\u4f8d\u5019\u90a3\u4e9b\u684c\u5b50\u6905\u5b50\u5427\uff1f\u300d (9) \u662f\u5c0f\u6642\u5019\u6551\u4eba\u624d\u6703\u6253\u7834\u6c34\u7f38\uff0c\u9577\u5927\u5f8c\u5c31\u4e0d\u5fc5\uff0c\u96d9\u65b9\u7684\u8faf\u89e3\u8a9e\u53e5\uff0c\u5404 \u6709\u5176\u5de7\u5999\u4e4b\u8655\u3002 b.\u4ee5\u4e0b\u53e5\u5b50\u300c\u4e0d\u5fc5\u300d\u662f\u4e3b\u8981\u8b02\u8a9e\u6216\u662f\u55ae\u7368\u56de\u7b54 (10) \u7ad9\u8d77\u4f86\uff0c\u8a9e\u6c23\u5f88\u6eab\u548c\u5730\u8aaa\uff1a\u5e78\u800c\uff0c\u4f60\u4e0d\u5fc5\u3002\u7136\u5f8c\u4f38\u624b\u53bb\u62ff\u5c24\u8774\u7528 \u904e\u800c\u6454\u5728\u6d17\u81c9\u53f0 (11) \u6211\u8981\u63d0\u4ed6\u7684\u5584\u884c\u55ce\uff1f\u8b1d\u8b1d\uff0c\u5be9\u5224\u9577\u8aaa\uff1a\u4e0d\u5fc5\u3002\u88ab\u544a\u8981\u4e0d\u8981\u63d0\u51fa\u7b54 \u8faf\uff1f\u672c\u5ead\u73fe\u5728\u8981 (12) \u6771\u5c3c\u9918\u6c23\u672a\u6d88\uff0c\u61a4\u61a4\u5730\u8aaa\uff1a \u300c\u4e0d\u5fc5\uff01\u6211\u6253\u904e\u96fb\u8a71\u4e86\uff0c\u65c5\u904b\u516c\u53f8\u7b54\u61c9 \u8a2d\u6cd5\uff0c c.\u8868\u6bd4\u8f03\u7684\u7a0b\u5ea6\u526f\u8a5e\u300c\u66f4\u300d\u4fee\u98fe\u300c\u4e0d\u5fc5\u300d \u96d6\u7136\u8a9e\u6599\u4e2d\u6c92\u6709\u51fa\u73fe\u300c\u4e0d\u5fc5\u300d\u88ab\u4e00\u822c\u7684\u7a0b\u5ea6\u526f\u8a5e\u300c\u5f88\u300d\u6216\u300c\u975e\u5e38\u300d\u7684\u4fee\u98fe\uff0c\u4f46\u6709 \u8a9e\u6599\u986f\u793a\u300c\u4e0d\u5fc5\u300d\u53ef\u5728\u52a9\u52d5\u8a5e\u4e4b\u524d\u6216\u5f8c\uff1a (16) \u5982\u679c\u6211\u5011\u80fd\u5920\u8b19\u5351\u4e00\u9ede\uff0c\u6211\u5011\u5c31\u4e0d\u5fc5\u4e00\u5b9a\u8981\u8aaa\u6211\u5011\u90fd\u6210\u4e86\u4e0a\u5e1d (17) \u9019\u4e09\u500b\u5973\u7684\u662f\u4e0d\u662f\u8981\u4ecb\u7d39\u4e00\u4e0b\uff1f\u61c9\u8a72\u4e0d\u5fc5\u4e86\u3002 (18) \u5c3c\u5967\u4e5f\u8a8d\u70ba\u8a0e\u8ad6\u4e8b\u9805\u53ef\u4ee5\u4e0d\u5fc5\u53c3\u52a0\uff0c\u5979\u4fbf\u53c8\u53bb\u7761\u4e86\u3002 (19) \u5e02\u6c11\u6c92\u6709\u8fa6\u6cd5\u5728\u751f\u6d3b\u5468\u570d\u96a8\u6642\u627e\u5230\u53ef\u4ee5\u4e0d\u5fc5\u82b1\u9322\uff0c\u6216\u4e0d\u5fc5\u82b1\u592a\u591a \u9322\u5c31\u53ef\u4ee5\u4eab\u53d7\u7684 (20) \u5c0d\u5df2\u5a5a\u7684\u4eba\u4f86\u8aaa\uff0c\u55ae\u8eab\u8cb4\u65cf\u4f3c\u4e4e\u4e0d\u5fc5\u64d4\u5fc3\u67f4\u7c73\u6cb9\u9e7d\uff0c\u4e0d\u5fc5\u6ce8\u610f\u5b69 \u5b50\u6559\u990a \u5f9e\u4ee5\u4e0a\u7684\u8a0e\u8ad6\u770b\u4f86\uff0c \u300c\u4e0d\u5fc5\u300d\u5177\u6709\u591a\u9805\u52a9\u52d5\u8a5e\u7684\u7279\u5fb5\uff0c\u96d6\u975e\u5178\u578b\u7684\u52a9\u52d5\u8a5e\uff0c\u5927\u81f4 \u53ef\u8996\u70ba\u300c\u5fc5\u9808\u300d\u7684\u5426\u5b9a\u5f62\u5f0f\u3002 2.3 \u52a9\u52d5\u8a5e\u51fa\u73fe\u7684\u4f4d\u7f6e Li &amp; Thompson (1981) \u8a8d\u70ba\u52a9\u52d5\u8a5e\u4e0d\u80fd\u5728\u4e3b\u8a9e\u4e4b\u524d\uff0c\u6e6f&amp;\u6e6f(1998)\u5247\u4e3b\u5f35\u52a9\u52d5\u8a5e \u53ef\u51fa\u73fe\u65bc\u53e5\u4e2d\u751a\u6216\u53e5\u9996\u7684\u4f4d\u7f6e\uff0c\u8209\u51fa\u7684\u4f8b\u5b50\u70ba\u300c\u53ef\u80fd/\u61c9\u8a72\u300d \uff0c\u672c\u7bc0\u5168\u9762\u6aa2\u8a0e\u52a9\u52d5 \u8a5e\u53ef\u51fa\u73fe\u7684\u4f4d\u7f6e(\u53e5\u9996\u3001\u53e5\u4e2d\u6216\u53e5\u5c3e)\uff0c\u5e95\u4e0b\u5206\u4e09\u5c0f\u7bc0\u52a0\u4ee5\u8a0e\u8ad6\u3002 2.3.1 \u51fa\u73fe\u65bc\u53e5\u9996\u7684\u4f4d\u7f6e \u7d93\u6aa2\u9a57\u8a9e\u6599\uff0c \u300c\u61c9\u8a72\u300d\u4e26\u7121\u51fa\u73fe\u65bc\u53e5\u9996\u7684\u4f8b\u5b50\u3002\u800c\u300c\u53ef\u80fd\u300d\u5728\u4e3b\u8a9e\u4e4b\u524d\u8005\u6709 5 \u4f8b\uff1a (1)\u7684\u6749\u539f\u8f1d\u96c4\uff0c\u5c0d\u65bc\u4ed6\u7684\u602a\u7570\u63ee\u687f\u59ff\u52e2\uff0c\u53ef\u80fd\u4e00\u822c\u4eba\u4e0d\u6562\u82df\u540c (2)\u82e5\u63db\u6210\u4e00\u822c\u7684\u3001\u6b63\u5e38\u7684\u884c\u6587\uff0c\u53ef\u80fd\u6b63\u78ba\u7387\u6703\u9ad8\u4e9b\u3002\u6578\u5b57\u6703\u8aaa\u8a71\u3002 (3)\u82e5\u751f\u5728\u592a\u5e73\u6cbb\u4e16\uff0c\u53ef\u80fd\u6709\u4e9b\u4eba\u6d3b\u5728\u9109\u4e0b\u5c0f\u5730\u65b9\uff0c\u4e00\u8f29\u5b50\u4e5f\u6c92\u60f3 (4)\u6211\u53ea\u770b\u5230\u90a3\u4f4d\u8ce3\u51b0\u8001\u592a\u592a\u7684\u8868\u9762\uff0c\u53ef\u80fd\u5979\u7684\u5bb6\u5ead\u662f\u7236\u6148\u5b50\u5b5d\u3001\u5144\u53cb\u5f1f \u606d (5)\u6211\u4e5f\u4e0d\u77e5\u9053\u600e\u9ebc\u56de\u7b54\uff0c\u9019\u4ec0\u9ebc\u539f\u56e0\uff0c\u53ef\u80fd\u524d\u4e00\u8f29\u6211\u662f\u4e2d\u570b\u4eba\uff0c\u6211\u4e5f\u4e0d \u77e5\u9053\u3002 \u5e95\u4e0b\u7684\u7121\u4e3b\u8a9e\u53e5\u4e0d\u5217\u5165\uff1a (6)\u90a3\u6642\u4f60\u7531\u80c3\u8178\u7684\u5495\u53eb\u8072\u97ff\u52d5\u983b\u7387\u63a8\u6e2c\uff0c\u53ef\u80fd\u5df2\u7d93\u508d\u665a\u4e86\u3002 \u5176\u4ed6\u52a9\u52d5\u8a5e\u6c92\u6709\u51fa\u73fe\u53e5\u9996\u7684\u60c5\u5f62\uff0c\u800c\u300c\u5fc5\u9808\u300d\u53ea\u51fa\u73fe\u4e0b\u5217\u4e00\u53e5\uff1a (7)\u5c31\u597d\u50cf\u662f\u5b8c\u6210\u4e00\u4ef6\u5de5\u4f5c\u6642\u7684\u559c\u6085\uff0c\u5fc5\u9808\u4f60\u89aa\u81ea\u53bb\u5de5\u4f5c\u3001\u53bb\u5b8c\u6210\uff0c\u7136\u5f8c \u81ea\u5fc3\u4e2d (3)\u672c\u6240\u63d0\u4f9b\u6b64\u9805\u81e8\u5e8a\u8a66\u9a57\u8a08\u756b\u5fc5\u9808\u7684\u8edf\u3001\u786c\u9ad4\u652f\u63f4\uff0c\u53ca\u81e8\u5e8a\u85e5\u7406\u5b78\u7814\u7a76\u3001 (4)\u672c\u9662\u73fe\u884c\u65b9\u6cd5\u64da\u4e86\u89e3\u6709\u4e09\u7a2e\u5fc5\u9808\u7684\u8655\u7406\u65b9\u5f0f\uff1a\u7b2c\u4e00\uff0c\u5148\u8a18\u9304\u767b\u9304\u865f\u4ee5 \u4fbf (5)\u5171\u5206\u70ba\u516b\u671f\uff0c\u8996\u9078\u624b\u72c0\u6cc1\u7f85\u5217\u5404\u7a2e\u5fc5\u9808\u7684\u8a13\u7df4\u5167\u5bb9\u3002 (6)\u8003\u8a66\u4ee3\u66ff\u6559\u5b78\uff0c\u672a\u63d0\u4f9b\u5b78\u751f\u9032\u884c\u601d\u8003\u6240\u5fc5\u9808\u7684\u6642\u7a7a\u689d\u4ef6\uff0c\u81f4\u4f7f\u5b78\u5b50\u4e0d \u662f\u9003\u907f\u6392\u65a5 2.3.3 \u52a9\u52d5\u8a5e\u51fa\u73fe\u65bc\u53e5\u5c3e Li &amp; Thompson \u8a0e\u8ad6\u540d\u8a5e\u5316\u6642\u6240\u8209\u7684\u53e5\u578b\uff1a \u300c*\u4ed6\u662f\u80fd\u7684\u300d\u548c\u5206\u88c2\u53e5\u985e\u4f3c\uff0c\u5169\u8005\u300c\u7684\u300d \u7684\u529f\u80fd\u662f\u5426\u76f8\u540c\u6709\u5f85\u9032\u4e00\u6b65\u7814\u7a76\uff0c\u56e0\u6b64\u672c\u6587\u4e0d\u5217\u5165\u8a0e\u8ad6\u3002\u52a9\u52d5\u8a5e\u51fa\u73fe\u65bc\u53e5\u5c3e\u6709\u4e09 \u7a2e\u60c5\u5f62\uff1a(a) \u52a9\u52d5\u8a5e\u5f8c\u7684\u52d5\u8a5e\u7d44\u7701\u7565\uff0c(b) \u52a9\u52d5\u8a5e\u4ee5\u5b50\u53e5\u70ba\u4e3b\u8a9e\u6216(c)\u5145\u7576\u5206\u88c2 \u53e5\u7684\u4fe1\u606f\u7126\u9ede\uff0c\u672c\u6587\u7684\u8a0e\u8ad6\u66ab\u6642\u6392\u9664(a)\u985e\u3002\u4e00\u822c\u52a9\u52d5\u8a5e\u6975\u5c11\u51fa\u73fe\u65bc\u53e5\u5c3e\uff0c\u5373\u4f7f \u6709\u4e5f\u662f\u4e00\u3001\u4e8c\u4f8b\u800c\u5df2\uff0c\u5982\u300c\u6562\u300d \u3001 \u300c(\u4e0d)\u80fd\u300d\u53ea\u898b\u4e00\u4f8b\uff0c\u800c\u300c\u5fc5\u9808\u300d\u4e5f\u53ea\u6709\u4e8c\u4f8b\uff1a (1)\u8a31\u591a\u4eba\u611f\u5230\u932f\u6115\uff1a\u5e79\u3002\u9023\u8abf\u67e5\u90fd\u4e0d\u6562\uff0c\u9019\u7f9e\u6b7b\u4eba\u7684\u570b\u6c11\u9ee8\uff01 (2)\u5e38\u898b\u7684\u7537\u6027\u4e0d\u5b55\u75c7\u6709\uff1a\u6027\u4ea4\u4e0d\u80fd\u3002\u5c3f\u9053\u4e0b\u88c2\u3002\u96b1\u776a\u75c7\u3002\u8166\u4e0b\u5782\u9ad4\u6a5f\u80fd \u4f4e\u4e0b (3)\u5340\u57df\u5716\u66f8\u9928\u6216\u793e\u5340\u5b78\u6821\uff0c\u5247 NREN \u662f\u5fc5\u9808\u7684\uff0c\u4f46\u5047\u5982\u53ea\u662f\u5efa\u7acb\u4e00\u500b\u5f88\u5927 \u7684\u7db2\u8def\uff0c (4)\u5f71\u97ff\u53ef\u80fd\u53ea\u662f\u5176\u4e2d\u7684\u4e00\u7a2e\uff0c\u4f46\u9810\u9632\u7e3d\u662f\u5fc5\u9808\u7684\uff0c\u4ed6\u5efa\u8b70\u96fb\u529b\u516c\u53f8\u67b6\u8a2d \u9ad8\u58d3\u96fb\u7dda\uff0c \u4e0b\u9762\u4ee5\u300c\u53ef\u4ee5\u3001\u61c9\u8a72\u3001\u6703\u300d\u70ba\u4f8b\u4f86\u8aaa\u660e\u52a9\u52d5\u8a5e\u51fa\u73fe\u65bc\u53e5\u5c3e\u7684\u9650\u5236\u6216\u529f\u80fd\u3002 \u82e5\u628a\u300c\u53ef\u300d\u8996\u70ba\u300c\u53ef\u4ee5\u300d\u7684\u7701\u7565\u6216\u8b8a\u9ad4\u800c\u5217\u5165\u8a0e\u8ad6\uff0c\u5247\u300c\u53ef\u300d\u7576\u5b50\u53e5\u4e3b\u8a9e\u7684\u8b02\u8a9e \u7684\u60c5\u5f62\u6050\u6015\u662f\u6240\u6709\u52a9\u52d5\u8a5e\u4e2d\u6700\u591a\u8005\uff0c\u7d04\u6709\u516b\u5341\u4f8b\uff1a (5)\u81ea\u958b\u666e\u6566\u524d\u5f80\u4e00\u5929\u5373\u53ef\u5f80\u8fd4\uff0c\u6240\u4ee5\u4f4f\u5728\u958b\u666e\u6566\u5373\u53ef\u3002 (8)\u6216\u8005\u4e00\u689d\u9f8d\uff0c\u6216\u662f\u96d5\u82b1\u7684\u7a97\u6afa\u7b49\u7b49\u5747\u53ef\u3002 8 773 \u7b46\u8cc7\u6599\u4e2d\u53ea\u6709\u4e00\u6b21\u51fa\u73fe\u65bc\u53e5\u5c3e(\u300c\u505a\u66f4\u591a\u66f4\u597d\u7684\u5b89\u6392\uff0c\u4f46\u662f\u611f\u60c5\u4e16\u754c\u537b\u4e0d\u4e00\u5b9a\u3002\u96e2\u5a5a\u5a66\u5973\u5728 11 \u5728\u5e73\u8861\u8a9e\u6599\u5eab\u9078\u4e2d\u7684 2000 \u500b\u300c\u80fd\u300d\u5b57\u4e2d\uff0c\u96d6\u7121\u6b63\u53cd\u554f\u53e5\uff0c\u4f46\u672a\u9078\u4e2d\u8005\u51fa\u73fe\u4e86\u4e00\u6b21\uff0c\u5982\u4e0b\u5217\u53e5 \u9304 0.36%\uff0c\u5171 9.85%\u3002 15 \u6b64\u8655\u5f62\u5bb9\u8a5e\u6307\u7684\u662f\u505a\u70ba\u540d\u8a5e\u4fee\u98fe\u8a9e\uff0c\u800c\u975e\u53e5\u5b50\u7684\u8b02\u8a9e\u3002 \u5c31\u300c\u53ef\u300d\u800c\u8a00\uff0c\u9808\u6709\u300c\u5373/\u4e0d/\u5747/\u624d/\u4fbf\u300d\u4fee\u98fe\u300c\u53ef\u300d\u624d\u80fd\u6210\u70ba\u5b50\u53e5\u4e3b\u8a9e\u7684\u8b02\u8a9e\uff1b \u4fbf\u62da\u547d\u7684\u300d)\u3002 (i)\u5728\u72c4\u65af\u8010\u7684\u5361\u901a\u4e16\u754c\uff0c\u6c92\u6709\u4ec0\u9ebc\u53ef\u80fd\u4e0d\u53ef\u80fd\uff0c\u97f3\u6a02\u7576\u7136\u4e5f\u662f\u3002 12 \u6839\u64da\u8a5e\u5eab\u5c0f\u7d44(1998\uff1a12)\uff0c\u53e3\u8a9e\u90e8\u4efd\u5305\u62ec\u6f14\u8b1b\u7a3f 1.38%\u3001\u5287\u5834\u53f0\u8a5e 0.82%\u3001\u6703\u8a71 7.29%\u53ca\u6703\u8b70\u8a18 \u5176\u4ed6\u526f\u8a5e\u5340\u9694\uff0c\u4ecd\u7a31\u70ba\u300c\u52a9\u52d5\u8a5e\u300d \uff1b\u7d71\u8a08\u7d50\u679c\u5247\u662f\u5e73\u8861\u8a9e\u6599\u5eab\u6240\u63d0\u4f9b\u7684\u3002 (9)\u56e0\u6b64\u53ea\u8981\u6309\u8aaa\u660e\u66f8\u5b89\u88dd\u4fbf\u53ef 7 159 \u7b46\u8cc7\u6599\u4e2d\u53ea\u51fa\u73fe\u4e00\u6b21( \u300c\u4ed4\u7d30\u89c0\u5bdf\u4e86\u4e00\u6703\uff0c\u5c31\u4ee4\u6211\u5012\u8db3\u4e86\u5473\u53e3\u3002\u5927\u6982\u5979\u8a8d\u5b9a\u4e86\u6211\u662f\u500b\u51a4\u5927\u982d\uff0c 14 \u6b64\u8655\u8a5e\u985e\u7684\u5224\u5b9a\u8acb\u53c3\u8003\u4e2d\u6587\u8a5e\u77e5\u8b58\u5eab\u5c0f\u7d44 (1993)\uff0c\u4ed6\u5011\u5c07\u60c5\u614b\u52a9\u52d5\u8a5e\u6b78\u65bc\u526f\u8a5e\uff0c\u6b64\u8655\u70ba\u4e86\u8207 10 \u300c\u53ef\u80fd\u4e0d\u53ef\u80fd\u300d\u53ea\u51fa\u73fe\u4e00\u6b21\uff0c\u537b\u662f\u540d\u8a5e\u7528\u6cd5(\u898b\u4e0b\u4f8b i)\uff0c\u56e0\u6b64\u4e0d\u8a08\u5165\uff1a 6 \u66f9(1990\u30011996)\u5c07\u300c\u597d\u50cf\u300d\u5206\u6790\u70ba\u8a8d\u77e5\u60c5\u614b\u52d5\u8a5e(\u5373\u672c\u6587\u4e4b\u52a9\u52d5\u8a5e)\u3002 \u5247\u6709 55 \u7b46\u8cc7\u6599\u3002 \u6b64\u5916\uff0c\u6211\u5011\u5728\u6574\u7406\u8a9e\u6599\u7684\u904e\u7a0b\u4e2d\u767c\u73fe\uff0c\u6709\u4e9b\u8a5e\u96d6\u88ab\u5217\u70ba\u52a9\u52d5\u8a5e\uff0c\u7136\u800c\u5728\u8a9e\u8a00\u4f7f (7)\u4f9d\u820a\u975e\u5f97\u5230\u5e7e\u500b\u516c\u7acb\u7684\u535a\u7269\u9928\u53bb\u4e0d\u53ef 9 \u6b64\u8655\u300c\u6703\u4e0d\u6703\u300d\u53ea\u6709 3 \u4f8b\u662f\u4ee5\u96a8\u6a5f\u53d6\u6a23\u5f97\u5230\u7684 2000 \u7b46\u70ba\u9650\uff1b\u82e5\u4ee5\u300c\u6703\u4e0d\u6703\u300d\u70ba\u95dc\u9375\u8a5e\u641c\u5c0b\u6642 \u8981\u300d\u5206\u6790\u70ba\u60c5\u614b\u540d\u8a5e\u6bd4\u52a9\u52d5\u8a5e\u5408\u9069\u3002 \u6216\u5f62\u5bb9\u8a5e(\u5982\u300c\u5f88\u4e0d\u9ad8\u8208\u300d)\u76f8\u540c\u3002\u96d6\u7136\u5b78\u8005\u90fd\u6307\u51fa\u52a9\u52d5\u8a5e\u53ef\u51fa\u73fe\u65bc\u6b63\u53cd\u554f\u53e5\u4e2d\uff0c (6)\u627e\u5e2b\u9577\u505a\u5fb5\u4fe1\u4eba\u6642\uff0c\u5fc5\u9808\u53d6\u5f97\u4ed6\u7684\u540c\u610f\u624d\u53ef\uff0c (16)\u5b69\u5b50\u5011\u770b\u5728\u773c\u88e1\uff0c\u5fc3\u60f3\uff1a\u4f60\u5011\u53ef\u4ee5\uff0c\u6211\u70ba\u4ec0\u9ebc\u4e0d\u53ef\u4ee5\uff1f (17)\u9ad4\u9a57\u7576\u4e2d\uff0c\u6211\u5011\u6703\u6025\u5207\u7684\u5436\u558a\uff1a \u300c\u5982\u679c\u53ef\u4ee5\uff0c\u8acb\u628a\u9019\u8981\u547d\u7684\u82e6\u676f\u79fb\u6389 \u5427\uff01\u300d (18)\uff1a \u300c\u4f60\u5728\u80e1\u8aaa\u4e9b\u4ec0\u9ebc\u554a\uff1f\u4e0d\u53ef\u4ee5\u5c31\u662f\u4e0d\u53ef\u4ee5\uff01\u6cd5\u5f8b\u5c31\u662f\u6cd5\u5f8b\u3002\u300d \u518d\u770b\u300c\u61c9\u8a72\u300d\u7684\u4f8b\u5b50\uff0c \u300c\u61c9\u8a72\u300d\u51fa\u73fe\u65bc\u53e5\u5c3e\u5145\u7576\u4e3b\u8981\u8ff0\u8a9e\u6642(\u5171 7 \u4f8b)\uff0c\u90fd\u662f\u7fa9\u52d9 \u7528\u6cd5\uff1a (19)\u505a\u4eba\u4e0d\u8981\u592a\u904e\u5206\uff0c\u8d95\u4ed6\u5011\u51fa\u53bb\u5df2\u4e0d\u61c9\u8a72\uff0c\u8b93\u4ed6\u5011\u5728\u8d70\u5eca\u5750\u5750\u53c8\u6709\u4f55 \u59a8\uff1f (20)\u540c\u5c45\u53c8\u592a\u8cbf\u7136\uff0c\u672a\u505a\u907f\u5b55\u63aa\u65bd\u771f\u4e0d\u61c9\u8a72\uff0c\u58ae\u80ce\u7576\u7136\u4e0d\u5f97\u5df2 \u53e6\u5916\uff0c \u300c\u61c9\u8a72\u300d\u6709 12 \u4f8b 5 \u51fa\u73fe\u65bc\u5206\u88c2\u53e5\uff0c\u4e5f\u90fd\u662f\u7fa9\u52d9\u7528\u6cd5\uff1a (21)\u6211\u7e3d\u89ba\u5f97\uff0c\u96a8\u4fbf\u5c0d\u4eba\u767c\u813e\u6c23\uff0c\u662f\u4e0d\u61c9\u8a72\u7684\u3002 (22)\u4e5f\u662f\u4e3b\u5f35\u6c11\u4e3b\u7684\u793e\u6703\uff0c\u591a\u5143\u5316\u662f\u61c9\u8a72\u7684\u4f46\u4e0d\u61c9\u8a72\u88ab\u5206\u6b67\u3001\u88ab\u6df7\u6dc6 (23)\u7236\u6bcd\u89aa\u7167\u9867\u4f60\u4e8c\u5341\u5e74\uff0c\u628a\u91ab\u5b78\u5ff5\u5b8c\u4e5f\u662f\u61c9\u8a72\u7684\u3002 2000 \u7b46\u300c\u6703\u300d\u505a\u70ba\u52d5\u8a5e\u51fa\u73fe\u65bc\u53e5\u5c3e\u8005\uff0c\u53ea\u6709 3 \u4f8b\uff0c3 \u4f8b\u7686\u662f\u8868\u793a\u61c2\u5f97\u4e4b\u610f\u7684\u4e00\u822c \u52d5\u8a5e\uff1a (24) \u662f\u505a\u4eba\u3002\u5982\u679c\uff0c\u9023\u6700\u8d77\u78bc\u7684\u505a\u4eba\u898f\u77e9\u90fd\u4e0d\u6703\uff0c\u5716\u756b\u5f97\u518d\u597d\uff0c\u4e5f\u6c92\u6709 \u7528\u3002 (25)\u4e26\u5728\u9069\u7576\u7684\u6a5f\u6703\u8868\u9054\uff0c\u5982\uff1a \u300c\u9019\u500b\u6211\u6703\uff0c\u6211\u53ef\u4ee5\u8a66\u8a66\u770b\u3002\u300d (26) \u6700\u8fd1\u4e00\u4ee3\u76ae\u7334\u4e86\uff0c\u73fe\u5728\u6c92\u4eba\u5b78\uff0c\u5c07\u4f86\u6c92\u4eba\u6703\uff0c\u5c31\u662f\u53bb\u535a\u7269\u9928\u770b 2.3.4 \u5c0f\u7d50 \u5c31\u8a9e\u6599\u6240\u898b\uff0c\u5404\u52a9\u52d5\u8a5e\u53ef\u51fa\u73fe\u7684\u4f4d\u7f6e\u5217\u8868\u5982\u4e0b\uff1a \u61c9\u7576 -\u253c -\u61c9 -\u253c -\u8a72 -\u253c -\u80fd\u5920 -\u253c -\u5f97 de -\u253c -\u5f97d e i -\u253c -\u5fc5\u5f97 -\u253c -\u80af -\u253c -\u9858\u610f -\u253c -\u5c31\u4ee5\u4e0a\u7684\u8a0e\u8ad6\u770b\u4f86\uff0c\u52a9\u52d5\u8a5e\u6700\u5e38\u898b\u7684\u4f4d\u7f6e\u662f\u53e5\u4e2d\uff0c\u9019\u4e00\u9ede\u548c\u60c5\u614b\u526f\u8a5e\u4e26\u7121\u592a\u5927\u5340 \u5206\u3002\u4e0b\u9762\u4ee5\u300c\u597d\u50cf\u3001\u5927\u6982\u3001\u4e00\u5b9a\u3001\u5fc5\u5b9a\u3001\u7d55\u5c0d\u300d\u4e94\u8a5e\u70ba\u4f8b\uff1a \u53e5\u9996 \u53e5\u4e2d \u53e5\u5c3e \u597d\u50cf 6 \u253c \u253c -\u5927\u6982 \u253c 7 \u253c -\u4e00\u5b9a -\u253c \u253c 8 \u5fc5\u5b9a -\u253c -\u7d55\u5c0d -\u253c -\u9019\u4e94\u500b\u8a5e\u4e00\u822c\u6b78\u70ba\u60c5\u614b\u526f\u8a5e\uff0c\u51fa\u73fe\u7684\u4f4d\u7f6e\u4e5f\u4ee5\u53e5\u4e2d\u70ba\u5178\u578b\u3002\u526f\u8a5e\u548c\u52a9\u52d5\u8a5e\u82e5\u8981\u6709 \u5340\u5225\uff0c\u6050\u6015\u662f\u526f\u8a5e\u6975\u5c11\u5728\u53e5\u5c3e\uff0c\u4e0d\u904e\u9019\u4e00\u9ede\u6709\u5f85\u5c0d\u526f\u8a5e\u505a\u8f03\u5168\u9762\u7684\u7814\u7a76\u624d\u80fd\u4e0b\u5b9a \u8ad6\u3002 2.4 \u6b63\u53cd\u554f\u53e5 \u96d6\u7136\u5169\u6587\u90fd\u80af\u5b9a\u52a9\u52d5\u8a5e\u53ef\u51fa\u73fe\u65bc\u6b63\u53cd\u554f\u53e5\u4e2d\uff0c\u4f46\u5b6b(1996\uff1a295)\u5c31\u6307\u51fa\u300c\u524d\u4eba\u6240 \u8a8d\u5b9a\u7684\u52a9\u52d5\u8a5e\u7684\u8a9e\u6cd5\u7279\u9ede\u591a\u534a\u90fd\u4e0d\u5177\u5099\u5c0d\u5167\u7684\u666e\u904d\u6027\u300d \u3002\u4ed6\u4ee5\u6b63\u53cd\u554f\u53e5(V-\u4e0d-V) \u70ba\u4f8b\uff0c\u8d99(1968)\u6240\u5217\u7684\u52a9\u52d5\u8a5e\u4e2d\uff0c\u6709 8 \u500b\u4e0d\u80fd\u7528\u3002\u6b64\u8655\u4ee5\u5e73\u8861\u8a9e\u6599\u5eab\u4e2d\u5404\u52a9\u52d5\u8a5e \u7684\u7528\u4f8b\u4f86\u770b\uff0c\u52a9\u52d5\u8a5e\u4f7f\u7528\u6b63\u53cd\u554f\u53e5\u7684\u983b\u7387\u4e0d\u50c5\u5f88\u4f4e\uff0c\u800c\u4e14\u51fa\u73fe\u9019\u7a2e\u53e5\u578b\u7684\u52a9\u52d5\u8a5e \u80fd 0 11 0 \u80fd\u5920 0 \u518d\u8005\u5c31\u524d\u56db\u5c0f\u7bc0\u8a0e\u8ad6\u7684\u52a9\u52d5\u8a5e\u7684\u8a9e\u6cd5\u7279\u9ede\u4e2d\uff0c\u82e5\u4e0d\u8003\u616e\u52a9\u52d5\u8a5e\u51fa\u73fe\u7684\u4f4d\u7f6e\uff0c\u4ee5 \u52d5\u8a5e 31(3.55) 36(2.47) 8(0.40) \u89c0\uff0c\u4e14\u4e0d\u7b26\u5408\u8aaa\u8a71\u8005\u7684\u8a9e\u8a00\u672c\u80fd\u3002 0 \u61c9\u7576 0 \u5176\u9918\u4e09\u500b\u689d\u4ef6\u4f86\u6aa2\u9a57 17 \u500b\u52a9\u52d5\u8a5e\u5728\u8a9e\u6599\u4e2d\u7684\u4f7f\u7528\u60c5\u5f62\uff0c\u5217\u51fa\u7c21\u8868\u5982\u4e0b\uff1a \u5f62\u5bb9\u8a5e 15 1609(80.45) 0 \u6562 0 0 \u5f97 dei 0 0 \u5f97 de 0 0 \u5fc5\u9808 0 0 \u5fc5\u5f97 0 0 \u5fc5\u8981 0 \u7a0b\u5ea6\u526f\u8a5e \u4ecb\u8a5e 16 \u53c3\u8003\u66f8\u76ee\uff1a 44(3.01) \u5426\u5b9a\u5f0f \u6b63\u53cd\u554f\u53e5 \u300c\u61c9\u8a72\u300d \u3001 \u300c\u61c9\u300d \u3001 \u300c\u8a72\u300d\u548c\u300c\u61c9\u7576\u300d\u56db\u8005\u5728\u8a9e\u6599\u5eab\u4e2d\u505a\u70ba\u52a9\u52d5\u8a5e\u7684\u983b\u7387\u5dee\u5225\u9817\u5927\u3002 \u6731\u5fb7\u7199 1982 \u8a9e\u6cd5\u8b1b\u7fa9 \u5546\u52d9\u5370\u66f8\u9928\u3002 \u4fee\u98fe \u6703 \u253c \u253c \u4ee5\u300c\u8a72\u300d\u70ba\u4f8b\uff0c\u8a9e\u6599\u4e2d\u6709\u516b\u6210\u662f\u505a\u70ba\u6307\u793a\u8a5e\u4f86\u4fee\u98fe\u540d\u8a5e\uff0c\u53ef\u8003\u616e\u628a\u6b64\u8a5e\u6392\u9664\u65bc\u52a9 \u5442\u53d4\u6e58 1984 \u6f22\u8a9e\u516b\u767e\u8a5e \u5546\u52d9\u5370\u66f8\u9928\u3002 \u253c \u61c9\u8a72 \u253c \u253c \u52d5\u8a5e\u4e4b\u5916\u3002\u7531\u6b64\u89c0\u4e4b\uff0c\u4ee5\u8a9e\u6599\u4e2d\u6240\u898b\u52a9\u52d5\u8a5e\u7684\u975e\u60c5\u614b\u7528\u6cd5\u70ba\u4f8b\u53ef\u7528\u4f86\u8aaa\u660e\uff0c\u8a9e\u8a00 \u5b6b\u5fb7\u91d1 1997 \u6f22\u8a9e\u52a9\u52d5\u8a5e\u7684\u7bc4\u570d\uff0c\u8a5e\u985e\u554f\u984c\u8003\u5bdf p.286-307\u3002 \u253c \u61c9 \u253c \u253c \u253c \u7684\u4f7f\u7528(\u6b64\u8655\u4ee5\u8a9e\u6599\u70ba\u4ee3\u8868)\u548c\u4e3b\u89c0\u5370\u8c61\u5f80\u5f80\u6709\u4e00\u6bb5\u843d\u5dee\u3002 \u8a5e\u5eab\u5c0f\u7d44 1993 \u4e2d\u6587\u8a5e\u985e\u5206\u6790(\u4e09\u7248)\uff0c\u4e2d\u592e\u7814\u7a76\u9662\u8cc7\u8a0a\u79d1\u5b78\u6240\uff0c\u8a5e\u5eab\u5c0f\u7d44\uff0c 0 \u53ef\u4ee5 \u253c \u253c \u253c \u5b6b\u5fb7\u91d1(1997:295)\u6307\u51fa\u300c\u901a\u7528\u8a9e\u6599\u4e2d\u7684\u52a9\u52d5\u8a5e\u5fc5\u5b9a\u662f\u975e\u540c\u8cea\u6027\u7684\uff0c\u6709\u53e3\u8a9e\u7684\uff0c\u6709 \u53f0\u5317\u5357\u6e2f\u3002 \u56e0\u4f8b\u5b50\u4e0d\u591a\uff0c\u5168\u90e8\u5217\u51fa\u5982\u4e0b\uff1a (1)\u7528\u624b\u53bb\u62c9\u5979\uff0c\u56e0\u70ba\u7537\u5973\u6388\u53d7\u4e0d\u89aa\u3002\u5230\u5e95\u8a72\u4e0d\u8a72\u62c9\u5462\uff1f (2)\u7236\u89aa\uff1a\u4f60\u62ff\u4e00\u500b\u5427\u3002\u5b54\u878d\uff1a (\u4e0d\u77e5\u9053\u8a72\u4e0d\u8a72\u62ff\uff0c\u56de\u982d\u770b\u770b\u6bcd\u89aa) \u3002 (3)\u4e0d\u77e5\u9053\u5927\u5bb6\u9858\u4e0d\u9858\u610f\u4e5f\u7528\u540c\u6a23\u7684\u7df4\u7fd2\uff0c\u53bb\u5c0d\u8457\u93e1\u5b50\uff0c\u597d\u597d\u7684 (4)\u6211\u771f\u80fd\u70ba\u5225\u4eba\u8457\u60f3\u55ce\uff1f\u90a3\u9ebc\uff0c\u6211\u61c9\u4e0d\u61c9\u8a72\u633d\u7559\u51f1\u6d1b\u7433\u5462\uff1f \u9858\u610f \u253c \u253c \u253c \u53ef\u80fd \u253c \u253c \u8a5e\u5eab\u5c0f\u7d44 1998 \u4e2d\u592e\u7814\u7a76\u9662\u5e73\u8861\u8a9e\u6599\u5eab\u7684\u5167\u5bb9\u8207\u8aaa\u660e(\u4fee\u8a02\u7248) \uff0c\u4e2d\u592e\u7814\u7a76 \u66f8\u9762\u8a9e\u7684\uff0c\u6709\u6587\u8a00\u7684\u300d \uff0c\u5c31\u6211\u5011\u7684\u8a9e\u6599\u4f86\u770b\uff0c\u96d6\u7136\u300c\u61c9\u8a72\u300d \u3001 \u300c\u61c9\u7576\u300d \u3001 \u300c\u61c9\u300d\u548c\u300c\u8a72\u300d -\u80fd \u253c \u253c \u9662\u8cc7\u8a0a\u79d1\u5b78\u6240\uff0c\u8a5e\u5eab\u5c0f\u7d44\uff0c\u53f0\u5317\u5357\u6e2f\u3002 \u505a\u70ba\u52a9\u52d5\u8a5e\u6642\uff0c\u5f80\u5f80\u88ab\u8996\u70ba\u540c\u7fa9\u8a5e\uff0c\u5c31\u6211\u5011\u7684\u8a9e\u611f\u800c\u8a00\uff0c \u300c\u61c9\u300d\u4f3c\u4e4e\u662f\u8f03\u70ba\u6587\u8a00 -\u61c9\u7576 \u253c \u253c \u7684\u8a5e\u5f59\uff0c\u800c\u300c\u61c9\u8a72\u300d\u6216\u300c\u8a72\u300d\u61c9\u5c6c\u53e3\u8a9e\u7684\u8a5e\u5f59\u3002\u7406\u8ad6\u4e0a\u5728\u65e5\u5e38\u751f\u6d3b\u4e2d\uff0c\u5f8c\u8005\u51fa\u73fe \u5f35\u975c 1961 \u8ad6\u6f22\u8a9e\u526f\u8a5e\u7684\u7bc4\u570d \u4e2d\u570b\u8a9e\u6587 8 \u6708\u865f 1-14\u3002 -\u6562 \u253c \u253c \u7684\u983b\u7387\u61c9\u9ad8\u65bc\u524d\u8005\uff0c\u7136\u800c\u8a9e\u6599\u537b\u544a\u8a34\u6211\u5011\uff0c \u300c\u61c9\u300d\u7684\u4f7f\u7528\u6b21\u6578\u53cd\u800c\u6bd4\u300c\u61c9\u8a72\u300d\u6216 \u6e6f\u5ef7\u6c60 1984 \u570b\u8a9e\u7684\u52a9\u52d5\u8a5e \u4e2d\u570b\u8a9e\u6587 22-28\u3002 -\u5fc5\u9808 \u253c \u253c -\u300c\u8a72\u300d\u9ad8\u5f97\u591a\uff0c\u9019\u53ef\u80fd\u662f\u76ee\u524d\u5e73\u8861\u8a9e\u6599\u5eab\u7684\u8a9e\u6599\u4f86\u6e90\u4ee5\u66f8\u9762\u8a9e\u70ba\u4e3b(\u9054\u767e\u5206\u4e4b\u4e5d \u6e6f\u5ef7\u6c60&amp;\u6e6f\u5fd7\u771f 1998 \u83ef\u8a9e\u60c5\u614b\u8a5e\u5e8f\u8ad6 \u7b2c\u4e94\u5c46\u83ef\u8a9e\u6587\u6559\u5b78\u7814\u8a0e\u6703\u8ad6\u6587\u96c6 (5)\u4e0d\uff0c\u6b63\u78ba\u7684\u8aaa\u6cd5\uff0c\u61c9\u8a72\u662f\u51f1\u6d1b\u7433\u61c9\u4e0d\u61c9\u8a72\u7559\u5728\u9019\u88e1\uff1f \u8a72 -\u253c \u253c \u5341\u4ee5\u4e0a)\u6240\u81f4\u3002\u76ee\u524d\u53d7\u9650\u65bc\u8a9e\u6599\u5eab\uff0c\u7121\u6cd5\u5f97\u77e5\u771f\u76f8\uff1b\u5f85\u5c07\u4f86\u8a9e\u6599\u5eab\u80fd\u6539\u9032\u5230\u53e3\u8a9e 177-197 \u9801\u3002 (6)\u7de8\u7e54\u91cd\u9022\u7684\u6545\u4e8b\uff1b\u53ef\u4e0d\u53ef\u4ee5\u7528\u611b\u628a\u904e\u53bb\u90fd\u7d50\u675f\uff0c\u771f\u5fc3\u651c\u624b\u80a9\u4e26\u80a9\uff0c \u80af -\u253c \u253c \u8207\u66f8\u9762\u8a9e\u8cea\u548c\u91cf\u76f8\u7576\u5f8c\uff0c\u53ef\u9032\u4e00\u6b65\u63a2\u8a0e\u8a9e\u5f0f\u548c\u52a9\u52d5\u8a5e\u4f7f\u7528\u7684\u95dc\u4fc2\u3002 \u912d \u7e08 1999 \u53f0\u7063\u570b\u8a9e\u3001\u95a9\u5357\u8a9e\u548c\u5ba2\u5bb6\u8a71\u5404\u985e\u60c5\u614b\u8a5e\u642d\u914d\u95dc\u4fc2\u7684\u6bd4\u8f03\uff0c\u570b\u79d1 (7)\u6700\u5f8c\u624d\u8aaa\u4e86\uff1a\u9019\u4e9b\u8c9d\u6bbc\u53ef\u4e0d\u53ef\u4ee5\u9001\u7d66\u6211\uff1f\u4e0d\u884c\uff0c\u6211\u53ea\u5269\u4e0b\u9019\u4e9b\u4e86\u3002 (8)\u3002\u5582\uff0c\u4f60\u4e0d\u8981\u8001\u770b\u80a1\u7968\u90a3\u4e00\u7248\uff0c\u53ef\u4e0d\u53ef\u4ee5\u807d\u807d\u6211\u8b1b\u8a71\uff1f (9)\u6211\u6703\u7acb\u523b\u52a0\u85e5\uff0c\u5343\u842c\u4e0d\u53ef\u5927\u8072\u56b7\u56b7\u3002\u958b\u5200\u6703\u4e0d\u6703\u75db\uff1f\u96c5\u9e97\u554f\u6211\u3002 (10)\u4e0a\u5b98\u8d74\u4efb\uff0c\u7d50\u5a5a\u5ac1\u5a36\u771f\u4e0d\u597d\u610f\u601d\uff0c\u4e0d\u66c9\u5f97\u6703\u4e0d\u6703\u5f71\u97ff\u4f60\u8fa6\u684c\uff1f\u3002 (11)\u4f46\u5979\u53c8\u731c\u60f3\u6703\u4e0d\u6703\u662f\u56e0\u70ba\u5979\u77e5\u9053\u7537\u53cb\u548c\u820a\u60c5\u4eba\u78b0\u9762\uff0c \u80fd\u5920 \u253c --\u5f97dei \u253c -\u6703\u8a08\u756b\u6210\u679c\u5831\u544a(NSC 88-2411-H-126-005)\u3002 3.\u7d50\u8a9e -\u5f97 de -\u253c \u7406\u8ad6\u6216\u5047\u8a2d\u95e1\u8ff0\u7684\u662f\u8aaa\u8a71\u8005\u7684\u8a9e\u8a00\u672c\u80fd\uff0c\u800c\u8a9e\u6599\u53cd\u6620\u7684\u662f\u8a9e\u8a00\u7684\u904b\u7528\uff0c\u7d9c\u5408\u4ee5\u4e0a \u912d\u7e08 2000 \u6f22\u8a9e\u60c5\u614b\u52d5\u8a5e\u7684\u8a5e\u5e8f, \u7b2c\u4e5d\u5c46\u570b\u969b\u6f22\u8a9e\u8a9e\u8a00\u5b78\u6703\u8b70(IACL-9) , -\u5fc5\u8981 -\u253c \u8a0e\u8ad6\u53ef\u77e5\uff0c\u4ee5\u5f80\u5b78\u8005\u7814\u7a76\u8a9e\u8a00\u73fe\u8c61\u6642\u591a\u4ee5\u500b\u4eba\u8a9e\u611f\u70ba\u6e96\uff0c\u9020\u6210\u5169\u8005\u4e4b\u9593\u5f80\u5f80\u6709\u843d \u65b0\u52a0\u5761\u3002 -\u5fc5\u5f97 --\u5dee\u3002\u4ee5\u80fd\u5426\u88ab\u7a0b\u5ea6\u526f\u8a5e(\u5f88/\u66f4)\u4fee\u98fe\u9019\u500b\u7279\u9ede\u70ba\u4f8b\uff0c\u5c0d Li &amp; Thompson \u800c\u8a00\u662f\u6c7a \u66f9\u9022\u752b 1996 \u6f22\u8a9e\u7684\u63d0\u5347\u52d5\u8a5e\uff0c\u4e2d\u570b\u8a9e\u6587\uff0c172-182 \u9801\u3002 -\u5c31\u4e0a\u8868\u4f86\u770b\uff0c17 \u500b\u52a9\u52d5\u8a5e\u4e2d\u53ea\u6709\u300c\u6703\u3001\u61c9\u8a72\u3001\u61c9\u3001\u53ef\u4ee5\u3001\u9858\u610f\u300d\u4e94\u500b\u8a5e\u5b8c\u5168 \u5b9a\u52a9\u52d5\u8a5e\u7684\u4e00\u500b\u689d\u4ef6\uff0c\u6e6f&amp;\u6e6f\u5247\u4ee5\u4e4b\u5340\u5206\u60c5\u614b\u52d5\u8a5e\u548c\u5f62\u5bb9\u8a5e\uff0c\u7d50\u679c\u8a9e\u6599\u986f\u793a\u591a\u6578 Charles N. Li and Sandra A. Thompson 1981 Mandarin Chinese :a (12)\u5728\u9032\u884c\u5efa\u4ea4\u4f5c\u696d\u6642\uff0c\u6211\u5011\u7adf\u6703\u53bb\u722d\u8ad6\u61c9\u4e0d\u61c9\u627f\u8a8d\u4ed6\u5011\uff1f \u6eff\u8db3\u52a9\u52d5\u8a5e\u7684\u9019\u4e09\u500b\u7279\u9ede\uff0c\u5176\u4ed6\u5247\u6216\u591a\u6216\u5c11\u5177\u5099\u5176\u4e2d\u90e8\u4efd\u7684\u7279\u9ede\uff0c\u800c\u300c\u5fc5\u5f97\u300d \u52a9\u52d5\u8a5e\u662f\u53ef\u8207\u300c\u5f88\u300d\u6216\u300c\u66f4\u300d\u642d\u914d\uff0c\u7136\u800c\u52a9\u52d5\u8a5e\u642d\u914d\u7a0b\u5ea6\u526f\u8a5e\u6642\uff0c\u53ef\u80fd\u6709\u9650\u5236\uff0c functional reference grammar, Berkeley :University of \u76f4\u89ba\u4e0a\uff0c\u52a9\u52d5\u8a5e\u90fd\u61c9\u8a72\u53ef\u4ee5\u51fa\u73fe\u65bc\u6b63\u53cd\u554f\u53e5\u4e2d\uff0c\u7136\u800c\u6aa2\u9a57\u8a9e\u6599\u7684\u7d50\u679c\u537b\u975e\u5982\u6b64\u3002 \u5247\u5b8c\u5168\u4e0d\u7b26\u5408\u4efb\u4f55\u4e00\u9ede\uff0c\u6240\u4ee5\u300c\u5fc5\u5f97\u300d\u61c9\u6392\u9664\u65bc\u52a9\u52d5\u8a5e\u4e4b\u5916\u3002\u53e6\u5916\uff0c\u96d6\u7136\u300c\u5fc5 \u5982\u300c\u6703\u300d\u53ef\u8868\u80fd\u529b\u6216\u9810\u65b7\uff0c\u8a9e\u6599\u4e2d\u524d\u8005\u4e3b\u8981\u642d\u914d\u300c\u5f88\u300d\u800c\u5f8c\u8005\u53ea\u8207\u300c\u66f4\u300d\u5171\u73fe\uff1b California Press\u3002 \u53ef\u80fd\u662f\u6b63\u53cd\u554f\u53e5\u662f\u8f03\u53e3\u8a9e\u7684\u53e5\u578b\uff0c\u4f46\u8a9e\u6599\u5eab\u56e0\u53e3\u8a9e\u8cc7\u6599\u4f86\u6e90\u8f03\u5c11\uff0c\u53e3\u8a9e\u90e8\u4efd(\u5305 \u8981\u300d\u53ef\u4ee5\u88ab\u300c\u4e0d\u300d\u5426\u5b9a\uff0c\u5728\u6aa2\u9a57\u8a9e\u6599\u7684\u904e\u7a0b\u4e2d\u537b\u767c\u73fe\uff0c \u300c\u5fc5\u8981\u300d\u7528\u70ba\u4e3b\u8cd3\u8a9e(\u540d \u300c\u61c9\u8a72\u300d\u5247\u517c\u6709\u7fa9\u52d9\u548c\u9810\u65b7\u7528\u6cd5\uff0c\u4f46\u53ea\u898b\u7fa9\u52d9\u7528\u6cd5\u548c\u7a0b\u5ea6\u526f\u8a5e\u4e00\u8d77\u51fa\u73fe\u3002\u52a9\u52d5\u8a5e Tsao, F-F(\u66f9\u9022\u752b) 1990 Sentence and clause structure in Chinese, \u53ea\u6709\u4e0b\u5217\u524d 7 \u500b\uff1a \u8a5e)\u6216\u5b9a\u8a9e(\u4fee\u98fe\u540d\u8a5e)\u8005\u9054 88.41% 13 \uff0c\u9019\u5728\u5176\u4ed6\u52a9\u52d5\u8a5e\u4e2d\u5341\u5206\u5c11\u898b\uff0c\u56e0\u6b64\u628a\u300c\u5fc5 \u7684\u5426\u5b9a\u5f0f\u5982 \u300c\u4e0d\u61c9\u8a72\u300d \u53ef\u4ee5\u88ab\u7a0b\u5ea6\u526f\u8a5e\u4fee\u98fe\uff0c\u9019\u4e00\u9ede\u548c\u5fc3\u7406\u52d5\u8a5e (\u5982 \u300c\u5f88\u4e0d\u559c\u6b61\u300d ) Taipei: Student Book Co.\u3002</td></tr></table>"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-1009.json

@@ -0,0 +1,302 @@
+{
+    "paper_id": "O00-1009",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:08.331120Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "Computer learner corpora have been widely used by SLA/EFL specialists since mid 1990s to gain better insights into authentic learner language. The work presented in this paper examines the inter-language of Taiwanese learners of English from a part-of-speech sequence perspective. Two pre-tagged corpora (one learner corpus and one native corpus) are involved in this work. The experimental results indicate that there are more than one third of eligible POS trigrams that are never practiced by the Taiwanese learners in their writing and the learners have stronger preference than native speakers in using pronouns, especially right after punctuations, verbs and conjunctions.",
+    "pdf_parse": {
+        "paper_id": "O00-1009",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "Computer learner corpora have been widely used by SLA/EFL specialists since mid 1990s to gain better insights into authentic learner language. The work presented in this paper examines the inter-language of Taiwanese learners of English from a part-of-speech sequence perspective. Two pre-tagged corpora (one learner corpus and one native corpus) are involved in this work. The experimental results indicate that there are more than one third of eligible POS trigrams that are never practiced by the Taiwanese learners in their writing and the learners have stronger preference than native speakers in using pronouns, especially right after punctuations, verbs and conjunctions.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "Rebecca H. Shih * , John Y. Chiang + and F. Tien +",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Part-of-speech Sequences and Distribution in a Learner Corpus of English",
+                "sec_num": null
+            },
+            {
+                "text": "With the recognition of its theoretical and practical potential, computer learner corpora (CLC) have been subsequently built up around the world since early 1990s. [1] CLC research aims to gain a better insight into learners' inter-language from the authentic data. The research often involves comparisons between inter-language that learners possess and native language on various linguistic features. For instance, the frequency distributions of most commonly-used words in a native and seven eastern European learner corpora are compared on various parts-of-speech categories [2] ; the use of complement clauses in terms of their frequencies in four learner corpora as contrasted with their native counterparts [3] is studied; the use of adverbial connectors by Swedish learners in comparison with the natives' is examined [4] . ",
+                "cite_spans": [
+                    {
+                        "start": 164,
+                        "end": 167,
+                        "text": "[1]",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 579,
+                        "end": 582,
+                        "text": "[2]",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 714,
+                        "end": 717,
+                        "text": "[3]",
+                        "ref_id": "BIBREF2"
+                    },
+                    {
+                        "start": 826,
+                        "end": 829,
+                        "text": "[4]",
+                        "ref_id": "BIBREF4"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Perplexity, in speech recognition community, is often referred to as the number of equi-probable choices at each step of word prediction in a language model such as a bigram/trigram model under the assumption that a word depends merely on the previous one/two words. In this work, given a corpus L, the perplexity of the corpus, S(L), can be viewed as a measure of diversity for the next POS in a language model, and it is defined as:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Corpus Perplexity in Bigram and Trigram models",
+                "sec_num": "3.1"
+            },
+            {
+                "text": ") ( 2 ) ( L H L S = \u2211 = c c k H N L H ) ( 1 ) ( \u2211 \u2212 = k c c k P c k P k H ) | ( log ) | ( ) ( 2",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Corpus Perplexity in Bigram and Trigram models",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "where H(L) is the entropy of the corpus L, N is the size of part-of-speech set, and P(k|c) is the probability that k will be the next POS when the current POS is c.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Corpus Perplexity in Bigram and Trigram models",
+                "sec_num": "3.1"
+            },
+            {
+                "text": ". In this experiment, the perplexities of BNC and TLCE corpora are calculated using both bigram and trigram models, and the results are shown in As can be seen in Table 1 , the perplexities of BNC corpus in the two language models are both greater than those of TLCE, especially in the trigram model where the degree of POS diversity in the learner corpus is only 2/3 of BNC's. The above phenomena can be explained by the limiting sentence structure varieties the learners possess.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 163,
+                        "end": 170,
+                        "text": "Table 1",
+                        "ref_id": "TABREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Corpus Perplexity in Bigram and Trigram models",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "In order to further understand the limit of structure variety in learners' writing, the numbers of POS trigrams, i.e. sequences of three POSs, used in the two corpora are compared and shown in Table 2 : the number of POS trigrams in the corpora Under the same assumption, Figure 1 depicts the divergence of learners' use of trigrams from BNC, the optimum indicated by the square curve, on the scale of top-ranking trigrams in use. The diamond curve denotes the number of the learners' trigrams that overlap with BNC at the same rank. As illustrated, the learners' curve moves away from the optimum when the scope of the rank enlarges, especially after the rank of 1000. Figure 1 : The divergence of the use of POS trigrams",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 193,
+                        "end": 200,
+                        "text": "Table 2",
+                        "ref_id": "TABREF2"
+                    },
+                    {
+                        "start": 272,
+                        "end": 280,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 670,
+                        "end": 678,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Structure Variety",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "As the learners have preference in using certain POS trigrams it is then desirable to understand the learners' preference in using POSs themselves as well. Figure 2 shows the POS distribution in each corpus, and only those taking up at east 5% of the corpus are indicated. Two significant phenomena are observed from the figure. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 156,
+                        "end": 164,
+                        "text": "Figure 2",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "POS Distribution",
+                "sec_num": "3.3"
+            },
+            {
+                "text": "The results of the preliminary experiments above show that there are more than one third of BNC trigrams that the learners never practice in their writing, whereas there are 4.5% of TLCE trigrams which do not appear in the BNC's. It is intended to believe that this small proportion of TLCE trigrams is contributed from the learner's writing errors. However, increasing the size of the native speaker corpus to observe any changes in the distribution of the trigrams will clarify the findings. It is also worth looking into those BNC trigrams that the learners do not know or are not aware of, and then isolating those with high frequency for the pedagogical purpose.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Discussions and future work",
+                "sec_num": "4."
+            },
+            {
+                "text": "The experimental results also suggest that the learners use pronouns excessively in their writing and that they have stronger preference than native speakers in using pronouns right after punctuations, verbs and conjunctions but less preference after prepositions and nouns. Pronouns often appear in the informal register, and as the corpus is composed of college students' compositions as well as their weekly journals, the informality of the journals may contribute partly to their excessive use of pronouns. So, it is desirable in the next stage of the work to divide the learner corpus in terms of its different registers and compare their POS distributions with the native speaker corpus.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Discussions and future work",
+                "sec_num": "4."
+            }
+        ],
+        "back_matter": [
+            {
+                "text": "The authors would like to thank the National Science Council, Taiwan, for supporting this project, and Prof. Ching-Yuan Tsai for his insightful comment.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Acknowledgements",
+                "sec_num": null
+            }
+        ],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "The International Corpus of Learner English, in English Language Corpora: Design, Analysis and Exploitation",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Granger",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "57--69",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Granger, S., The International Corpus of Learner English, in English Language Corpora: Design, Analysis and Exploitation, J. Aarts, P.d. Haan, and N. Oostdijk, Editors. 1993, Rodopi: Amsterdam. p. 57-69.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Overstatement in advanced learners' writing: stylistic aspects of adjective intensification",
+                "authors": [
+                    {
+                        "first": "G",
+                        "middle": [],
+                        "last": "Lorenz",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "53--66",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lorenz, G., Overstatement in advanced learners' writing: stylistic aspects of adjective intensification, in Learner English on Computer, S. Granger, Editor. 1998, Addison Wesley Longman Limited. p. 53-66.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Comparing native and learner perspectives on English grammar: a study of complement clauses",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Biber",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Reppen",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "Learner English on Computer",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Biber, D. and R. Reppen, Comparing native and learner perspectives on English grammar: a study of complement clauses, in Learner English on Computer, S.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "The use of adverbial connectors in advanced Swedish learners' written English",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Tapper",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Tapper, M., The use of adverbial connectors in advanced Swedish learners' written English, in Learner English on Computer, S. Granger, Editor. 1998,",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "type_str": "figure",
+                "text": "Firstly, although N(Noun) and VB(Verb) are the first two leading POSs in both corpora, there exists a distinct discrepancy of the percentage difference between the two. The difference in distribution percentage between N and VB in BNC reaches 9%, whereas merely 1% difference in TLCE. Secondly, PRON(pronoun), the 3rd highest distribution in the learner corpus but the 7 th in BNC, apparently is overused the learners. Distribution of Preceding POSs in PRON bigrams As the previous figure indicates the excessive use of PRON in the learner corpus, the phenomenon is further analyzed by examining the likelihood of each POS preceding PRON in the bigrams. Figure 3 shows the distribution of preceding POSs of PRON in each corpus. As seen, PUNC(punctuation), VB and CONJ(conjuction) are the three most likely POSs in TLCE to be followed by PRON, and the learners also have stronger preference in using these bigrams than the native speakers. By contrast, the bigrams, PREP(preposition)+PRON and N+PRON, are used more frequently by the native speakers than the learners. Distribution of Preceding POSs in PRON bigrams",
+                "num": null,
+                "uris": null
+            },
+            "TABREF0": {
+                "content": "<table><tr><td>It is based on two corpora: Taiwanese Learner corpus of English (TLCE) and British</td></tr><tr><td>National Corpus (BNC). Both corpora are tagged by TOSCA tagger, using the</td></tr><tr><td>TOSCA-ICLE tagset. The details of the corpora and the tagger will be stated</td></tr><tr><td>subsequently in Section 2, which is followed by a series of experiments in Section 3.</td></tr><tr><td>Conclusions are drawn in Section 4 with future work.</td></tr><tr><td>2. Methodology</td></tr><tr><td>2.1 Corpora: TLCE and BNC</td></tr><tr><td>As stated in the introduction, CLC-research often compares non-native data with</td></tr><tr><td>native data in order to reveal the overuse and/or underuse phenomena in a learner</td></tr><tr><td>corpus. In this work, the Taiwanese Learner Corpus of English (TLCE) is under</td></tr><tr><td>investigation and the British National Corpus (BNC) is used for comparison. TLCE of</td></tr><tr><td>455,000 words is Appendix A)</td></tr></table>",
+                "num": null,
+                "html": null,
+                "type_str": "table",
+                "text": ", what particular areas of language behavior that are shared by learners with different backgrounds, and to what extent these phenomena appear in learner English. The aim of the work in this paper is to discover distinctive inter-language features of Taiwanese learners of English in terms of part-of-speech sequences and distribution. For instance, word forms such as takes, took, taken, and taking have the same lemma take. This function facilitates the collocation analysis under the same lemma. TOSCA operates with a lexicon, which currently contains about 160,000 lemma-tag pairs, covering about 90,000 lemmas. The TOSCA-ICLE tagset contains 270 different tags within 18 major word classes.For simplicity, only the major word classes are considered in the current study (see"
+            },
+            "TABREF1": {
+                "content": "<table><tr><td>:</td></tr></table>",
+                "num": null,
+                "html": null,
+                "type_str": "table",
+                "text": ""
+            },
+            "TABREF2": {
+                "content": "<table><tr><td>BNC</td><td>TLCE</td><td>overlap</td></tr><tr><td>2531</td><td>1649</td><td>1574</td></tr></table>",
+                "num": null,
+                "html": null,
+                "type_str": "table",
+                "text": "As seen in the table, there are 2531 trigram patterns in BNC, 1649 in TLCE, and 1574 in both. If those appearing in BNC can be viewed"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-1011.json

@@ -0,0 +1,533 @@
+{
+    "paper_id": "O00-1011",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:12.678572Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-1011",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "\u5047\u8a2d\u6709\u4e00\u5305\u542b M \u500b \u7684 \u9663\u5217\uff0c\u6bcf\u4e00\u7d44\u76f8\u9130\u7684 \u7684\u8ddd \u70ba d\uff0c \u6709\u4e00\u8a9e\u97f3 \u8a0a (\u5047\u8a2d\u70ba\u5e73\u9762\u6ce2)\u5f9e\u6211\u5011 \u51fa\u7684\u6700\u4f73\u65b9\u5411 s \u50b3 \u904e\u4f86\uff0c \u7684\u8f38\u51fa\u70ba i t x \uff0c M i 1 \uff0c \u5247\u5728\u6642\u9593 t \u7684\u6642\u5019\uff0c\u7576\u7b2c i \u5230\u5e73\u9762\u6ce2\u7684\u8a0a \uff0c\u7b2c i+1 \u5247\u9700\u7b49\u5230 \u6ce2\u518d\u524d\u9032 \u8ddd R ( s d R cos = )\u65b9\u53ef \u5230\u8a0a \uff0c\u5982\u5716\u4e8c\u6240\u793a\u3002 \u82e5 \u6ce2\u7684\u901f\u5ea6\u70ba C\uff0c\u5247\u7b2c i+1 \u500b \u7684\u6642\u9593 C d C R s cos = = (1) 1 + + = i t i t x x \uff0c\u56e0\u6b64\u6211\u5011\u53ef\u4ee5\u4f30\u7b97\u7b2c i \u500b \u8207\u7b2c 1 \u500b \u7684\u95dc\u4fc2\u5982\u4e0b\uff1a 1 ) 1 ( \u2212 + = i t i t x x (2)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Delay-and-Sum Beamformer",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "\u800c\u6574\u500b Delay-and-Sum Beamformer \u7684\u8f38\u51fa t",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Delay-and-Sum Beamformer",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "x \uff0c\u5982\u5716\u4e09\u6240\u793a\uff0c\u5c31\u662f\u5c07\u6bcf\u500b \u9593\u7684\u6642\u9593",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Delay-and-Sum Beamformer",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "\u4f5c\u88dc \u5f8c\u5408\u6210\u518d\u53d6\u5e73\u5747\u800c\u5f97 \u2211 = \u2212 + = M i i i t t M 1 ) 1 ( 1 x x (3)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Delay-and-Sum Beamformer",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "\u5716\u4e8c\u3001\u76f8\u9130 \u7684\u6642\u9593 \u5716\u4e09\u3001Delay-and-Sum Beamformer \u6d41\u7a0b\u5716 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Delay-and-Sum Beamformer",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5\u7684\u4e3b\u8981\u76ee\u7684\u5728\u65bc\u4f30 \u51fa\u8a9e\u8005\u767c \u7684\u65b9\u5411\uff0c\u5176\u7cfb\u7d71\u6d41\u7a0b\u5716\u5982\u5716\u56db\u6240\u793a\uff0c\u6211 \u5011\u5c07\u5206\u6210\u4e0b\u5217\u4e09\u90e8\u4efd\u505a\u8aaa\u660e\u3002 \u5716\u56db\u3001\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5(SLA)\u6d41\u7a0b\u5716 \u9996\u5148\uff0c\u6211\u5011\u5c07 M \u500b \u5728\u6642\u57df\u4e0a\u7684\u8a9e\u97f3\u8a0a M i i t ,..., 1 , = x \u7d93\u904e\u5feb\u901f \u5229\u8449\u8f49\u63db\u5f8c\u5f97\u5230 \u7684 \u5728\u983b\u7387\u4e0a\u7684\u8a0a 1 ,..., 0 , ,..., 1 ), ( \u2212 = = K k M i k i t X \uff0c\u5176\u4e2d i \u8868\u793a \u7684 \u6578\uff0ck \u8868\u793a \u983b\u7387\u7684 \u6578\uff0ct \u8868\u793a\u97f3 \u7684 \u6578\u3002 \u7b2c\u4e8c\u90e8\u5206\uff0c\u6211\u5011\u8a08\u7b97\u4e0d\u540c \u97f3\u65b9\u5411\u89d2\u5ea6 180 ,..., 1 = \u7684 \u9593\u529f\u7387\u983b \u2211 \u2212 = = 1 0 ) , ( ) ( K k t t k P P \uff0c 180 , , 1 L = (4) \u5176\u4e2d 2 1 } cos ) 1 ( 2 exp{ ) ( ) , ( \u2211 = \u2212 = M i k i t t c d i f j k k P \u03c0 X",
+                        "eq_num": "(5)"
+                    }
+                ],
+                "section": "\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5(Speaker Localization Algorithm, SLA)",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "\u6642\u9593 \u3002\u5176\u57fa\u672c\u60f3\u6cd5\u662f\u5047\u8a2d\u7b2c i \u500b \u548c\u5176\u6240\u76f8\u9130\u7684\u7b2c i+1 \u500b \u5728\u7b2c L \u500b\u6578\u4f4d\u9ede\u5f8c \u7684\u8a9e\u97f3\u8a0a \u5206\u5225\u8868\u793a\u5982\u4e0b\uff1a i N L i L i L + + x x x ,..., , 1 \u548c 1 1 1 1",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5(Speaker Localization Algorithm, SLA)",
+                "sec_num": "2.2"
+            },
+            {
+                "text": ",..., , ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5(Speaker Localization Algorithm, SLA)",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "+ + + + + + + + i N L i L i L x x x \u5176\u4e2d\u6211\u5011\u53d6\u51fa N \u500b\u6578\u4f4d\u9ede\uff0c\u5982\u5716\u4e94\u6240\u793a\u3002 \u5728\u4e0d\u8003\u616e \u97f3\u4ee5\u53ca\u8a0a \u6e1b\u7684\u60c5\u5f62\u4e0b\uff0c\u82e5 \u70ba i \u548c i+1 \u9593\u7684\u6642\u9593 \uff0c\u5247 i t x \u548c 1 + + i t x \u4e4b\u9593\u5177\u6709\u6700\u5927\u7684\u76f8\u95dc\u6027\u4e14 \u2211 + = + + N L L t i t i t 1 x x \u9ede \u548c\u70ba\u6700\u5927\uff0c\u6b64\u4e00 \u548c\u53ef\u7a31\u4e4b\u70ba Time Domain Cross Correlation\u3002 \u5716\u4e94\u3001TDCC\u793a\u610f\u5716 \u7d93\u7531\u4ee5\u4e0a\u7684\u60f3\u6cd5\u6211\u5011\u767c\u5c55\u51fa TDCC \u7684\u6f14\u7b97\u6cd5\uff1a\u82e5\u73fe\u6709\u4e00 \u9663\u5217\u5305\u542b\u6709 M \u500b \uff0c i \u65bc\u6642\u9593 t \u6240 \u5230\u7684\u8a0a \u7a31\u70ba i t x \uff0c\u5247\u5c0d\u8a9e\u97f3\u8a0a \u4e2d\u4efb\u4e00\u97f3 m \u7684 TDCC \u5b9a\u7fa9\u5982\u4e0b \u2211\u2211 = = + \u2212 + \u2212 = M i N j i j p m j p m m C 2 1 ) 1 ( 1 ) 1 ( ) ( x x",
+                        "eq_num": "("
+                    }
+                ],
+                "section": "\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5(Speaker Localization Algorithm, SLA)",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "\u5176\u4e2d \u2211\u2211 = = \u2212 + + \u2212 + \u2212 = M i N j i i j p m j p m m C 2 1 ) 1 ( ) 1 ( 1 ) 1 ( ) , ( x x (9) \u9019\u88e1\u6211\u5011\u662f\u4ee5\u8a9e\u53e5\u4e2d\u80fd\u91cf\u6700\u9ad8\u7684\u97f3 \u70ba\u57fa\u6e96\u4f86\u8a08\u7b97\u6642\u9593 \uff0c\u53e6\u5916\u82e5\u5c07\u8a9e\u53e5 \u5168\u90e8\u97f3 \u7684 TDCC \u52a0\u8d77\u4f86\uff0c\u6839\u64da\u6b64 \u52a0\u503c\u5247\u76f8\u9130 \u9593\u7684\u6642\u9593 A \u02c6\u70ba \u2211 = m A ) C(m, max",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5(Speaker Localization Algorithm, SLA)",
+                "sec_num": "2.2"
+            },
+            {
+                "text": ") ( ) ( 2 / 1 2 / 1 2 / 1 ) 2 ( 1 ) , | ( s t s s t o o s n s s t e o P \u00b5 \u00b5 \u03c0 \u00b5 \u2212 \u03a3 \u2212 \u2212 \u2212 \u03a3 = \u03a3 (11) \u5176\u4e2d s \u00b5 \u8868\u793a\u5e73\u5747\u503c\u5411\u91cf\uff0c s \u03a3 \u8868\u793a\u8b8a\u7570\u6578\u77e9\u9663\uff0c t o \u70ba\u89c0 \u5230\u7684\u7279\u5fb5\u5411\u91cf\uff0cn \u70ba\u5411\u91cf\u7684 \u5ea6\u3002 \u5b9a\u7fa9\u4e00\u500b\u5927\u5c0f\u70ba ) 1 ( + \u00d7 n n \u7684\u8f49 \u77e9\u9663 s W \uff0c \u53ef\u5c07 \u5c55\u5f8c\u7684\u5e73\u5747\u503c\u5411\u91cf s \u8abf\u6574\u800c\u5f97\u5230\u65b0\u7684 \u5e73\u5747\u503c\u5411\u91cf s s s W \u00b5 = (12) \u5176\u4e2d = n s \u00b5 \u00b5 \u00b5 ,..., , , 2 1 \uff0c \u662f\u5728\u9032\u884c\u56de\u6b78\u8a08\u7b97\u6642\u8003\u616e\u662f\u5426\u4f7f\u7528 \u5dee\u91cf(\u4f7f\u7528\u5247 \u70ba 1\uff0c\u4e0d\u4f7f \u7528\u5247\u70ba 0)\u3002\u56e0\u6b64\u8abf\u6574\u904e\u5f8c\u7684\u9ad8\u65af\u6a5f\u7387\u5206\u4f48\u5982\u4e0b\u6240\u793a ) ( ) ( 2 / 1 2 / 1 2 / 1 ) 2 ( 1 ) , ,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5(Speaker Localization Algorithm, SLA)",
+                "sec_num": "2.2"
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Maximum Likelihood from Incomplete Data via the EM Algorithm",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [
+                            "P"
+                        ],
+                        "last": "Dempster",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "N",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Laird",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [
+                            "B"
+                        ],
+                        "last": "Rubin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1977,
+                "venue": "J. Roy. Stat. Soc",
+                "volume": "39",
+                "issue": "1",
+                "pages": "1--38",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "A. P. Dempster and N. M. Laird, D. B. Rubin. \"Maximum Likelihood from Incomplete Data via the EM Algorithm\", J. Roy. Stat. Soc., 39(1) : 1-38, 1977.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Maximum a Posterior Estimation for Multivariate Gaussian Mixture Observations of Markov Chains",
+                "authors": [
+                    {
+                        "first": "J.-L",
+                        "middle": [],
+                        "last": "Gauvain",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C.-H",
+                        "middle": [],
+                        "last": "Lee",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "IEEE Trans. Speech, Audio Processing",
+                "volume": "2",
+                "issue": "",
+                "pages": "291--298",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "J.-L. Gauvain and C.-H. Lee, \"Maximum a Posterior Estimation for Multivariate Gaussian Mixture Observations of Markov Chains\", IEEE Trans. Speech, Audio Processing, Volume 2, pages 291-298, April 1994.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Experiments of Speech Recognition In a Noisy and Reverberant Environment Using a Microphone Array and HMM Adaptation",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Giuliani",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Omologo",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Svaizer",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Proc. of ICSLP '96",
+                "volume": "",
+                "issue": "",
+                "pages": "1329--1332",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "D. Giuliani, M. Omologo and P. Svaizer, \"Experiments of Speech Recognition In a Noisy and Reverberant Environment Using a Microphone Array and HMM Adaptation\", In Proc. of ICSLP '96, pages 1329-1332, October 1996.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Microphone Array Design Measures for Hands-Free Speech Recognition",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Inoue",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Nakamura",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Yamada",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Shikano",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proc. of Eurospeech '97",
+                "volume": "1",
+                "issue": "",
+                "pages": "331--334",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "M. Inoue, S. NAKAMURA, T. YAMADA and K. SHIKANO, \"Microphone Array Design Measures for Hands-Free Speech Recognition\", In Proc. of Eurospeech '97, Volume 1, pages 331-334, September 1997.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Maximun Likelihood Linear Regression for Speaker Adaptation of Continuous Density Hidden Markov Models",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Leggetter",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Woodland",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Computer Speech and Language",
+                "volume": "9",
+                "issue": "",
+                "pages": "171--185",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C. J. Leggetter and P. C. Woodland, \"Maximun Likelihood Linear Regression for Speaker Adaptation of Continuous Density Hidden Markov Models\", Computer Speech and Language, Volume 9, pages 171-185, September 1995.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Combined Wiener and Coherence Filtering in Wavelet Domain For Microphone Array Speech Enhancement",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Mahmoudi",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proc. of ICASSP '98",
+                "volume": "",
+                "issue": "",
+                "pages": "385--388",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "D. Mahmoudi, \"Combined Wiener and Coherence Filtering in Wavelet Domain For Microphone Array Speech Enhancement\", In Proc. of ICASSP '98, pages 385-388, May 1998.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Acoustic Event Localization Using a Crosspower-Spectrum Phase Based Technique",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Omologo",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Svaizer",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "Proc. of ICASSP '94",
+                "volume": "2",
+                "issue": "",
+                "pages": "273--276",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "M. Omologo and P. Svaizer, \"Acoustic Event Localization Using a Crosspower-Spectrum Phase Based Technique\", In Proc. of ICASSP '94, Volume 2, pages 273-276, 1994.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Acoustic Source Location in Noisy and Reverberant Environment Using CSP Analysis",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Omologo",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Svaizer",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Proc. of ICASSP '96",
+                "volume": "",
+                "issue": "",
+                "pages": "921--924",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "M. Omologo and P. Svaizer, \"Acoustic Source Location in Noisy and Reverberant Environment Using CSP Analysis\", In Proc. of ICASSP '96, pages 921-924, 1996.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Robust Speech Recognition with Speaker Localization by a Microphone Array",
+                "authors": [
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Yamada",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Nakamura",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Shikano",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Proc. of ICSLP '96",
+                "volume": "",
+                "issue": "",
+                "pages": "1317--1320",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "T. YAMADA, S. Nakamura and K. Shikano, \"Robust Speech Recognition with Speaker Localization by a Microphone Array\", In Proc. of ICSLP '96, pages 1317-1320, October 1996.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Hands-Free Speech Recognition Based on a 3-D Viterbi Search Using a Microphone Array",
+                "authors": [
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Yamada",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Nakamura",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Shikano",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proc. of ICASSP '98",
+                "volume": "",
+                "issue": "",
+                "pages": "245--248",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "T. YAMADA, S. Nakamura and K. Shikano, \"Hands-Free Speech Recognition Based on a 3-D Viterbi Search Using a Microphone Array\", In Proc. of ICASSP '98, pages 245-248, May 1998a.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "An Effect of Adaptive Beamforming on 3-D Viterbi Search",
+                "authors": [
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Yamada",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Nakamura",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Shikano",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proc. of ICSLP '98",
+                "volume": "",
+                "issue": "",
+                "pages": "381--384",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "T. YAMADA, S. Nakamura and K. Shikano, \"An Effect of Adaptive Beamforming on 3-D Viterbi Search\", In Proc. of ICSLP '98, pages 381-384, December 1998b.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Simultaneous Recognition of Multiple Sound Sources Based on 3-D N-Best Search Using Microphone Array",
+                "authors": [
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Yamada",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Nakamura",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Shikano",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proc. of Eurospeech '99",
+                "volume": "1",
+                "issue": "",
+                "pages": "69--72",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "T. YAMADA, S. Nakamura and K. Shikano, \"Simultaneous Recognition of Multiple Sound Sources Based on 3-D N-Best Search Using Microphone Array\". In Proc. of Eurospeech '99, Volume 1, Page 69-72, September 1999.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF0": {
+                "content": "<table><tr><td colspan=\"4\">\u4e00\u822c\u7684\u8a9e\u97f3\u8fa8\u8a8d\u7686\u4f7f\u7528\u55ae\u4e00</td><td>\u505a\u70ba\u8a9e\u97f3\u8a0a \u7684\u8f38\u5165\uff0c\u5728</td><td>\u7684</td><td>\u4e0b\u5df2\u6709\u4e0d\u932f\u7684\u8fa8</td></tr><tr><td colspan=\"5\">\u8b58\u6210\u679c\uff0c\u7136\u800c\uff0c\u7576\u61c9\u7528\u5728 \u97f3\u5f88\u5927\u7684</td><td>\u88e1\uff0c\u8a9e\u97f3\u8fa8\u8b58\u7684\u6548\u679c\u5c07\u5927</td><td>\uff0c\u56e0\u6b64\uff0c\u5982\u4f55 \u5236</td></tr><tr><td colspan=\"4\">\u97f3\u4e26\u52a0 \u8a9e\u97f3\u8a0a \u5df2\u6210\u70ba</td><td>\u8a9e\u97f3\u8fa8\u8b58\u7684\u95dc \u6027\u6280\u8853\u3002\u56e0\u6b64\u672c\u8ad6\u6587\u4e2d\u6211\u5011\u4f7f\u7528\u4e00\u7d44\u9060\u8ddd</td></tr><tr><td colspan=\"5\">\u9663\u5217\u505a\u8a9e\u97f3\u8a0a \u8f38\u5165\uff0c\u7136\u5f8c\u4f7f\u7528 TDCC \u5c07\u6bcf\u500b</td><td>\u4e4b\u9593\u7684\u6642\u9593</td><td>\u8a08\u7b97\u51fa\u4f86\uff0c\u518d\u5229</td></tr><tr><td colspan=\"5\">\u5efa \u7528 Delay-and-Sum Beamformer \u7684\u65b9\u5f0f\uff0c\u5f97\u5230\u4e00\u7d44\u5177 \u7c21</td><td>\u6027\u4e14\u52a0 \u904e\u5f8c\u7684\u8a9e\u97f3\u8a0a \u3002\u70ba\u4e86\u4f7f\u52a0</td></tr><tr><td colspan=\"5\">\u570b\u7acb\u6210\u529f\u5927\u5b78\u8cc7\u8a0a\u5de5\u7a0b\u5b78\u7cfb \u904e\u7684\u8a9e\u97f3\u8a0a \u5728\u9032\u884c\u96b1\u85cf\u5f0f\u99ac\u53ef\u592b\u6a21\u578b(Hidden Markov Model, HMM)\u70ba\u4e3b\u8a9e\u97f3\u8fa8\u8b58\u6642\u6709\u66f4</td></tr><tr><td colspan=\"5\">Email\uff1ajtchien@mail.ncku.edu.tw \u4f73\u7684\u8fa8\u8b58\u6548\u679c\uff0c\u6211\u5011\u4f7f\u7528\u6700\u4f73\u76f8 \u5ea6\u7dda\u6027\u56de\u6b78\u7406\u8ad6(maximum likelihood linear regression,</td></tr><tr><td colspan=\"5\">MLLR) (Leggetter and Woodland, 1995)\u5c07\u539f \u8a9e\u97f3\u8a13\u7df4\u51fa\u4f86\u7684\u96b1\u85cf\u5f0f\u99ac\u53ef\u592b\u6a21\u578b\u53c3\u6578\u505a\u8abf</td></tr><tr><td>\u6574\uff0c\u4ee5\u88dc</td><td colspan=\"4\">\u6458\u8981 \u8a9e\u97f3\u8207\u6a21\u578b\u53c3\u6578\u4e4b\u9593\u7684\u4e0d \u914d\u3002</td></tr><tr><td colspan=\"5\">\u672c\u7bc7\u8ad6\u6587\u63d0\u51fa\u4e00\u7a2e\u53ef\u61c9\u7528\u65bc \u97f3 \u76ee\u524d \u7684\u5b78\u8853\u7814\u7a76\u6a5f\u69cb\u5c0d\u65bc\u67b6\u69cb\u65bc</td><td>\u4e0b</td><td>\u9663\u5217(Microphone Array)\u7684\u8a9e\u97f3\u8fa8\u8b58\u6f14\u7b97 \u9663\u5217\u4e0a\u7684\u8a9e\u97f3\u8655\u7406\u6280\u8853\u5c1a\u5c6c\u8d77\u6b65\u968e\u6bb5\uff0c\u5728\u4e2d</td></tr><tr><td colspan=\"5\">\u6cd5\uff0c\u5176\u4e3b\u8981\u7684\u76ee\u7684\u5728\u65bc \u6587\u8a9e\u97f3\u8fa8\u8b58\u4e0a\u7684\u61c9\u7528\u767c\u8868\u5728\u76f8\u95dc\u5b78\u8853\u6703 \u53ca \u50b3\u7d71\u96fb \u8a9e\u97f3\u8fa8\u8b58\u7cfb\u7d71\u9700\u8981\u4f7f\u7528\u8005\u982d \u6216 \u8ad6\u6587\u5c1a\u4e0d\u591a\u898b\uff0c\u7136\u800c\uff0c\u570b\u5916\u7684\u7814\u7a76\u6a5f\u69cb\u5247 \u7684\u4e0d\u65b9 \u3002</td></tr><tr><td colspan=\"5\">\u70ba\u4e86 \u9664\u9060\u8ddd \u5df2 \u5165\u6b64\u4e00 \u57df\uff0c\u4e26\u4e14\u7372\u5f97\u4e0d\u932f\u7684\u6210\u679c\uff0c\u6bd4\u8f03\u6709\u540d\u7684\u5305\u62ec\u4e09\u5927\u985e\u7684\u65b9\u6cd5\uff0c\u7b2c\u4e00\u985e\u662f\u8457\u91cd\u5728\u4e0d \u7684 \u97f3 \uff0c\u6211\u5011\u7684\u65b9\u6cd5\u662f\u5148\u5c07\u6bcf\u500b \u96c6\u5230\u7684\u8a9e\u97f3\uff0c\u5229\u7528\u8a9e\u97f3</td></tr><tr><td>\u5230 \u6bcf\u500b \u540c</td><td colspan=\"4\">\u89d2\u5ea6\u7684\u4e0d\u540c\uff0c\u4f7f\u7528 Time Domain Cross Correlation (TDCC)\u6f14\u7b97\u6cd5 \u51fa\u8a9e\u8005\u767c \u9593\u6642\u9593 \u7684\u8a08\u7b97\uff0c \u4e3b\u8981\u662f\u5229\u7528\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5\u4f86\u8a08\u7b97\u4e0d\u540c \u9593\u7684\u6642\u9593</td></tr><tr><td colspan=\"5\">\u97f3\u7684\u65b9\u5411\u53ca\u8a9e\u97f3\u5230 \u6bcf\u500b \u53ef \u7684\u6642\u9593</td><td>\uff0c\u518d\u61c9\u7528 Delay-and-Sum Beamformer \u9663\u5217\u8a0a \u8655</td></tr><tr><td colspan=\"5\">\u7406\u6280\u8853\u5c07\u8a9e\u97f3\u8a0a \u52a0 \uff0c\u6700\u5f8c\u6211\u5011\u518d\u5c07\u52a0 \u904e\u7684\u8a9e\u97f3\u8a0a \u548c\u8a9e\u97f3\u6a21\u578b\u53c3\u6578\u9593\u7684\u4e0d \u914d\u7528\u6700\u4f73</td></tr><tr><td colspan=\"5\">\u76f8 \u5ea6\u7dda\u6027\u56de\u6b78(MLLR)\u7684\u6a21\u578b\u8abf\u6574\u6f14\u7b97\u6cd5\u4f86 \u9593\u6642\u9593 \u4f75\u5165\u96b1\u85cf\u5f0f\u99ac\u53ef\u592b\u6a21\u578b\u7684\u53c3\u6578\uff0c \u7684\u89c0 \u662f \u3002\u5728 \u97f3</td><td>\u4e0b\u4f7f\u7528 \u50b3\u7d71\u8a9e\u97f3\u96b1\u85cf\u5f0f\u99ac\u53ef\u592b\u6a21\u578b\u7684\u53c3 \u9663\u5217\u4e4b\u9023\u7e8c\u6578\u5b57</td></tr><tr><td colspan=\"5\">\u8fa8\u8b58\u5be6\u9a57\u4e2d\uff0c\u6211\u5011\u63d0\u51fa\u4f86\u7684\u65b9\u6cd5\u5c0d\u65bc\u63d0\u5347\u8fa8\u8b58\u7387\u6709 \u597d\u7684\u6548\u679c\u3002 \u6578\uff0c\u52a0\u5165\u5404\u7a2e\u4e0d\u540c\u7684\u8a9e\u8005\u89d2\u5ea6\uff0c\u4e26\u4f7f\u7528\u4e00\u7a2e\u4e09 \u7684 \u7279\u6bd4\u6f14\u7b97\u6cd5(Three-Dimensional Viterbi</td></tr><tr><td colspan=\"2\">Search)\u4f5c\u8a9e\u97f3\u8fa8\u8b58</td><td/><td/></tr><tr><td colspan=\"2\">1. \u5c0e\u8ad6</td><td/><td/></tr><tr><td colspan=\"2\">\u73fe\u5be6\u751f</td><td>\u4e2d\uff0c</td><td colspan=\"2\">\u4e86\u5404\u5f0f\u5404\u6a23\u7684 \u97f3\u548c\u56de\u97f3\uff0c\u9019\u4e9b</td><td>\u6703 \u91cd\u7684 \u4f4e\u8a9e\u97f3\u8fa8\u8a8d\u7cfb\u7d71</td></tr><tr><td colspan=\"5\">\u7684\u6548\u80fd\uff0c\u5176\u4e2d\u4e4b\u4e00\u7684\u89e3\u6c7a\u65b9\u5f0f\u662f\u4f7f\u7528\u982d \u5f0f</td><td>(Head-Mounted Microphone)\uff0c\u4f7f\u5f97 \u97f3</td></tr><tr><td>\u548c</td><td colspan=\"3\">\u76e1\u53ef\u80fd\u7684 \u8fd1\uff0c\u4f86 \u4f4e</td><td>\u97f3\u548c\u56de\u97f3\u7684\u5f71\u97ff\u3002\u7136\u800c\u4f7f\u7528\u982d \u5f0f</td><td>\u8a2d \u6703\u9020\u6210</td></tr><tr><td colspan=\"5\">\u4f7f\u7528\u8005\u7684\u4e0d \uff0c\u56e0\u6b64\u5982\u4f55\u767c\u5c55\u4ee5\u514d \u5f0f</td><td>(Hands-Free Microphone)\u70ba\u4e3b\u7684\u8a9e\u97f3\u8fa8\u8a8d\u7cfb\u7d71\u5df2</td></tr><tr><td colspan=\"4\">\u6210\u70ba\u4e00\u500b\u76f8\u7576\u91cd\u8981\u7684\u7814\u7a76 \u984c\u3002</td></tr><tr><td colspan=\"2\">\u57fa\u672c\u4e0a\uff0c\u4f7f\u7528</td><td colspan=\"3\">\u9663\u5217\u53ef\u4ee5\u9032\u884c\u9060\u8ddd</td><td>\u97f3\uff0c\u56e0\u6b64\u53ef\u4ee5\u89e3\u6c7a\u982d \u5f0f</td><td>\u9020\u6210\u4f7f\u7528\u8005</td></tr><tr><td colspan=\"3\">\u4e0d \u7684\u554f\u984c\uff0c\u800c\u6211\u5011\u5e38\u7528\u7684</td><td/><td>\u9663\u5217\u8a0a \u8655\u7406\u6280\u8853\u662f \u7528 Delay-and-Sum Beamformer\uff0c</td></tr><tr><td>\u53ef\u4ee5</td><td colspan=\"4\">\u97f3\u548c\u56de\u97f3\u5c0d\u8a9e\u97f3\u8a0a \u7684\u5f71\u97ff\uff0c\u9084\u539f\u51fa</td><td>\u7684\u8a9e\u97f3\u3002\u800c\u4e14\u6b64\u4e00\u6280\u8853\u4e26\u975e\u91dd\u5c0d\u7279</td></tr><tr><td>\u5b9a \u97f3</td><td colspan=\"4\">\uff0c \u53ef\u9069\u7528\u65bc\u4efb\u4f55 \u97f3</td><td>\u4e0b\uff0c\u5f97\u5230 \u4eba \u610f\u7684\u6548\u679c\u3002\u5728\u672c\u8ad6\u6587\u4e2d\uff0c\u6211\u5011\u5c07</td></tr><tr><td colspan=\"2\">\u9663\u5217\u61c9\u7528\u65bc \u4f4e</td><td/><td colspan=\"2\">\u97f3\u7684</td><td>\uff0c\u4ee5 \u5230\u63d0\u9ad8\u8a9e\u97f3\u8fa8\u8a8d\u7387\u4e4b\u76ee\u7684\u3002</td></tr></table>",
+                "text": "",
+                "num": null,
+                "html": null,
+                "type_str": "table"
+            },
+            "TABREF6": {
+                "content": "<table><tr><td>\u591a\u901a \u97f3 SigC31-4 \u63a5\u982d\u7d93\u7531\u5de5\u4f5c\u96fb\u8def\u548c \u76f8\u9023\u63a5\uff0c\u900f\u904e\u6240\u9644\u7684 \u9ad4\u5373\u53ef\u5229\u7528 4 \u500b \u6211\u5011\u4f7f\u7528\u570b \u97f3 \u516c \u6240\u751f\u7522\u7684\u5168\u65b9\u5411\u96fb\u5bb9\u5f0f (Omni-directional Condenser \u540c\u6642 \u97f3\uff0c\u96fb\u5bb9\u5f0f Microphone)\uff0c\u578b \u70ba ECM9D\uff0c\u6240\u5c0d\u61c9\u7684\u983b \u70ba 20~10000Hz\uff0c \u5ea6\u70ba-38 3dB\uff0c\u8a0a \u6bd4 (signal-to-noise ratio, SNR)\u5927\u65bc 60dB\uff0c\u5de5\u4f5c\u96fb\u58d3\u5247\u4ecb\u65bc DC 3V \u81f3 DC 10V \u4e4b\u9593\u3002\u5de5\u4f5c\u96fb\u8def\u4e3b \u524d\u5f8c\u7684 \u97f3\u548c\u6578\u5b57\u9593\u7684 \u97f3\uff0c\u6240\u4ee5\u7e3d\u5171\u7684 \u6578\u76ee\u70ba 73 \u500b\u3002\u6bcf\u4e00\u500b \u5305\u542b 4 \u500b \u5408\u6578\uff0c \u56e0\u6b64\u5171\u6709 292 \u500b \u5408\u6578\u3002 \u8a9e\u6599\u662f\u5728\u5be6\u9a57 \u4e2d\u4f7f\u7528\u9060\u8ddd \u9663\u5217\u6240 \uff0c\u6211\u5011\u6a21 \u4e86\u4e09\u7a2e\u4e0d\u540c \u901f\u7684\u8def \u6cc1\uff0c\u5206\u5225\u70ba 0 km/h\u300150 km/h \u548c 90 km/h\u3002\u5728 0 km/h \u8def\u6cc1\u4e0b\u4e0d\u52a0\u4efb\u4f55 \u97f3\uff0c\u800c 50 km/h \u548c 90 km/h \u8def\u6cc1\u5247\u5229\u7528 \u653e\u51fa \u65bc\u6642\u901f 50 km/h \u548c 90 km/h \u6642\u6240 \u4e0b\u7684 \u97f3\u4f86\u6a21 \u3002 \u97f3\u6642\u8a9e\u8005\u8ddd \u4e2d\u5fc3 120 \u516c\u5206\u548c \u9663\u5217\u7684 \u89d2 \u70ba 60 \u5ea6\uff0c \u97f3 \u5247 \u653e\u65bc\u8ddd \u4e2d\u5fc3 80 \u516c\u5206\u8655\u548c \u7684\u7684 \u89d2 75 \u5ea6\uff0c \u9663\u5217\u662f\u7dda\u6027\u914d \u7684\uff0c\u76f8\u9130 \u7684\u9593\u8ddd\u70ba 10 \u516c Delay-and-Sum Beamformer \u8655\u7406\u5f8c\u7684\u8a9e\u97f3\u8a0a \u5206\u5225\u8a08\u7b97\u5176\u8fa8\u8a8d\u7d50\u679c\uff0c\u8fa8\u8a8d\u7d50\u679c\u5982\u8868\u4e8c\u6240\u793a\u3002 \u89c0\u5bdf\u5be6\u9a57\u7d50\u679c\u6211\u5011\u53ef\u4ee5\u767c\u73fe\u5728\u4e0d\u540c\u8def\u6cc1\u4e0b\u6700\u4f73\u8fa8\u8a8d\u7387\u90fd\u51fa\u73fe\u5728 60 \uff0c\u6b64\u4e00\u7d50\u679c\u548c\u6211\u5011\u5be6 \u4e0a \u8a9e\u97f3\u6642\u7684\u65b9\u5411\u662f\u5341\u5206 \u5408\u7684\u3002 Digit Error Rate (%) \u8def\u6cc1 0 km/h 50 km/h 90 km/h Mic 1 47.0 52.1 55.1 Mic 2 42.0 48.3 3.4 \u53d6\u6a23\u983b\u7387\u5c0dSLA\u548cTDCC\u7684\u5f71\u97ff \u7d93\u7531\u4ee5\u4e0a\u7684\u5be6\u9a57\u7d50\u679c\uff0c\u6211\u5011\u767c\u73fe\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5\u7684\u6548\u80fd\u8f03\u5dee\u3002 \u7814\u7a76\u5176\u539f\u56e0\u5f8c\u767c\u73fe\uff0c \u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5\u662f\u5148\u6c42\u51fa\u8a9e\u8005\u7684\u65b9\u5411 s \uff0c\u518d\u5229\u7528\u516c\u5f0f C d b s cos Rate) (Sampling Rate) Sampling ( = = 3.5 \u52a0\u5165\u8a9e\u97f3\u6a21\u578b\u8abf\u6574\u7684\u5be6\u9a57\u7d50\u679c \u63a5\u4e0b\u4f86\u7684\u5be6\u9a57\u8457\u91cd\u65bc \u89e3\u8a9e\u97f3\u6a21\u578b\u8abf\u6574\u5c0d \u9663\u5217\u8a9e\u97f3\u8fa8\u8a8d\u7cfb\u7d71\u7684\u5f71\u97ff\u3002\u57fa\u672c\u7cfb\u7d71\u7d93 \u518d\u5e73\u5747\u3002 \u884c \u7684\u96fb \u914d \u70ba B a s e l i n e S L A T D C C H T D C C A \u7531 Digit Error Rate (%) \u8def\u6cc1 0 km/h 50 km/h Without MLLR 0.28 0.58 0.41 0.56 90 km/h (19) Mic 1 + MLLR 29.0 31.7 33.4 With MLLR 0.50 0.79 0.63 0.77 53.4 Mic 3 51.0 54.8 58.7 \u4f86\u6c42\u51fa\u53d6\u6a23\u9ede\u4e0a\u7684\u4f4d b \u3002\u56e0\u70ba \u97f3\u7684\u901f\u5ea6 C \u548c \u7684\u9593\u8ddd d \u90fd\u662f\u56fa\u5b9a\u7684\uff0c\u56e0\u6b64\u6211\u5011\u8a2d\u8a08 Mic 2 + MLLR 25.5 29.9 33.1 \u8868\u4e5d\u3001SLA \u8207 TDCC \u884c\u901f\u5ea6\u4e4b\u6bd4\u8f03 (\u901f\u5ea6\u8a08\u7b97\u55ae\u4f4d\u70ba \u53e5)</td></tr><tr><td>\u8981\u7684\u4f5c\u7528\u662f\u5c07\u96fb \u4f9b\u61c9\u5668\u6240\u63d0\u4f9b\u7684\u96fb\u529b\u9032\u884c \u58d3\uff0c\u4e4b\u5f8c\u518d \u81f3 \u529b\uff0c\u4e26\u4fdd \u97f3\u6642\u8a0a \u7684 \u5b9a\uff0c\u4e26\u5c07\u8a0a \u81f3\u591a\u901a \u97f3 SigC31-4\u3002\u6b64\u5de5\u4f5c\u96fb\u8def\u662f \u63d0\u4f9b \u97f3\u6642\u6240\u9700\u7684\u96fb \u5206\u3002\u8a9e\u8005\u548c \u9663\u5217\u4ee5\u53ca \u97f3\u7684\u76f8\u5c0d\u4f4d \u5982\u5716\u4e03\u6240\u793a\u3002\u7e3d\u5171\u6709 15 \u4eba\u53c3\u8207 \u97f3\uff0c\u5305\u542b 12 \u4f4d Mic 4 46.5 53.0 57.2 Mic \u5e73\u5747 46.6 52.1 56.1 Mic 3 + MLLR 31.6 33.4 36.2 \u4e86\u4e00\u4e9b\u5be6\u9a57\u4f86 \u89e3\u53d6\u6a23\u983b\u7387\u5c0d SLA \u548c TDCC \u5169\u7a2e\u6f14\u7b97\u6cd5\u7684\u5f71\u97ff\u3002 Mic 4 + MLLR 28.3 31.4 34.5 \u751f\u548c 3 \u4f4d \u751f\uff0c\u6bcf\u4e00\u7a2e\u8def\u6cc1\u6709 30 \u53e5\u4e0d\u540c\u7684\u4e2d\u6587\u9023\u7e8c\u6578\u5b57\u3002\u6bcf\u7a2e\u8def\u6cc1\u7e3d\u5171 \u5f97 450 \u53e5\u97f3 \u3002 \u8868\u4e00\u3001\u57fa\u672c\u7cfb\u7d71\u7684\u8fa8\u8a8d\u7d50\u679c \u5be6\u9a57\u6642\u6211\u5011\u5148\u5c0d \u8a9e\u6599\u5229\u7528 \u5dee\u6cd5\u63d0\u9ad8\u53d6\u6a23\u983b\u7387\uff0c\u7d93\u7531 Delay-and-Sum Beamformer \u6c42 (Mic + MLLR)\u5e73\u5747 28.6 31.6 34.4 4. \u7d50\u8ad6</td></tr><tr><td>\u4f9d\u64da\u97f3 \u516c \u91dd\u5c0d\u96fb\u5bb9\u5f0f \u9023\u7e8c\u8a9e\u97f3\u8fa8\u8a8d\u6240\u4f7f\u7528\u7684\u6f14\u7b97\u6cd5\u70ba\u4e00\u968e\u6bb5\u6f14\u7b97\u6cd5\u3002\u5be6\u9a57\u7d50\u679c\u6211\u5011\u4ee5\u6578\u5b57\u932f\u8aa4\u7387(Digit Error Rate) \u7684\u5efa \u96fb\u8def \u52a0\u4fee\u6539\u5f8c\u7531\u6211\u5011\u81ea\u884c \u63a5 \u4f5c\u7684\u3002 \u4f86\u8868\u793a\u3002 \u51fa\u589e \u904e\u7684\u8a9e\u97f3\u8a0a \u5f8c\uff0c\u518d\u5c07\u53d6\u6a23\u983b\u7387 \u70ba 8KHz\u3002\u7136\u5f8c\u518d\u9032\u884c\u8fa8\u8b58\uff0c\u8fa8\u8a8d\u7d50\u679c\u5982\u8868\u56db(8KHz)\u3001 Mic \u5e73\u5747 46.6 52.1 56.1 \u672c\u8ad6\u6587\u4e2d\u6211\u5011\u5efa\u7acb\u4e00\u500b\u61c9\u7528 \u9663\u5217\u7684\u8a9e\u97f3\u8fa8\u8a8d\u7cfb\u7d71\uff0c\u6b64\u4e00\u7cfb\u7d71\u5229\u7528 Delay-and-Sum \u8def\u6cc1\uff0fDigit Error Rate (%)\uff0f\u89d2\u5ea6 30 60 90 120 150 0 km/h 35.2 31.9 60.6 58.6 55.4 \u8868\u4e03\u3001\u57fa\u672c\u7cfb\u7d71\u7d93\u7531MLLR\u8abf\u6574\u5f8c\u7684\u5be6\u9a57\u7d50\u679c \u8868\u4e94(16KHz)\u548c\u8868\u516d(24KHz)\u6240\u793a\u3002 Beamformer \u4f86 \u4f4e \u97f3\u5c0d\u65bc\u8a9e\u97f3\u8a0a \u7684\u5f71\u97ff\u3002\u540c\u6642\u6211\u5011\u4e5f\u63d0\u51fa\u4e86\u4e00\u500b\u61c9\u7528\u65bc \u9663\u5217</td></tr><tr><td>50 km/h \u6211\u5011\u53ef\u4ee5\u767c\u73fe SLA \u7684\u8fa8\u8a8d\u932f\u8aa4\u7387\u6709\u660e\u986f\u7684\u6539\u8b8a\uff0c\u53d6\u6a23\u983b\u7387\u7531 8KHz \u63d0\u9ad8\u81f3 16KHz \u548c 40.9 38.1 66.9 68.3 62.6 \u6f14\u7b97\u6cd5 Digit Error Rate (%) \u8def\u6cc1 0 km/h 50 km/h 90 km/h \u4e0a\u8a08\u7b97\u6642\u9593 \u7684\u6f14\u7b97\u6cd5 TDCC\u3002\u5be6\u9a57\u7684\u90e8\u5206\u6211\u5011\u9032\u884c\u4e86\u57fa\u672c\u7cfb\u7d71\u7684\u5be6\u9a57\u3001 \u5b9a\u5404\u7a2e\u4e0d\u540c\u89d2</td></tr><tr><td>90 km/h 24KHz \u6642\u932f\u8aa4\u7387\u5728 3 \u7a2e\u4e0d\u540c\u8def\u6cc1\u90fd\u6709\u986f\u8457\u7684\u4e0b \u3002\u800c\u63d0\u9ad8\u81f3 16KHz \u548c\u63d0\u9ad8\u81f3 24KHz \u76f8\u6bd4\u6642 42.8 40.4 70.0 69.6 65.7 (Mic + MLLR)\u5e73\u5747 28.6 31.6 34.4 \u5ea6\u7684\u5be6\u9a57\u3001\u53d6\u6a23\u983b\u7387\u6539\u8b8a\u7684\u5be6\u9a57\u3001\u4f7f\u7528 SLA \u6f14\u7b97\u6cd5\u3001\u4f7f\u7528\u6700\u5927\u80fd\u91cf\u97f3 \u7684 TDCC \u6f14\u7b97\u6cd5\u548c\u4f7f \u8868\u4e8c\u3001\u4e0d\u540c \u97f3 \u89d2\u5ea6\u4e0bDelay-and-Sum Beamformer\u7684\u8fa8\u8a8d\u7d50\u679c \u6f14\u7b97\u6cd5 Digit Error Rate (%) \u8def\u6cc1 0 km/h 50 km/h 90 km/h Mic \u5e73\u5747 46.7 52.1 56.1 \u56fa\u5b9a\u89d2\u5ea6 60 31.9 38.1 40.4 SLA 43.8 48.6 52.1 TDCC H 37.5 43.6 SLA + MLLR 29.9 31.5 35.1 \u5247\u5728 3 \u7a2e\u4e0d\u540c\u8def\u6cc1\u4e0a\u932f\u8aa4\u7387 \u6709\u4e9b\u8a31\u7684\u6539\u8b8a\u3002\u81f3\u65bc TDCC \u5247\u56e0\u70ba\u5176 \u7b97\u7684\u5c0d\u8c61\u5c31\u662f\u6642\u57df\u4e0a\u7684 \u53d6\u6a23\u9ede\uff0c\u56e0\u6b64\u8fa8\u8a8d\u932f\u8aa4\u7387\u4e26\u7121\u660e\u986f\u6539\u8b8a\u3002\u6b64\u4e00\u7d50\u679c\u986f\u793a\u7531\u65bc TDCC \u4e0d\u9700\u8981\u8a08\u7b97\u8a9e\u8005\u65b9\u5411\uff0c\u56e0 \u6b64\u53ef\u4ee5\u9069\u7528\u65bc\u5404\u7a2e\u53d6\u6a23\u983b\u7387\uff0c\u80fd \u4e00\u5b9a\u7684\u8fa8\u8a8d\u6548\u679c\u3002 \u6f14\u7b97\u6cd5 Digit Error Rate (%) \u8def\u6cc1 0 km/h 50 km/h TDCC H + MLLR 25.2 28.8 31.4 TDCC A + MLLR 21.1 24.9 28.2 \u8868\u516b\u3001SLA\u548cTDCC\u7d93\u7531MLLR\u8abf\u6574\u5f8c\u7684\u8fa8\u8a8d\u7d50\u679c\u6bd4\u8f03\u5716 \u7d93\u7531\u8868\u4e03\u7684\u5be6\u9a57\u7d50\u679c\u6211\u5011\u53ef\u4ee5\u767c\u73fe\uff0c\u57fa\u672c\u7cfb\u7d71\u4f7f\u7528 MLLR \u7684\u8a9e\u97f3\u6a21\u578b\u8abf\u6574\u6280\u8853\u5f8c\uff0c\u4e0d\u7ba1 \u7528\u5168\u90e8\u97f3 \u7684 \u672c\u8ad6\u6587\u4e2d\u4ea6\u7d50\u5408\u4e86\u8a9e\u97f3\u6a21\u578b\u8abf\u6574\u7684\u6280\u8853\u3002\u7d93\u7531\u5be6\u9a57\u6211\u5011\u53ef\u4ee5\u767c\u73fe\uff0c\u55ae\u7d14\u53ea\u4f7f\u7528 MLLR \u4f86 90 km/h 47.0 TDCC A 31.7 38.4 40.9 SLA H 43.79 48.64 52.12 TDCC H 37.50 43.58 47.03 \u8abf\u6574\u8a9e\u97f3\u6a21\u578b\u5373\u53ef\u7372\u5f97\u4e0d\u932f\u7684\u6548\u679c\u3002\u7136\u800c\u82e5\u5c07 \u9663\u5217\u548c\u8a9e\u97f3\u6a21\u578b\u8abf\u6574\u7684\u6280\u8853\u76f8\u7d50\u5408\uff0c\u5c0d \u5728 \u4e00\u7a2e\u8def\u6cc1\u90fd\u53ef\u6709\u6548\u7684 \u4f4e\u8fa8\u8a8d\u932f\u8aa4\u7387(0 \u65bc \u4e0b\u6240 \u65bc \u4f4e\u8fa8\u8a8d\u932f\u8aa4\u7387(\u5728\u4e0d\u540c\u8def\u6cc1\u4e0b\u5e73\u5747 \u53ef \u4f4e 25%\u7684\u8fa8\u8a8d\u932f\u8aa4\u7387)\u6703\u7522\u751f\u66f4\u986f\u8457\u7684\u6548\u679c\u3002\u5f9e \u8868\u4e09\u3001SLA\u548cTDCC\u8fa8\u8a8d\u7d50\u679c\u6bd4\u8f03\u5716 \u8868\u56db\u3001\u53d6\u6a23\u983b\u7387 8KHz \u6642 SLA \u548c TDCC \u7684\u8fa8\u8a8d\u7d50\u679c \u7684\u8a13\u7df4\u8a9e\u6599\u548c\u5229\u7528 \u9663\u5217\u65bc \u97f3 \u4e0b\u6240 \u7684 \u8a9e\u6599\u9593\u7684\u4e0d \u914d\u73fe\u8c61\u76f8\u7576 \u91cd\u3002 \u6211\u5011\u7814\u7a76\u7684\u7d50\u679c\uff0c\u53ef\u4ee5\u767c\u73fe\u4ecd\u7136\u9084\u6709\u8a31\u591a\u503c\u5f97\u7814\u7a76\u7684 \u984c\uff0c\u5982\u66f4 \u78ba\u8a9e\u8005\u65b9\u5411\u7684\u5b9a\u4f4d\u3001</td></tr><tr><td>\u7b2c\u4e8c\u7d44\u5be6\u9a57\u662f\u5c07\u6240 \u5f97\u7684 \u6f14\u7b97\u6cd5 Digit Error Rate (%) \u8def\u6cc1 \u8a9e\u6599\u5206\u5225\u7d93\u7531\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5\u548c TDCC \u6c42\u53d6\u4e0d\u540c 0 km/h 50 km/h 90 km/h \u5206\u6790\u8868\u516b\u7684\u7d50\u679c\uff0c\u6211\u5011\u767c\u73fe\u52a0\u5165\u8a9e\u97f3\u6a21\u578b\u8abf\u6574\u7684 SLA \u548c TDCC \u5728 \u4f4e\u8fa8\u8a8d\u932f\u8aa4\u7387\u4e0a\u4ea6\u6709 \u9593 \u4f4d \u7684\u8003\u91cf\u2026\u7b49\u3002\u672a\u4f86\u6211\u5011\u5c07\u4e3b\u8981\u81f4\u529b\u65bc\u7814\u7a76 \u9663\u5217\u4e2d \u7684 \u653e\u4f4d \u548c\u8fa8\u8a8d\u7387\u9593</td></tr><tr><td>\u5716\u4e03\u3001 \u97f3\u6642 \uff0c\u7d93\u904e\u88dc \u5f8c\u7522\u751f\u52a0 \u904e\u5f8c\u7684\u8a9e\u97f3\u8a0a \uff0c\u518d\u5206\u5225\u9032\u884c\u8fa8\u8a8d\uff0c\u5be6\u9a57\u7d50\u679c\u5982\u8868\u4e09\u6240\u793a\u3002 \u9663\u5217\u548c\u8a9e\u8005\u4ee5\u53ca \u97f3\u9593\u7684\u76f8\u5c0d\u4f4d \u5716 SLA H 34.94 39.97 42.00 \u986f\u8457\u7684\u6548\u679c\uff0c\u5728\u4e09\u7a2e\u4e0d\u540c\u8def\u6cc1\u4e0a TDCC \u7684\u6548\u80fd\u4ecd\u7136\u512a\u65bc SLA\u3002\u6700\u4f4e\u7684\u8fa8\u8a8d\u932f\u8aa4\u7387(21.10%)\u70ba \u7684\u6642\u9593 \u7684\u95dc\u4fc2\uff0c\u4ee5\u53ca\u5be6 \u5c07 \u9663\u5217\u7684\u6f14\u7b97\u6cd5\u61c9\u7528\u5728 \u6216\u6709\u56de\u97f3\u3001 \u97f3\u7684\u8a9e\u97f3\u8fa8\u8b58\u7cfb\u7d71\u4e0a\u3002</td></tr><tr><td>\u5716\u516d\u3001 3.3 Delay-and-Sum Beamformer\u5be6\u9a57\u7d50\u679c \u9663\u5217 \u97f3\u8a2d \u9023\u63a5\u5716 \u5176\u4e2d TDCC \u6709\u5169\u7a2e\u4e0d\u540c\u7684\u8a08\u7b97\u65b9\u6cd5\uff0cTDCC H \u8868\u793a\u6bcf\u4e00\u500b\u8a9e\u53e5 \u4f7f\u7528\u6700\u9ad8\u80fd\u91cf\u7684\u97f3 \u4f86\u8a08\u7b97 \u6642\u9593 \uff0c\u800c TDCC A \u5247\u8868\u793a\u6bcf\u4e00\u500b\u8a9e\u53e5\u7684\u6240\u6709\u97f3 \u7686\u88ab\u8003\u616e\u4f86\u8a08\u7b97\u6642\u9593 \u3002\u6b64\u5916\u8868\u4e09\u4ea6 TDCC H 36.90 42.69 47.09 \u8868\u4e94\u3001\u53d6\u6a23\u983b\u7387 16KHz \u6642 SLA \u548c TDCC \u7684\u8fa8\u8a8d\u7d50\u679c \u8def\u6cc1 0 km/h \u4e0b\u4f7f\u7528\u5168\u90e8\u97f3 \u4f86\u8a08\u7b97\u7684 TDCC \u6f14\u7b97\u6cd5\u3002</td></tr><tr><td>3. \u5be6\u9a57\u7d50\u679c 3.1 3.2 \u8a9e\u6599\u5eab \u5c0d\u65bc\u6bcf\u4e00\u500b \u5217\u51fa \u9663\u5217\u7684\u5e73\u5747\u8fa8\u8b58\u7387\u548c\u56fa\u5b9a\u89d2\u5ea6 60 \u6642\u7684\u8fa8\u8a8d\u932f\u8aa4\u7387\u4ee5\u65b9 \u6bd4\u8f03\u3002 (Mic1, Mic2, Mic3, Mic4)\u6240 \u96c6\u7684\u8a9e\u97f3\u8a0a \u5176\u500b\u5225\u7684\u5b57\u5143\u932f\u8aa4\u7387\u4ee5\u53ca \u6f14\u7b97\u6cd5 Digit Error Rate (%) \u8def\u6cc1 0 km/h 50 km/h 90 km/h 3.6 \u8fa8\u8a8d\u6642\u9593\u7684\u6bd4\u8f03 \u932f\u8aa4\u7387\u7684\u5e73\u5747\u503c\u5982\u8868\u4e00\u6240\u793a\uff0c\u6b64\u7d50\u679c\u53ef\u8996\u70ba\u57fa\u672c\u7cfb\u7d71(Baseline)\u7684\u932f\u8aa4\u7387\u3002\u5728\u6b64\u6211\u5011\u4f7f\u7528\u6240\u6709 \u89c0\u5bdf\u8868\u4e09\u7684\u8fa8\u8a8d\u7d50\u679c\u6bd4\u8f03\uff0c\u6211\u5011\u53ef\u4ee5\u767c\u73fe\u4f7f\u7528 Delay-and-Sum Beamformer \u7684 SLA \u548c SLA H 34.17 39.52 42.79 \u8fa8\u8a8d\u6642\u9593\u7684\u8a08\u7b97\u662f\u7d71\u8a08\u6240\u6709 \u8a9e\u6599(\u5171 1350 \u53e5\uff0c\u5e73\u5747\u4e00\u53e5\u5305\u542b 6 \u500b\u4e2d\u6587\u9023\u7e8c\u6578\u5b57)\u7d93\u7531 \u9663\u5217 \u97f3\u8a2d \u9663\u5217 \u97f3\u8a2d \u5982\u5716\u516d\u6240\u793a\uff0c\u4e3b\u8981\u7684\u90e8\u5206\u5305\u542b\u591a\u901a \u97f3 SigC31-4\u3001\u56db\u500b\u5168\u65b9\u5411 \u99ac\u53ef\u592b\u6a21\u578b\u4f86\u8868\u793a\uff0c\u6bcf\u4e00\u500b\u4e2d\u6587\u6578\u5b57\u4f7f\u7528 7 \u500b \uff0c \u97f3\u5247\u4f7f\u7528 3 \u500b \uff0c\u5206\u5225\u8868\u793a\u97f3 \u8fa8\u8a8d\u932f\u8aa4\u7387\u7684\u5e73\u5747\u503c\u4f5c\u70ba \u9663\u5217\u7684\u6574\u9ad4\u932f\u8aa4\u7387\u3002 TDCC \u78ba\u5be6\u80fd \u6709\u6548 \u4f4e\u932f\u8aa4\u7387\uff0c\u800c TDCC \u548c\u50b3\u7d71\u7684\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5 SLA \u76f8\u6bd4\uff0cTDCC \u66f4\u80fd TDCC H 38.27 44.08 49.07 \u8868\u516d\u3001\u53d6\u6a23\u983b\u7387 24KHz \u6642 SLA \u548c TDCC \u7684\u8fa8\u8a8d\u7d50\u679c \u6642\u9593 \u7684\u8a08\u7b97\u3001Delay-and-Sum Beamformer \u7684\u8655\u7406\u3001\u8a9e\u97f3\u7279\u5fb5\u53c3\u6578\u7684\u6c42\u53d6\u548c\u8a9e\u97f3\u8fa8\u8a8d\u7684\u6240 \u6211\u5011\u6240\u9032\u884c\u7684\u7b2c\u4e00\u7d44\u5be6\u9a57\u662f\u4e8b\u5148\u9810\u8a2d\u4e00\u4e9b \u97f3 \u89d2\u5ea6\u503c\u4f86\u9032\u884c\u5be6\u9a57\u4ee5 \u51fa\u6700\u4f73\u6548\u679c\u7684\u89d2 \u5ea6 \uff0c \u9019 \u88e1 \u6211 \u5011 \u5c07 \u97f3 \u7684 \u65b9 \u5411 \u56fa \u5b9a \u5f9e 30 \u5230 150 \u6bcf \u9593 \u9694 30 \u505a \u4e00 \u6b21 \u5be6 \u9a57 \uff0c \u518d \u5c0d \u7d93 \u7531 \u6709\u6548 \u4f4e\u8fa8\u8a8d\u932f\u8aa4\u7387\uff0c\u4e14 TDCC \u4f7f\u7528\u6240\u6709\u7684\u97f3 \u7684\u6548\u679c\u4e0d\u4f46\u6bd4\u4f7f\u7528\u4e00\u500b\u97f3 \u6548\u679c\u9084\u597d\uff0c\u800c\u4e14 \u6709\u6642\u9593\u518d\u505a\u5e73\u5747\u800c\u5f97\uff0c\u5be6\u9a57\u7d50\u679c\u5982\u8868\u4e5d\u6240\u793a\u3002\u57fa\u672c\u7cfb\u7d71\u5247 \u8a08\u7b97\u7279\u5fb5\u53c3\u6578\u548c\u8a9e\u97f3\u8fa8\u8a8d\u7684\u6642\u9593</td></tr></table>",
+                "text": "\u662f\u7531 \u570b Signalogic \u516c \u6240\u751f\u7522\u7684\uff0c\u70ba\u4e00 4 \u500b\u901a \u7684 \u97f3 \uff0c\u4f7f \u7528\u7684\u6578\u4f4d\u8a0a \u8655\u7406 \u7247(DSP processor)\u70ba \u5668\u516c (TI)\u6240\u751f\u7522\u7684 TM8320C31\uff0c\u53ef\u540c\u6642\u63d0\u4f9b 4 \u500b\u901a \u9032\u884c \u97f3\u7684\u52d5\u4f5c\uff0c\u6b64 \u97f3 \u7684\u4ecb\u9762\u70ba ISA \u4ecb\u9762\u53ef \u65bc\u500b\u4eba\u96fb \u4e0a\uff0c\u4e26\u6709\u63d0\u4f9b D37 \u578b MLLR \u8abf\u6574\u5f8c\u7684\u5be6\u9a57\u7d50\u679c\u5982\u8868\u4e03\u6240\u793a\u3002\u6b64\u5916\uff0cSLA \u548c TDCC \u5206\u5225\u52a0\u4e0a MLLR \u7684\u5be6\u9a57\u7d50\u679c \u5982\u8868\u516b\u6240\u793a\u3002 km/h \u7531 46.64% \u81f3 28.61%\uff0c50 km/h \u7531 52.07% \u81f3 31.60%\uff0c90 km/h \u7531 56.12% \u81f3 34.42%)\uff0c\u5176\u539f\u56e0\u70ba\u4f7f\u7528\u50b3\u7d71 Pentium II 350 \u8655\u7406\u5668\u548c 128MB \u8a18 \u9ad4\u7684\u500b\u4eba\u96fb \uff0c\u4f5c\u696d\u7cfb \u7d71\u5247\u70ba Windows 98\u3002\u89c0\u5bdf\u5be6\u9a57\u7d50\u679c\uff0c\u6211\u5011\u767c\u73fe\u4e0d\u7ba1\u662f \u4f7f\u7528\u6700\u5927\u80fd\u91cf\u7684\u97f3 \u6216\u662f\u5168\u90e8\u97f3 \u7684 TDCC \u6f14\u7b97\u6cd5\u5728 \u884c\u901f\u5ea6\u4e0a\u7686\u512a\u65bc\u50b3\u7d71\u7684 SLA \u6f14\u7b97\u6cd5\u3002 TDCC \u6f14\u7b97\u6cd5\u4ee5\u53ca \u884c\u901f\u5ea6\u6bd4\u8f03\u3002\u7d93\u7531\u5be6\u9a57\u7d50\u679c\u6211\u5011\u53ef\u4ee5\u8b49\u660e TDCC \u7684\u6709\u6548\u6027 (\u5728\u4e0d\u540c\u8def\u6cc1\u4e0b\u5e73\u5747 \u53ef \u4f4e 15%\u7684\u8fa8\u8a8d\u932f\u8aa4\u7387)\u3002\u548c\u50b3\u7d71\u7684\u8a9e\u8005\u5b9a\u4f4d\u6f14\u7b97\u6cd5 SLA \u76f8\u6bd4\u8f03\uff0c TDCC \u4e0d\u8ad6\u662f\u5728\u8fa8\u8a8d\u932f\u8aa4\u7387 \u4f4e\u7684 \u5ea6\u4e0a\u6216 \u884c\u901f\u5ea6\u4e0a\u7686\u512a\u65bc SLA\u3002",
+                "num": null,
+                "html": null,
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-1012.json

@@ -0,0 +1,389 @@
+{
+    "paper_id": "O00-1012",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:11.606804Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-1012",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "]\uff0c\u900f\u904e\u524d\u8a5e\u9810\u6e2c\u5f8c\u8a5e(Word Prediction) [ Koester and Levine, 1994; Hunnicut, 1990 ]\u3001\u82f1\u6587\u8a5e\u5f59\u7c21\u7a31(Abbreviation Expansion) [ Vanderheiden, 1984] \u3001 \u8a9e\u610f\u7de8\u78bc(Semantic Coding, e.g. Consumer Product-Minspeak System) [ Chang, 1992; Baker, 1982 ]\uff0c\u5176\u4e3b\u8981\u76ee\u7684\u4e43\u8f14\u52a9\u5927\u91cf\u8a5e\u5f59\u5eab\u4e2d\u9078\u53d6\u7279\u5b9a\u8a5e\u5f59\uff0c\u7136\u800c\u5176\u8a2d\u8a08\u4e4b\u6839\u672c\u4ee5\u8a5e\u5f59\u8a9e\u610f\u70ba\u57fa \u790e\uff0c\u4f7f\u7528\u8005\u5fc5\u9808\u8a18\u61b6\u5927\u91cf\u7684\u8a5e\u5f59\u53ca\u5176\u5c0d\u61c9\u7b26\u865f\uff0c\u9020\u6210\u8a8d\u77e5\u4e0a\u7684\u8ca0\u64d4\uff1b2.)\u7531\u5317\u7f8e\u5fa9\u5065\u5de5\u7a0b\u5354 \u6703 (RESNA) \u6240\u63d0\u51fa\u4e4b Sentence Compansion (compress/expansion) \u7684\u89c0\u5ff5 [ Demasco, et. al., 1989; Demasco and McCoy, 1992] ",
+                "cite_spans": [
+                    {
+                        "start": 28,
+                        "end": 55,
+                        "text": "[ Koester and Levine, 1994;",
+                        "ref_id": "BIBREF4"
+                    },
+                    {
+                        "start": 56,
+                        "end": 70,
+                        "text": "Hunnicut, 1990",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 104,
+                        "end": 125,
+                        "text": "[ Vanderheiden, 1984]",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 189,
+                        "end": 203,
+                        "text": "[ Chang, 1992;",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 204,
+                        "end": 215,
+                        "text": "Baker, 1982",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 357,
+                        "end": 382,
+                        "text": "[ Demasco, et. al., 1989;",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 383,
+                        "end": 407,
+                        "text": "Demasco and McCoy, 1992]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "N-Gram \u8a5e\u6027\u6a21\u7d44\u4e43\u4ee5\u8a5e\u6027(POS)\u70ba\u57fa\u672c\u55ae\u4f4d(\u4f9d\u4e2d\u7814\u9662\u8a5e\u6027\u5206\u985e\u5b9a\u7fa9) \uff0c\u6bcf\u4e00\u500b\u5167\u90e8\u7bc0\u9ede \u7686\u4ee3\u8868\u4e00\u500b\u8a5e\u6027\u6a23\u672c\uff0c\u7b2c n \u500b\u7bc0\u9ede\u51fa\u73fe\u7684\u6a5f\u7387\u53ef\u7528\u4e0b\u5217\u5f0f\u5b50\u4f86\u63cf\u8ff0\uff1a ) ( ) ( ) | ( 1 1 1 1 1 \u2212 \u2212 = n n n n POS POS P POS POS P POS POS POS P L L L (1) \u5176\u4e2d\uff0cC(POS 1 \u2026POS n )\u4ee3\u8868 POS 1 \u2026POS n \u8a5e\u6027\u4e32\u5217\u51fa\u73fe\u7684\u6b21\u6578\u3002\u5176\u8655\u7406\u6d41\u7a0b\u5716\u4e09\u6240\u793a \u69cb \u53e5 \u958b \u59cb \u53e5 \u578b \u9810 \u6e2c \u6a21 \u7d44 V a r ia b le N -G r a m \u8a5e \u5f59 \u64f7 \u53d6 \u69cb \u53e5 \u7d50 \u675f \u8a5e \u5f59 \u8cc7 \u6599 \u5eab \u9ad8 \u983b \u8a5e \u5f59 \u5e8f \u5217 \u662f \u5426 \u66f4 \u65b0 \u8a5e \u983b \u6587 \u6cd5 \u66f4 \u6b63 \u66f4 \u65b0 \u53e5 \u578b \u8cc7 \u6599 \u5eab \u6587 \u53e5 \u751f \u6210 \u8a5e \u5f59 \u9078 \u53d6 \u5716\u4e8c \u53e5\u578b\u9810\u6e2c\u6a21\u7d44\u8655\u7406\u6d41\u7a0b\u5716 \u8209\u4f8b\u4f86\u8aaa\uff0c\u7cfb\u7d71\u4f9d\u64da\u4f7f\u7528\u8005\u4e4b\u524d\u7684\u69cb\u53e5-\u300e\u6211\u53bb\u53f0\u5317\u958b\u6703\u300f \u3001 \u300e\u6211\u53bb\u9ad8\u96c4\u73a9\u300f \u3001 \u300e\u6211\u559c\u6b61\u82b1\u300f \u3001 \u300e\u6211\u559c\u6b61\u5403\u51b0\u6dc7\u6dcb\u300f \uff0c\u5efa\u7acb\u4e86\u4e00\u500b Variable N-grams \u8a5e\u6027\u6a21\u7d44\uff0c\u5982\u5716\u4e09\u6240\u793a\uff1a Root Nh VCL Nc VA VC \u6211 \u53bb \u53f0\u5317 \u958b\u6703 \u73a9 VK VC Na \u559c\u6b61 \u5403 \u51b0\u6dc7\u6dcb Na \u82b1 \u9ad8\u96c4 P(.|Nh)=1 P(.|Nh VCL)=0.5 P(.|Nh VK)=0.5 P(.|Nh VK VC)=0.5 P(.|Nh VK VC Na)=1 P(.|Nh VK Na)=0.5 P(.|Nh VCL Nc)=1 P(.|Nh VCL Nc VC)=0.5 P(.|Nh VCL Nc VA)=0.5 \u5716\u4e09 Variable N-grams \u8a5e\u6027\u6a21\u7d44 3-1-3. \u6ce8\u97f3\u7e2e\u5beb\u8f14\u52a9\u67e5\u8a62 \u63a1\u985e\u4f3c\u82f1\u6587\u7e2e\u5beb\u65b9\u5f0f\uff0c\u900f\u904e\u5b8c\u6574\u6ce8\u97f3\u6216\u9023\u7e8c\u555f\u59cb\u97f3\u8f38\u5165\uff0c\u5e6b\u52a9\u5177\u6ce8\u97f3\u8a8d\u77e5\u80fd\u529b\u7684\u4f7f\u7528\u8005 \u5f9e\u9f90\u5927\u7684\u624b\u8a9e\u8cc7\u6599\u5eab\u4e2d\u64f7\u53d6\u6240\u9700\u7684\u8a5e\u5b57\u5f59\u3002\u672c\u7814\u7a76\u4e43\u6839\u64da\u4e2d\u6587\u6587\u53e5\u8207\u97f3\u97fb\u7279\u6027\uff0c\u904b\u7528 Augmented Transition Network \u8a2d\u8a08\u67b6\u69cb(\u5716\u56db)\uff0c\u4ee5\u64f7\u53d6\u5c0d\u61c9\u7684\u5b57/\u8a5e\u3002\u5efa\u69cb\u898f\u5247\u70ba\uff1a1).\u9023\u7e8c \u5169\u500b\u540c\u985e\u97f3\u76f8\u63a5\uff0c\u70ba\u8a5e\uff1b2).First Vowel \u63a5 Consonant\uff0c\u70ba\u8a5e\uff1b3).Second Vowel \u63a5 Consonant \u6216 First Vowel\uff0c\u70ba\u8a5e\uff1b4).\u7121\u8072\u8abf(1 \u8072) \uff0c\u53ef\u80fd\u70ba\u8a5e\uff1b5).\u5305\u542b 2~5 \u8072\u8abf\uff0c\u5fc5\u70ba\u5b57\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "III. \u5916\u90e8\u7bc0\u9ede(external node\u3001terminal node) \uff0c\u70ba\u53e5\u578b\u7d50\u675f\u4f4d\u7f6e\uff0c\u5305\u542b 1.)\u53c3\u8003\u6b21\u6578\uff0c\u8a18\u9304 \u53e5\u578b\u6a23\u7248\u51fa\u73fe\u6b21\u6578\uff1b2.)\u7bc0\u9ede\u500b\u6578\uff0c\u53e5\u578b\u7684\u7bc0\u9ede\u500b\u6578\uff1b3.)\u540d\u8a5e\u500b\u6578\uff0c\u53e5\u578b\u7684\u540d\u8a5e\u7e3d\u6578\uff1b 4.)\u52d5\u8a5e\u500b\u6578\uff0c\u53e5\u578b\u7684\u52d5\u8a5e\u7e3d\u6578\uff1b5.)\u865b\u8a5e\u500b\u6578\uff0c\u53e5\u578b\u7684\u865b\u8a5e\u7e3d\u6578\uff1b6.)\u552f\u4e00\u7684\u7236\u7bc0\u9ede\u3002 \u53e5\u578b\u6a23\u7248\u6a39\u4e4b\u5efa\u69cb\u6d41\u7a0b\uff0c\u5982\u5716\u4e94\u6240\u793a\uff0c\u5176\u5efa\u69cb\u539f\u5247\u70ba\uff1a 1. \u7bc0\u9ede\u5339\u914d\u5617\u8a66\uff1a1).\u82e5\u5c6c\u6027\u7b26\u5408\uff0c\u5247\u5339\u914d\u4e4b\uff1b2).\u82e5\u8a5e\u6027\u7b26\u5408\uff0c\u5339\u914d\u4e4b\uff1b3).\u7bc0\u9ede\u5339\u914d \u8005\uff0c\u82e5\u8a5e\u5f59\u5df2\u5b58\u5728\u65bc\u7bc0\u9ede\u4e4b\u4e2d\uff0c\u5247\u8a5e\u5f59\u53c3\u8003\u6b21\u6578\u52a0 1\uff1b\u5426\u5247\uff0c\u5c07\u65b0\u8a5e\u5f59\u52a0\u5165\u7bc0\u9ede \u4e4b\u4e2d\uff0c\u65b0\u8a5e\u5f59\u53c3\u8003\u6b21\u6578\u8a2d\u70ba 1\u3002 2. \u5ee3\u5ea6\u512a\u5148\u641c\u5c0b\u53e5\u578b\u662f\u5426\u5b58\u5728\uff1a1).\u82e5\u53e5\u578b\u5b58\u5728\uff0c\u53e5\u578b\u6a23\u7248\u51fa\u73fe\u6b21\u6578\u52a0 1\uff1b2).\u82e5\u53e5\u578b \u4e0d\u5b58\u5728\uff0c\u5247\u7e7c\u7e8c\u5ef6\u4f38\u8def\u5f91\u76f4\u5230\u53e5\u578b\u751f\u6210\u70ba\u6b62\uff1b3).\u53e5\u578b\u751f\u6210\u5f8c\uff0c\u7522\u751f\u5916\u90e8\u7bc0\u9ede\uff0c\u53e5 \u578b\u6a23\u7248\u51fa\u73fe\u6b21\u6578\u8a2d\u70ba 1\u3002\u76ee\u524d\u672c\u7814\u7a76\u6240\u5efa\u7acb\u4e4b\u53e5\u578b\u6a23\u7248\u6a39\u5305\u542b 553 \u689d\u53e5\u578b\u8def\u5f91\uff0c \u5e73\u5747\u8def\u5f91\u9577\u5ea6 3.5 \u500b\u7bc0\u9ede\u3002 \u4e2d \u7814 \u9662 \u81ea \u52d5 \u65b7 \u8a5e \u7cfb \u7d71 \u8a5e \u5f59 / \u8a5e \u6027 / \u53e5 \u578b \u4fee \u6b63 \u65b7 \u8a5e \u7d50 \u679c \u53ca \u8a5e \u6027 \u6a19 \u8a18 H o w -N e t \u641c \u5c0b \u53e5 \u578b \u662f \u5426 \u5b58 \u5728 \u7bc0 \u9ede \u5339 \u914d \u7b56 \u7565 \u53e5 \u578b \u5b58 \u5728 \u53e5 \u578b \u5ef6 \u4f38 \u898f \u5247 \u53e5 \u578b \u6a23 \u7248 \u6a39 \u5ef6 \u4f38 \u905e \u8ff4 \u5ef6 \u4f38 \u7d50 \u675f \uff1f \u5c0d \u8a71 \u8a9e \u6599 \u5426 \u662f \u662f \u5426 \u5ef6 \u4f38 \u4e4b \u53e5 \u578b \u9084 \u6709 \u8a9e \u6599 \u672a \u8655 \u7406 \u7d50 \u675f \u5426 \u662f \u5716\u4e94 \u53e5\u578b\u6a23\u7248\u6a39\u5efa\u69cb\u6d41\u7a0b\u5716 3-2-3. \u6587\u53e5\u751f\u6210\u6a5f\u5236\u4e4b\u5efa\u7acb \u4ee5\u53e5\u578b\u6a23\u7248\u6a39\u70ba\u4f9d\u64da\uff0c\u7d93\u7531\u4ee5\u4e0b\u6b65\u9a5f\uff1a1.)\u91dd\u5c0d\u4f7f\u7528\u8005\u8f38\u5165\u4e4b\u95dc\u9375\u8a5e\u4e32\uff0c\u6839\u64da\u5256\u6790\u898f\u5247 \u505a\u7247\u8a9e\u5408\u4f75\u53ca\u55ae\u4f4d\u8a5e\u5d4c\u5165\u4e4b\u8655\u7406\uff0c\u627e\u51fa\u95dc\u9375\u8a5e\u4e32\u5c0d\u61c9\u7684\u7247\u8a9e\uff0c\u4f5c\u70ba\u7bc0\u9ede\u5339\u914d\u4e4b\u57fa\u672c\u55ae\u4f4d\uff1b2.) \u4f9d\u64da\u8a9e\u610f\u5206\u6790\u53ca\u7bc0\u9ede\u5339\u914d\u7b56\u7565\u5c07\u95dc\u9375\u8a5e\u6240\u7d44\u6210\u7684\u7247\u8a9e\u586b\u5165\u53e5\u578b\u6a23\u7248\u7684\u7bc0\u9ede\u4e4b\u4e2d\uff1b\u63a5\u8457\u4f9d\u7167\u7bc0 \u9ede\u5c6c\u6027\u5339\u914d\u7a0b\u5ea6\u4ee5\u53ca\u53e5\u578b\u6a23\u7248\u53c3\u8003\u6b21\u6578\u4e4b\u7d71\u8a08\u8cc7\u6599\uff0c\u7be9\u9078\u51fa\u5408\u9069\u7684\u53e5\u578b\u6a23\u7248\uff1b3.)\u53c3\u7167\u53e5\u578b \u6a23\u7248\uff0c\u900f\u904e\u6642\u9593\u3001\u5730\u65b9\u7247\u8a9e\u4e4b\u6a23\u7248\u5d4c\u5165\u4ee5\u53ca Variable N-Gram \u865b\u8a5e\u88dc\u7db4\u7b49\u8655\u7406\uff0c\u751f\u6210\u5408\u4e4e\u8a9e\u6cd5 \u53ca\u8a9e\u610f\u7684\u81ea\u7136\u5b8c\u6574\u8a9e\u53e5\u3002 \u7bc0\u9ede\u5339\u914d\u539f\u5247 \u57fa\u65bc\u964d\u4f4e\u8a9e\u6cd5\u8907\u96dc\u5ea6\u53ca\u7e2e\u6e1b\u641c\u5c0b\u7a7a\u9593\u4e4b\u8003\u91cf\uff0c\u672c\u7814\u7a76\u900f\u904e\u7247\u8a9e\u5408\u4f75\u53ca \u55ae\u4f4d\u8a5e\u5d4c\u5165\u898f\u5247\uff0c\u5148\u627e\u51fa\u95dc\u9375\u8a5e\u4e32\u5c0d\u61c9\u7684\u7247\u8a9e\uff0c\u7bc0\u9ede\u5339\u914d\u55ae\u4f4d\u3002\u8655\u7406\u539f\u5247\u70ba\uff1a1",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Implementing Augmentative and Alternative Communication: strategies for learners with severe disabilities",
+                "authors": [
+                    {
+                        "first": "Joe",
+                        "middle": [],
+                        "last": "Reichle",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Reichle, Joe, \"Implementing Augmentative and Alternative Communication: strategies for learners with severe disabilities.\", Paul H.Bookes Publishing Co.,1992.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Adaptive One-Switch Row-Column Scanning",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Simpson",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "H",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Koester",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "IEEE Transaction on Rehabilitation Engineering",
+                "volume": "7",
+                "issue": "4",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Simpson, R. C. and Koester, H. H., \" Adaptive One-Switch Row-Column Scanning\", IEEE Transaction on Rehabilitation Engineering, Vol. 7, No. 4., 1999.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Electronic Devices for Rehabilitation",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [
+                            "G"
+                        ],
+                        "last": "Webster",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1985,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Webster, J.G., et. al., \" Electronic Devices for Rehabilitation\", John Wiley & Sons, Inc., 1985.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Augmentative and Aternative Communication",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [
+                            "B"
+                        ],
+                        "last": "David",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Pat",
+                        "middle": [],
+                        "last": "Mirendan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "David, R. B. and Mirendan, Pat, \"Augmentative and Aternative Communication\", Paul H. Bookes Publishing Co., 1992.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Modeling the Speed of Text Entry with a Word Prediction Interface",
+                "authors": [
+                    {
+                        "first": "H",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Koester",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "P"
+                        ],
+                        "last": "Levine",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "IEEE Transaction on Rehabilitation Engineering",
+                "volume": "2",
+                "issue": "3",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Koester, H. H. and Levine, S. P., \"Modeling the Speed of Text Entry with a Word Prediction Interface\", IEEE Transaction on Rehabilitation Engineering, Vol. 2, No. 3., 1994.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Word Prediction: Exploring the Use of Semantic information",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Hunnicut",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Augmentative and Alternative Communication",
+                "volume": "6",
+                "issue": "2",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Hunnicut, S., \"Word Prediction: Exploring the Use of Semantic information\", Augmentative and Alternative Communication, Vol. 6, No. 2., 1990.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "A High-Efficiency Flexible Keyboard Input Acceleration Technigues: SPEEDKEY",
+                "authors": [
+                    {
+                        "first": "G",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Vanderheiden",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1984,
+                "venue": "Proceedings of the Second International Conference on Rehabilitation Engineering, RESNA",
+                "volume": "",
+                "issue": "",
+                "pages": "353--354",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Vanderheiden, G. C., \"A High-Efficiency Flexible Keyboard Input Acceleration Technigues: SPEEDKEY\", Proceedings of the Second International Conference on Rehabilitation Engineering, RESNA, pp. 353-354., 1984.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "A Methodology for Iconic Language Design with Application to",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chang, S. K., et. al., \" A Methodology for Iconic Language Design with Application to",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Augmentative Communication",
+                "authors": [],
+                "year": 1992,
+                "venue": "Preceedings of the 1992 IEEE Workshop on Visual Language",
+                "volume": "",
+                "issue": "",
+                "pages": "110--116",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Augmentative Communication\", Preceedings of the 1992 IEEE Workshop on Visual Language, pp110-116., 1992.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Towards More Intelligent AAC Interfaces: The Use of Natural Language Processing",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Demasco",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1989,
+                "venue": "The12 th Annual Conference of RESNA",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Demasco, P., et. al., \"Towards More Intelligent AAC Interfaces: The Use of Natural Language Processing\", The12 th Annual Conference of RESNA, 1989.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Generating Text from Compressed Input: An Intelligent Interface for Prople with Severe Motor Impairements",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Demasco",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Mccoy",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "Communication of the ACM",
+                "volume": "35",
+                "issue": "5",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Demasco, P. and McCoy, K. F., \"Generating Text from Compressed Input: An Intelligent Interface for Prople with Severe Motor Impairements\", Communication of the ACM, Vol. 35, No. 5., 1992. \u8a9e\u8a00\u5b78\u7814\u8a0e\u6703\uff0cpp.61-83\uff0c\u6c11\u570b 85 \u5e74\u3002 \u5433\u5b97\u61b2\u3001\u9673\u662d\u5b8f\u3001\u6797\u8d85\u7fa4\uff0c\"\u4e2d\u6587\u6587\u53e5\u7ffb\u8a9e\u97f3\u7cfb\u7d71\u4e2d\u9023\u97f3\u8655\u7406\u4e4b\u7814\u7a76\"\uff0c\u7b2c\u4e5d\u5c46\u8a08\u7b97\u8a9e\u8a00\u5b78 \u7814\u8a0e\u6703\uff0cpp.85-104\uff0c\u6c11\u570b 85 \u5e74\u3002",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF1": {
+                "uris": null,
+                "text": "Compansion \u53ca Bottom-Up Parsing/Top-Down filtering \u7684\u89c0\u5ff5\uff1b\u4ee5\u4e2d\u7814\u9662\u81ea\u52d5\u65b7\u8a5e\u53ca\u77e5\u7db2(HowNet)\u77e5\u8b58\u70ba\u57fa\u790e\uff0c\u53e5\u578b\u6a23\u7248\u53ca\u6982\u5ff5\u5f9e\u5c6c\u4e4b\u8a9e\u683c\u6587\u6cd5 \u70ba\u69cb\u53e5\u57fa\u672c\u67b6\u69cb\uff0c\u900f\u904e\u8a9e\u6cd5\u5256\u6790\u3001\u8a9e\u610f\u5206\u6790\u4ee5\u53ca Variable N-grams \u88dc\u8a5e\u7b49\u81ea\u7136\u8a9e\u8a00\u8655\u7406\u6280\u8853\uff0c \u5c07\u4f7f\u7528\u8005\u7d93\u7531\u624b\u8a9e\u9375\u76e4\u6a21\u7d44\u8f38\u5165\u7684\u95dc\u9375\u8a5e\u5f59\u8f49\u70ba\u81ea\u7136\u5b8c\u6574\u7684\u8a9e\u53e5\u3002 3\u53e5\u578b\u6a23\u7248\u6a39\u4ee5\u5be6\u969b\u6536\u96c6\u5f59\u6574\u4e4b 1000 \u53e5\u65e5\u5e38\u751f\u6d3b\u5c0d\u8a71\u8a9e\u6599\u5eab\u70ba\u57fa\u790e(\u5e73\u5747 4.9 \u5b57/\u53e5) \uff0c\u5229 \u7528\u4e2d\u7814\u9662\u81ea\u52d5\u65b7\u8a5e\u7a0b\u5f0f\u53d6\u5f97\u8a9e\u53e5\u4e4b\u65b7\u8a5e\u7d50\u679c\u3001\u8a5e\u6027(POS)\u53ca\u8a9e\u6cd5\u8cc7\u8a0a\uff0c\u900f\u904e\u77e5\u7db2(How-Net) \u53d6\u5f97\u8a5e\u5f59\u4e4b\u6982\u5ff5\u5f9e\u5c6c\u5c6c\u6027\u5f8c\uff0c\u63a5\u8457\u7d93\u7531\u89c0\u5bdf\u5c0d\u8a71\u4f7f\u7528\u7fd2\u6163\u53ca\u8a55\u4f30\u8a9e\u6cd5\u5ef6\u4f38\u4e4b\u5408\u7406\u6027\uff0c\u6b78\u7d0d\u53e5 \u578b\u5ef6\u4f38\u898f\u5247\uff0c\u905e\u8ff4\u5efa\u7acb\u7d50\u5408\u8a9e\u53e5\u3001\u8a9e\u6cd5\u53ca\u8a9e\u610f\u8cc7\u8a0a\u7684\u53e5\u578b\u6a23\u7248\u6a39\uff0c\u4ee5\u63d0\u4f9b\u65e5\u5f8c\u53e5\u578b\u6bd4\u5c0d\u4ee5\u53ca \u865b\u8a5e\u88dc\u7db4\u4e4b\u4f9d\u64da\uff0c\u4ee5 Nh + Nd \u2192 Nd(\u7701\u7565 Nh)\u70ba\u4f8b\u3002\u900f\u904e\u53e5\u578b\u5ef6\u4f38\u53ca\u7d50\u69cb\u5316\u8655\u7406\u5c0d\u8a71\u8a9e\u6599 \u7684\u65b9\u5f0f\uff0c\u4e0d\u50c5\u53ef\u4ee5\u5f4c\u88dc\u5c0d\u8a71\u53e5\u578b\u4e0d\u8db3\u4e4b\u7f3a\u5931\uff0c\u66f4\u53ef\u7e2e\u6e1b\u5c0d\u8a71\u8a9e\u6599\u8cc7\u6599\u5eab\u7684\u5132\u5b58\u7a7a\u9593\u3001\u52d5\u614b\u66f4 \u65b0/\u5ef6\u4f38\u5c0d\u8a71\u53e5\u578b\u53ca\u52a0\u5feb\u53e5\u578b\u641c\u5c0b\u901f\u5ea6\u3002 \u53e5\u578b\u6a23\u7248\u6a39\u4e4b\u7bc0\u9ede\u8cc7\u6599\u7d50\u69cb I. \u8d77\u59cb\u7bc0\u9ede(root) \uff0c\u70ba\u53e5\u578b\u8d77\u59cb\u4f4d\u7f6e\uff0c\u6240\u6709\u53e5\u578b\u6a23\u7248\u7686\u7531\u6b64\u958b\u59cb\u5ef6\u4f38\uff0c\u6709\u591a\u7d44\u5b50\u7bc0\u9ede\u3002 II. \u5167\u90e8\u7bc0\u9ede(internal node) \uff0c\u5305\u542b 1.)\u5c6c\u6027\uff0c\u540c\u5c6c\u6027\u7684\u8a5e\u5f59\u53ef\u540c\u6642\u5b58\u5728\u65bc\u540c\u4e00\u7bc0\u9ede\uff1b2.)\u8a5e \u6027\uff0c\u540c\u8a5e\u6027\u7684\u8a5e\u5f59\u53ef\u540c\u6642\u5b58\u5728\u65bc\u540c\u4e00\u7bc0\u9ede\uff1b3.)\u53c3\u8003\u6b21\u6578\uff0c\u8a18\u9304\u7bc0\u9ede\u5c0d\u61c9\u7684\u53c3\u8003\u6b21\u6578\uff1b 4.)\u8a5e\u5f59\u7d44\uff0c\u7bc0\u9ede\u5167\u53ef\u540c\u6642\u5b58\u5728\u591a\u7d44\u8a5e\u5f59\uff1b5.)\u8a5e\u5f59\u53c3\u8003\u6b21\u6578\uff0c\u7bc0\u9ede\u5167\u6240\u6709\u8a5e\u5f59\u7686\u6709\u5176\u5c0d \u61c9\u7684\u53c3\u8003\u6b21\u6578\uff1b6.)\u7236\u7bc0\u9ede\uff0c\u53ea\u6709\u552f\u4e00\u7684\u7236\u7bc0\u9ede\uff1b7.)\u5b50\u7bc0\u9ede\uff0c\u53ef\u6709\u591a\u7d44\u5b50\u7bc0\u9ede\u3002",
+                "num": null,
+                "type_str": "figure"
+            },
+            "TABREF0": {
+                "content": "<table><tr><td colspan=\"2\">Fax\uff1a+886-6-274-7076</td></tr><tr><td>\u6458</td><td>\u8981</td></tr><tr><td colspan=\"2\">\u8072\u97f3\u6216\u8a9e\u8a00\u6a5f\u80fd\u55aa\u5931\u7684\u807d\u8a9e\u969c\u7919\u8005\uff0c\u5e38\u5e38\u767c\u751f\u96e3\u4ee5\u8207\u4e00\u822c\u4eba\u6b63\u5e38\u6e9d\u901a\u6216\u6e9d\u901a\u6642\u767c\u751f\u660e\u986f</td></tr><tr><td colspan=\"2\">\u7684\u969c\u7919\u3002\u672c\u7814\u7a76\u4e43\u8003\u91cf\u672c\u571f\u807d\u8a9e\u969c\u7919\u65cf\u7fa4\u5be6\u969b\u6e9d\u901a\u8f14\u52a9\u7684\u9700\u6c42\uff0c\u7814\u767c\u7b26\u5408\u672c\u571f\u5316 PC-based \u53f0</td></tr><tr><td colspan=\"2\">\u7063\u624b\u8a9e\u8f49\u8a9e\u97f3\u6e9d\u901a\u8f14\u52a9\u7cfb\u7d71\uff0c\u5305\u62ec 1).\u624b\u8a9e\u9375\u76e4\uff0c\u4f9d\u64da Row-Column Scanning \u53ca\u8003\u91cf\u8a8d\u77e5\u3001\u6ce8</td></tr><tr><td colspan=\"2\">\u610f\u96c6\u4e2d\u53ca\u5b78\u7fd2\u53cd\u61c9\u4e4b\u968e\u5c64\u5f0f\u5b89\u7f6e(Hierarchical Arrangement)\u7684\u7b56\u7565\uff0c\u4ee5\u4f5c\u70ba\u64cd\u4f5c\u8f38\u5165\u5a92\u4ecb\uff1b</td></tr><tr><td colspan=\"2\">2).\u904b\u7528\u8a5e\u983b\u9810\u6e2c\u3001\u8a5e\u6027\u7be9\u9078\u3001\u53e5\u578b\u9810\u6e2c\u53ca\u6ce8\u97f3\u7e2e\u5beb\u6a21\u7d44\u7b49\u6a5f\u5236\u4f86\u8f14\u52a9\u95dc\u9375\u8a5e\u5f59\u8f38\u5165\u8207\u624b\u8a9e\u7b26</td></tr><tr><td colspan=\"2\">\u865f\u641c\u5c0b\uff1b3).\u7d50\u5408\u53e5\u578b\u6a23\u7248\u53ca\u6982\u5ff5\u5f9e\u5c6c\u4e4b\u8a9e\u683c\u6587\u6cd5\u4f86\u5efa\u7acb\u95dc\u9375\u8a5e\u5f59\u9810\u6e2c\u5b8c\u6574\u6587\u53e5\u4e4b\u8f49\u8b6f\u7cfb\u7d71\u3002</td></tr><tr><td colspan=\"2\">\u5728\u7cfb\u7d71\u529f\u80fd\u6027\u8a55\u4f30\u90e8\u5206\uff0c\u7531\u7279\u6559\u8001\u5e2b\u9078\u53d6\u65e5\u5e38\u751f\u6d3b 1000 \u53e5\u5c0d\u8a71\u8a9e\u6599(\u5e73\u5747\u9577\u5ea6\u70ba 4.9 \u5b57/\u53e5)\u3002</td></tr><tr><td colspan=\"2\">\u514d\u9664\u865b\u8a5e\u8f38\u5165\u53ef\u7bc0\u7701 26.25%\u6309\u9375\u6578\uff1b\u52a0\u5165\u8a5e\u5f59\u3001\u53e5\u578b\u9810\u6e2c\u53ca\u6ce8\u97f3\u7e2e\u5beb\u7b49\u8f14\u52a9\u69cb\u53e5\u65b9\u5f0f\uff0c\u8207\u672a</td></tr><tr><td colspan=\"2\">\u52a0\u4efb\u4f55\u9810\u6e2c\u529f\u80fd\u4e4b\u6aa2\u7d22\u901f\u5ea6\u6539\u5584\u7387\u5206\u5225\u70ba 67.71%\u300179.50%\u300196.87%\u3002\u5728\u9069\u7528\u6027\u8a55\u4f30\u90e8\u5206\uff0c\u7d93</td></tr><tr><td colspan=\"2\">\u7531\u6559\u5b78\u3001\u8abf\u9069\u53ca\u8a55\u4f30\u6642\u671f\u7684\u8a13\u7df4\uff0c\u69cb\u53e5\u6210\u529f\u7387\u5206\u5225\u70ba 47.37%\u300165.0%\u300168.38%\uff1b\u69cb\u53e5\u901f\u5ea6\u8207</td></tr><tr><td colspan=\"2\">\u4e3b\u89c0\u6eff\u610f\u5ea6\u8a55\u91cf\u4ea6\u6709\u986f\u8457\u6539\u5584\u3002\u56e0\u6b64\uff0c\u672a\u4f86\u53ef\u63d0\u4f9b\u7b26\u5408\u672c\u571f\u6240\u9700\u4e4b\u8f14\u52a9\u624b\u8a9e\u8a13\u7df4\u8207\u6559\u5b78\u7cfb\u7d71\u3002</td></tr><tr><td colspan=\"2\">\u95dc\u9375\u5b57\uff1a\u807d\u8a9e\u969c\u7919\u3001\u6e9d\u901a\u8f14\u52a9\u7cfb\u7d71\u3001\u624b\u8a9e\u9375\u76e4\u3001\u95dc\u9375\u8a5e\u5f59\u9810\u6e2c\u3001\u53e5\u578b\u6a23\u677f</td></tr><tr><td>1. \u7dd2\u8ad6</td><td/></tr><tr><td colspan=\"2\">\u807d\u8a9e\u969c\u7919\u6307\u8072\u97f3\u6216\u8a9e\u8a00\u529f\u80fd\u6027\u7684\u640d\u50b7\uff0c\u56e0\u800c\u9020\u6210\u96e3\u4ee5\u8207\u4e00\u822c\u4eba\u6b63\u5e38\u6e9d\u901a\uff0c\u6216\u6e9d\u901a\u6642\u767c\u751f</td></tr><tr><td colspan=\"2\">\u660e\u986f\u7684\u969c\u7919\uff0c\u4e14\u7531\u65bc\u6e9d\u901a\u969c\u7919\u8005\u5728\u6e9d\u901a\u554f\u984c\u4e0a\u5448\u73fe\u5f88\u5927\u7684\u500b\u5225\u5dee\u7570\uff0c\u6709\u6642\u53c8\u96e3\u4ee5\u9451\u5225\uff0c\u5728\u5be6</td></tr><tr><td colspan=\"2\">\u969b\u751f\u6d3b\u4e2d\uff0c\u96e3\u4ee5\u4f7f\u7528\u807d\u8a9e\u529f\u80fd\u8868\u9054\u57fa\u672c\u751f\u7406\u9700\u6c42\uff1b\u5728\u6c42\u5b78\u968e\u6bb5\u4e2d\uff0c\u4e5f\u9020\u6210\u8a31\u591a\u7684\u5b78\u7fd2\u969c\u7919\u3002</td></tr><tr><td colspan=\"2\">\u6b50\u7f8e\u5148\u9032\u570b\u5bb6\u5728 1970 \u5e74\u4ee3\u958b\u59cb\u7814\u767c\u53ef\u63d0\u4f9b\u8a9e\u8a00\u5b78\u7fd2\u53ca\u6e9d\u901a\u66ff\u4ee3\u6b98\u969c\u8f14\u52a9\u5fa9\u5065\u79d1\u6280\u8207\u8f14\u5177</td></tr><tr><td colspan=\"2\">(Augmentative and Alternative Communication, AAC)\uff0c\u4e3b\u8981\u767c\u5c55\u8207\u6539\u826f\u7c21\u6613\u578b\u6e9d\u901a\u677f\u3001\u96fb\u8166\u64cd\u4f5c\u8f38</td></tr><tr><td colspan=\"2\">\u5165\u4ecb\u9762\u53ca\u8f14\u52a9\u6027\u5468\u908a\u88dd\u7f6e\u30021980 \u5e74\u4ee3\uff0c\u7531\u65bc\u96fb\u8166\u3001\u8a9e\u97f3\u8a0a\u865f\u8655\u7406\u53ca\u6b98\u969c\u8f14\u52a9\u79d1\u6280\u7684\u767c\u5c55\uff0c\u5168\u529b</td></tr><tr><td colspan=\"2\">\u6574\u5408\u5de5\u7a0b\u3001\u5fa9\u5065\u3001\u91ab\u7642\u53ca\u6559\u80b2\u8a13\u7df4\u4f86\u6539\u5584\u807d\u8a9e\u969c\u7919\u8005\u65e5\u5e38\u751f\u6d3b\u7684\u529f\u80fd\u6027\u30021990 \u5e74\u4ee3\uff0c\u5247\u8457\u91cd\u61c9</td></tr><tr><td>\u7528\u5148\u524d\u8f14\u52a9\u79d1\u6280\u8207\u8f14\u5177\u63d0\u4f9b\u4e4b\u7d93\u9a57\u65bc\u6559\u80b2\u8a13\u7df4\u8207\u81e8\u5e8a\u904b\u7528</td><td/></tr></table>",
+                "text": "\u90b1\u6bd3\u8ce2\u3001\u5433\u5b97\u61b2\u3001\u90ed\u555f\u7965\u3001*\u937e\u9ad8\u57fa\u570b\u7acb\u6210\u529f\u5927\u5b78\u8cc7\u8a0a\u5de5\u7a0b\u7814\u7a76\u6240\u3001*\u91ab\u5b78\u5de5\u7a0b\u7814\u7a76\u6240 Email\uff1ap7888107@ccmail.ncku.edu.tw, chwu@csie.ncku.edu.tw",
+                "html": null,
+                "type_str": "table",
+                "num": null
+            },
+            "TABREF4": {
+                "content": "<table><tr><td colspan=\"4\">\u5982\uff1a\u751c/\u860b\u679c\u21d2\u751c\u7684\u860b\u679c\uff1b2).\u55ae\u4f4d\u8a5e\u5d4c\u5165\uff1a\u53f0\u7063\u624b\u8a9e\u4e26\u7121\u55ae\u4f4d\u8a5e\uff0c\u56e0\u6b64\uff0c\u63a1\u7528\u55ae\u4f4d\u8a5e\u5c0d node(ai), token(bj)</td></tr><tr><td colspan=\"4\">\u61c9\u8868\uff0c\u4e26\u642d\u914d\u8a9e\u6cd5\u898f\u5247\uff0c\u5c07\u55ae\u4f4d\u8a5e\u5d4c\u5165\u540d\u8a5e\u7247\u8a9e\u4e4b\u4e2d\uff0c\u5982\uff1a\u4e8c/\u978b\u5b50\u21d2\u5169\u96d9\u978b\u5b50\u3002 \u53e5\u578b\u6a23\u7248\u5339\u914d \u7531\u65bc\u53f0\u7063\u624b\u8a9e\u6587\u6cd5\u8f38\u5165\u8a5e\u5e8f\u7684\u554f\u984c\uff0c\u672c\u6587\u4ee5\u95dc\u9375\u8a5e\u5408\u4f75\u5f8c\u7684\u7247\u8a9e\u7d44 n f b a SN tternScore SentencePa n i j i  *  \uf8f7 \uf8f8 \uf8f6 \uf8ec \uf8ed \uf8eb = \u2211 =1 , ) Yes ( (2) POS(ai) =POS(bj) or attribute(ai)=attribute(bj) SN(ai,bj)=1.0</td></tr><tr><td colspan=\"4\">No \u70ba\u7bc0\u9ede\u5339\u914d\u55ae\u4f4d\uff0c\u914d\u5408\u7bc0\u9ede\u5339\u914d\u7b56\u7565\u5c07\u95dc\u9375\u7247\u8a9e\u7d44\u586b\u5165\u53e5\u578b\u6a23\u7248\u4e4b\u4e2d\uff1b\u63a5\u8457\u53c3\u7167\u7bc0\u9ede\u5c6c \u5176\u4e2d a i \u70ba\u7b2c i \u500b\u7bc0\u9ede\uff1bb j \u70ba\u8207\u7bc0\u9ede a i \u5339\u914d\u4e4b\u95dc\u9375\u8a5e\u5f59\uff1bn \u70ba\u53e5\u578b\u6a23\u7248\u7bc0\u9ede\u500b\u6578\uff1bf\uff1a\u53e5 Yes</td></tr><tr><td colspan=\"4\">\u6027\u5339\u914d\u7a0b\u5ea6\u4ee5\u53ca\u53e5\u578b\u6a23\u7248\u53c3\u8003\u6b21\u6578\u4e4b\u7d71\u8a08\u8cc7\u6599\uff0c\u7be9\u9078\u51fa\u5408\u9069\u7684\u53e5\u578b\u6a23\u7248\uff0c\u4ee5\u63d0\u4f9b\u81ea\u7136\u8a9e \u578b\u6a23\u7248\u7684\u51fa\u73fe\u6a5f\u7387(\u5373\u53e5\u578b\u6a23\u7248\u7d42\u7aef\u7bc0\u9ede\u4e4b\u53c3\u8003\u6b21\u6578) \uff1bSN(a i, b j )\u70ba\u7bc0\u9ede\u76f8\u4f3c\u5ea6\uff0c\u5176 POS(ai) is Functional Word SN(ai,bj)=0.9</td></tr><tr><td colspan=\"4\">\u53e5\u751f\u6210\u4e4b\u4f9d\u64da\u3002\u5176\u8655\u7406\u7b56\u7565\u70ba\uff1a1).\u7bc0\u9ede\u5339\u914d\u7b56\u7565\uff1a\u900f\u904e\u5c6c\u6027\u6216\u8a5e\u6027\u6bd4\u5c0d\uff0c\u5c07\u6240\u6709\u672a\u5339 \u914d\u5206\u503c\u70ba\uff1a No</td></tr><tr><td colspan=\"4\">\u914d\u7684\u95dc\u9375\u7247\u8a9e\u8207\u53e5\u578b\u6a23\u7248\u7684\u7bc0\u9ede\u4f5c\u4e00\u5339\u914d\u5617\u8a66\u3002\u672c\u7814\u7a76\u4ee5\u4e2d\u7814\u9662\u7c21\u5316\u7684 46 \u985e\u8a5e\u985e\u4f5c\u70ba \u7bc0\u9ede\u4e4b\u8a5e\u6027\u6a19\u8a18\uff0c\u5176\u52d5\u8a5e\u8a5e\u985e\u5206\u70ba 12 \u985e(VA-VL) \uff0c\u5176\u8a73\u76e1\u7684\u52d5\u8a5e\u5206\u985e\u56fa\u7136\u6709\u52a9\u65bc\u964d \u4f4e\u8a9e\u6cd5\u7684\u4e0d\u660e\u78ba\u6027\uff0c\u537b\u4e5f\u9020\u6210\u53e5\u578b\u4fb7\u9650\u4ee5\u53ca\u53e5\u578b\u5339\u914d\u4e0d\u6613\u4e4b\u7f3a\u5931\u3002\u4f8b\u5982\uff0c\u82e5\u55ae\u7d14\u4ee5\u8a5e\u6027 \u70ba\u5339\u914d\u4f9d\u64da\uff0c\u5247\u300e\u611b(VL) \u300f\u8207\u300e\u559c\u6b61(VK) \u300f\u8a5e\u6027\u4e0d\u540c\uff1b\u82e5\u653e\u5bec\u52d5\u8a5e\u8a5e\u6027\u9650\u5236\uff0c\u5176\u5169 \uf8f4 \uf8f3 75 . 0 Yes + 0.25 otherwise \uf8f4 \uf8f4 \uf8f4 \uf8f1 = = b attribute a attribute or b pos a pos if j i j i ) ( ) ( ) ( ) ( POS(ai) &amp; POS(bj) is Verb 1 \uf8f4 \uf8f2 \u2260 = b pos clsss a pos class if word functional is a pos if b a SN j i i j i ) ( _ ) ( _ 25 . 0 ) ( 9 . 0 ) ( , Yes + 0.25 (3) No \uf8f4 \u2260 b pos state a pos state if j i ) ( _ ) ( _ 5 . 0 class_POS(ai) = class_POS(bj) SN(ai,bj)=0.25</td></tr><tr><td colspan=\"4\">\u8005\u7686\u5c6c\u65bc\u72c0\u614b\u53ca\u7269\u985e\u52d5\u8a5e\u3002\u56e0\u6b64\uff0c\u672c\u7814\u7a76\u5c07\u7bc0\u9ede\u5339\u914d\u689d\u4ef6\u9069\u5ea6\u653e\u5bec\uff0c\u82e5\u5c6c\u6027\u76f8\u7b26\uff0c\u5247\u5339 \u6027\u5206\u985e\u7684\u9650\u5236\u7a0b\u5ea6\u800c\u5b9a) \uff0c\u52d5\u8a5e\u8a5e\u6027\u76f8\u4f3c\u5ea6\u4f9d\u95dc\u9375\u8a5e\u8207\u7bc0\u9ede\u5c6c\u6027\u95dc\u806f\u7a0b\u5ea6\u800c\u7570\uff0c\u9054\u5230\u03b1 1 \u914d\u4e4b\uff1b\u5426\u5247\uff0c\u82e5\u540d\u8a5e\u8a5e\u6027\u76f8\u7b26\uff0c\u5339\u914d\u4e4b\uff1b\u82e5\u52d5\u8a5e\u8a5e\u6027\u76f8\u4f3c\uff0c\u5339\u914d\u4e4b\u30022).\u53e5\u578b\u6a23\u7248\u5339\u914d\uff1a \u7bc0\u9ede\u76f8\u4f3c\u5ea6\u4fc2\u6839\u64da 1.)\u4e2d\u7814\u9662\u52d5\u8a5e\u7d50\u69cb(\u5982\u5716\u516d\u6240\u793a\uff0c\u5176\u53c3\u6578\u503c\u4e43\u53c3\u7167\u8a5e\u6027\u968e\u5c64\u5f0f\u5c6c No state_POS(ai) = state_POS(bj) SN(ai,bj)=0.5</td></tr><tr><td colspan=\"4\">Yes + 0.25 \u4ee5\u53e5\u578b\u6a23\u7248\u6a39\u70ba\u4f9d\u64da\uff0c\u914d\u5408\u7bc0\u9ede\u5339\u914d\u7b56\u7565\u53ca\u6bd4\u5c0d\u7a0b\u5e8f\uff0c\u5c07\u95dc\u9375\u7247\u8a9e\u7d44\u586b\u5165\u5408\u9069\u7684\u53e5\u578b\u6a23 SN(ai,bj)=0.75</td></tr><tr><td colspan=\"4\">\u7248\u3002\u57fa\u65bc\u865b\u8a5e\u88dc\u7db4\u4e4b\u9700\u6c42\u53ca\u7e2e\u6e1b\u641c\u5c0b\u7a7a\u9593\u4e4b\u8003\u91cf\uff0c\u5176\u689d\u4ef6\u9650\u5236\u70ba\uff1a\u6a23\u7248\u52d5\u8a5e\u500b\u6578\u7b49\u65bc\u95dc \u4e0d \u53ca \u5716\u4e03 \u7bc0\u9ede\u76f8\u4f3c\u5ea6\u6bd4\u5c0d\u6d41\u7a0b\u5716 VA \u9375\u7247\u8a9e\u52d5\u8a5e\u500b\u6578\u3001\u6a23\u7248\u540d\u8a5e\u500b\u6578\u5c0f\u65bc\u95dc\u9375\u7247\u8a9e\u540d\u8a5e\u500b\u6578\u3001\u6a23\u7248\u865b\u8a5e\u500b\u6578\u5c0f\u65bc 2\u3002\u5176\u4e2d\uff0c \u7269 VB \u81ea\u7136\u8a9e\u53e5\u751f\u6210 \u4ee5\u7be9\u9078\u51fa\u7684\u5019\u9078\u53e5\u578b\u6a23\u7248\u70ba\u4f9d\u64da\uff0c\u5229\u7528\u7247\u8a9e\u5d4c\u5165\u898f\u5247\uff0c\u5c07\u4e4b\u524d\u672a\u505a</td></tr><tr><td colspan=\"4\">\u95dc\u9375\u7247\u8a9e\u500b\u6578\u4fc2\u6307\u4f7f\u7528\u8005\u8f38\u5165\u4e4b\u95dc\u9375\u8a5e\u5f59\u5408\u4f75\u5f8c\u7684\u7247\u8a9e\u500b\u6578\u3002\u4f9d\u64da\u4ee5\u4e0a\u9650\u5236\uff0c\u5f9e\u53e5\u578b\u6a23 VC \u7bc0\u9ede\u5339\u914d\u7684\u6642\u9593\u53ca\u5730\u65b9\u7247\u8a9e\u5d4c\u5165\u53e5\u578b\u6a23\u672c\uff0c\u751f\u6210\u5408\u4e4e\u8a9e\u6cd5\u53ca\u8a9e\u610f\u7684\u8a9e\u53e5\uff0c\u4e26\u900f\u904e</td></tr><tr><td colspan=\"4\">VD \u7248\u6a39\u4e2d\u6311\u9078\u51fa\u689d\u4ef6\u7b26\u5408\u7684\u53e5\u578b\u6a23\u7248\u4e4b\u5f8c\uff0c\u958b\u59cb\u4f5c\u53e5\u578b\u6a23\u7248\u6bd4\u5c0d\uff0c\u5176\u6bd4\u5c0d\u7a0b\u5e8f\u5982\u4e0b\uff1a \u52d5 VE Variable N-Gram \u8a9e\u8a00\u6a21\u578b\u865b\u8a5e\u88dc\u7db4\u7684\u6280\u8853\uff0c\u8b93\u751f\u6210\u4e4b\u8a9e\u53e5\u66f4\u52a0\u81ea\u7136\u5b8c\u6574\u3002</td></tr><tr><td colspan=\"4\">i. \u6a23\u7248\u5339\u914d\u958b\u59cb\uff0c\u5c07\u6240\u6709\u95dc\u9375\u7247\u8a9e\u8a2d\u70ba\u672a\u5339\u914d\u3002 \u4f5c \u53ca Step1.)\u7247\u8a9e\u5d4c\u5165\uff1a\u6839\u64da\u6240\u6b78\u7d0d\u4e4b\u6642\u9593\u53ca\u5730\u65b9\u7247\u8a9e\u7684\u5d4c\u5165\u898f\u5247\u70ba\uff1a\u6642\u9593\u7247\u8a9e\u4e43\u7f6e\u65bc\u884c VF</td></tr><tr><td colspan=\"4\">ii. \u7531 root \u7bc0\u9ede\u958b\u59cb\uff0c\u914d\u5408\u7bc0\u9ede\u5339\u914d\u7b56\u7565\uff0c\u6aa2\u67e5\u6240\u6709\u672a\u5339\u914d\u7684\u95dc\u9375\u7247\u8a9e\u4e2d\uff0c\u662f\u5426\u6709\u7b26\u5408 \u52d5 \u7269 VG \u70ba\u8005(agent)\u4e4b\u524d\u6216\u4e4b\u5f8c\uff0c\u5426\u5247\uff0c\u7f6e\u65bc\u53e5\u9996\uff1b\u5730\u65b9\u7247\u8a9e\u4e43\u7f6e\u65bc\u52d5\u4f5c\u985e\u52d5\u8a5e</td></tr><tr><td>\u8a5e</td><td/><td>\u4e0d</td></tr><tr><td colspan=\"3\">\u5176\u5b50\u7bc0\u9ede\u5339\u914d\u689d\u4ef6\u3002 \u4e4b\u524d\uff0c\u5426\u5247\uff0c\u7f6e\u65bc\u53e5\u5c3e\uff0c\u5982\u5716\u516b\u6240\u793a\u3002 \u72c0 \u53ca \u7269</td><td>VH</td></tr><tr><td colspan=\"4\">\u614b iii. \u82e5\u6709\u7b26\u5408\u4e4b\u7247\u8a9e\uff0c\u5247\u5c07\u7247\u8a9e\u586b\u5165\u5176\u5b50\u7bc0\u9ede\u4e4b\u4e2d\uff0c\u56de\u5230\u6b65\u9a5f ii.\u7e7c\u7e8c\u4e0b\u4e00\u7bc0\u9ede\u4e4b\u5339\u914d\u3002 VI root</td></tr><tr><td colspan=\"4\">VJ iv. \u82e5\u7121\u7b26\u5408\u4e4b\u7247\u8a9e\uff0c\u5247\u6aa2\u67e5\u5176\u5b50\u7bc0\u9ede\u662f\u5426\u70ba\u95dc\u9375\u7bc0\u9ede(\u8a5e\u6027\u70ba V \u6216 N \u7684\u7bc0\u9ede) \uff0c\u5982\u679c \u53ca \u7269 VK \u6211 \u4e0d\u662f\u95dc\u9375\u7bc0\u9ede\uff0c\u5247\u8df3\u904e\u6b64\u5b50\u7bc0\u9ede\uff0c\u56de\u5230\u6b65\u9a5f ii.\u7e7c\u7e8c\u4e0b\u4e00\u7bc0\u9ede\u4e4b\u5339\u914d\uff1b\u5426\u5247\uff0c\u4e2d\u65b7\u6b64 VL (agent )</td></tr><tr><td>\u53e5\u578b\u6a23\u7248\u4e4b\u6bd4\u5c0d\u3002 \u03b1 1</td><td>\u03b1 2</td><td>\u03b1 3</td></tr><tr><td colspan=\"4\">v. \u5b8c\u6210\u6bd4\u5c0d\u4e4b\u689d\u4ef6\u70ba\uff1a\u6240\u6709\u95dc\u9375\u7247\u8a9e(\u6642\u9593\u3001\u5730\u65b9\u7247\u8a9e\u9664\u5916) \uff0c\u7686\u6709\u5339\u914d\u4e4b\u7bc0\u9ede\uff1b\u6700\u5f8c \u5716\u516d \u4e2d\u7814\u9662\u52d5\u8a5e\u5206\u985e\u6a39\u72c0\u7d50\u69cb\u5716</td></tr><tr><td colspan=\"4\">).\u7247\u8a9e 2.)\u7d50\u5408\u865b\u8a5e(functional word)\u7701\u7565\u7b56\u7565\uff0c\u5c07\u7121\u95dc\u9375\u8a5e\u5339\u914d\u7684\u865b\u8a5e\u7bc0\u9ede\u8a02\u70ba 0.9 \u5206\uff1b3.) \u7bc0\u9ede\u7684\u5b50\u7bc0\u9ede\u70ba\u5916\u90e8\u7bc0\u9ede(terminal node) \u3002</td></tr><tr><td colspan=\"4\">\u5408\u4f75\uff1a\u5206\u5225\u8655\u7406\u4eba\u3001\u6642\u3001\u5730\u3001\u7269\u7b49\u985e\u5225\uff0c\u4eba\u3001\u7269\u985e\u5225\u5728\u8a9e\u8a00\u5b78\u88e1\u7a31\u70ba argument\uff0c\u6642\u3001 \u53e5\u578b\u6a23\u7248\u8a08\u5206\u8655\u7406\u8207\u7be9\u9078 \u7531\u65bc\u7bc0\u9ede\u5339\u914d\u7b56\u7565\u5c07\u52d5\u8a5e\u8a5e\u6027\u9650\u5236\u653e\u5bec\uff0c\u96d6\u589e\u52a0\u53e5\u578b\u6210 \u8207\u95dc\u9375\u8a5e\u5c6c\u6027\u76f8\u7b26\u6216\u8a5e\u6027\u76f8\u540c\u7684\u7bc0\u9ede\u8a02\u70ba 1.0 \u5206\uff0c\u5982\u5716\u4e03\u6240\u793a\uff0c\u5176\u4e2d node(a)\u4ee3\u8868\u53e5\u578b\u6a23</td></tr><tr><td colspan=\"4\">\u5730\u7b49\u985e\u5225\u7a31\u70ba adjunct\uff1b\u7c21\u8a00\u4e4b\uff0c\u6642\u9593\u53ca\u5730\u65b9\u7247\u8a9e\u4e26\u975e\u69cb\u53e5\u521d\u671f\u6240\u9700\u7269\u4ef6\u3002\u56e0\u6b64\uff0c\u5728\u7bc0 \u529f\u5339\u914d\u7684\u6a5f\u7387\uff1b\u4f46\u653e\u5bec\u9650\u5236\u4ea6\u589e\u52a0\u932f\u8aa4\u5339\u914d\u7684\u53ef\u80fd\u6027\u3002\u56e0\u6b64\uff0c\u672c\u7814\u7a76\u63d0\u51fa\u521d\u6b65\u7684\u7d71\u8a08 \u7248\u7684\u7bc0\u9ede\uff0ctoken(b)\u4ee3\u8868\u8207\u8a72\u7bc0\u9ede\u5339\u914d\u4e4b\u95dc\u9375\u8a5e\uff0c\u4f9d\u64da\u7bc0\u9ede\u5339\u914d\u7b56\u7565\uff1a1.)\u9996\u5148\uff0c\u6aa2\u67e5\u95dc</td></tr><tr><td colspan=\"4\">\u9ede\u5339\u914d\u6642\u53ef\u66ab\u6642\u5ffd\u7565\u6642\u9593\u6216\u5730\u65b9\u7247\u8a9e\uff0c\u7b49\u5230\u8a9e\u53e5\u751f\u6210\u6642\uff0c\u518d\u5229\u7528\u7247\u8a9e\u5d4c\u5165\u7684\u65b9\u5f0f\uff0c\u5c07\u672a \u8a08\u5206\u6a5f\u5236\uff0c\u900f\u904e\u7bc0\u9ede\u5c6c\u6027\u76f8\u4f3c\u5ea6\u53ca\u53e5\u578b\u6a23\u7248\u53c3\u8003\u6b21\u6578\u4e4b\u7d71\u8a08\u8cc7\u6599\uff0c\u4f5c\u70ba\u5019\u9078\u53e5\u578b\u6a23\u7248 \u9375\u8a5e b \u8207\u7bc0\u9ede a \u4e4b\u5c6c\u6027\u6216\u8a5e\u6027\u662f\u5426\u76f8\u7b26\uff1b2.)\u5426\u5247\uff0c\u6aa2\u67e5\u7bc0\u9ede a \u662f\u5426\u70ba\u865b\u8a5e\uff1b3.)\u5426\u5247\uff0c</td></tr><tr><td colspan=\"4\">\u4f5c\u7bc0\u9ede\u5339\u914d\u7684\u6642\u9593\u53ca\u5730\u65b9\u7247\u8a9e\u4f75\u5165\u53e5\u578b\u6a23\u7248\u4e2d\uff0c\u4e26\u540c\u6642\u7e7c\u627f\u88ab\u4fee\u98fe\u8005\u7684\u5c6c\u6027\u53ca\u8a5e\u6027\u3002 \u9032\u884c\u95dc\u9375\u8a5e b \u8207\u7bc0\u9ede a \u4e4b\u52d5\u8a5e\u76f8\u4f3c\u5ea6\u6bd4\u5c0d\u3002</td></tr></table>",
+                "text": "\u5c64\u7d1a\u70ba 0.25 \u5206\uff0c\u03b1 2 \u5c64\u7d1a\u70ba 0.5 \u5206\uff0c\u03b1 3 \u5c64\u7d1a\u70ba 0.75 \u5206\uff1b",
+                "html": null,
+                "type_str": "table",
+                "num": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-2001.json

@@ -0,0 +1,707 @@
+{
+    "paper_id": "O00-2001",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:05.301759Z"
+    },
+    "title": "A Lexical-Semantic Analysis of Mandarin Chinese Verbs: Representation and Methodology",
+    "authors": [
+        {
+            "first": "",
+            "middle": [],
+            "last": "\u5f35\uf988\uf988",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Academia Sinica",
+                "location": {}
+            },
+            "email": ""
+        },
+        {
+            "first": "Li-Li",
+            "middle": [],
+            "last": "Chang",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Academia Sinica",
+                "location": {}
+            },
+            "email": ""
+        },
+        {
+            "first": "Keh-Jiann",
+            "middle": [],
+            "last": "Chen",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Academia Sinica",
+                "location": {}
+            },
+            "email": ""
+        },
+        {
+            "first": "Chu-Ren",
+            "middle": [],
+            "last": "Huang",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Academia Sinica",
+                "location": {}
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-2001",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "\u5206\uff0c\u5206\u5225\u662f 1)\u52d5\u8a5e\u8a5e\u5f59\u8a9e\u610f\u7684\u8868\u9054\u6a21\u5f0f; 2)\u52d5\u8a5e\u8a5e\u5f59\u8a9e\u610f\u7684\u5340\u5206\u548c\u8a5e\u7fa9\u5ef6\u4f38; 3) \u52d5\u8a5e\u8a5e\u5f59\u8a9e\u610f\u548c\u5176\u4ed6\u8a5e\u5f59\u8a9e\u610f\u7684\u7d50\u5408\u60c5\u6cc1\u3002\u63a5\u4e0b\uf92d\uff0c\u6211\u5011\u6703\u8a73\uf99c\u5be6\u969b\u5206\u6790\u7684\u65b9 \u6cd5\u3001\u6b65\u9a5f\u548c\u7bc4\uf9b5\u3002\u6700\u5f8c\u4f5c\u500b\u7e3d\u7d50\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u5728\u6587\u4e2d\u6211\u5011\u5c07\u5148\u4ecb\u7d39\u7814\u7a76\u80cc\u666f\uff0c\u7dca\u63a5\u8457\u9032\u5165\u4e3b\u984c\uff0c\u4ecb\u7d39\uf9e4\uf941\u67b6\u69cb\u7684\u4e09\u500b\u90e8",
+                "sec_num": null
+            },
+            {
+                "text": "In this paper we will briefly introduce the Module-Attribute Representation of Verbal Semantics (MARVS) and present in detail the methods used to analyze verbal semantics by the CKIP group. The theory and the methodology are based on the analysis of forty synonym pairs of verbs as well as verbs from ten different semantic fields. This paper will focus on the linguistic data and our research methodology. For more information on the theoretical issues performing to MARVS, please see Huang et al. [this volume] . The research results published by the members of CKIP group on a certain synonym pairs or semantic fields will also be discussed in this paper, such as Chang et al. [this volume] on mental verbs, Liu et al. [1997] on building verbs JIAN, GAI and ZAO, Liu et al. [this volume] on throwing verbs TOU, ZHI, DIU and RENG, Liu et al. [1999] on chasing verbs ZHUI and GAN, and Chief et al. [this volume] on verbs meaning \"beneficial\", FANGBIAN and BIANLI. This paper will be organized in the following way. In section 1 we will first introduce our basic ideas on verbal semantics. In section 2 we will discuss three related research topics, i.e. the MARVS theory, the distinction and extension of verbal meanings, and the co-occurrence of verbs with certain sentence patterns or adjuncts. In section 3 the methodology used for analyzing synonym pairs and verbs in a particular semantic field will be presented. In section 4 we will give an example of the near-synonym verbs KUAILE \"happy\" and GAOXING \"glad\" and show precisely what to observe, how to compare and how to explain the differences in detail. This paper is a record of our research methodology and will be used as a technical guide for the CKIP group. We will keep on modifying our research methods and the theory in the future and we look forward to feedback from readers of this paper. ",
+                "cite_spans": [
+                    {
+                        "start": 486,
+                        "end": 512,
+                        "text": "Huang et al. [this volume]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 667,
+                        "end": 679,
+                        "text": "Chang et al.",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 711,
+                        "end": 728,
+                        "text": "Liu et al. [1997]",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 766,
+                        "end": 776,
+                        "text": "Liu et al.",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 833,
+                        "end": 850,
+                        "text": "Liu et al. [1999]",
+                        "ref_id": "BIBREF12"
+                    },
+                    {
+                        "start": 882,
+                        "end": 898,
+                        "text": "and Chief et al.",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u9019\u7bc7\u6587\u7ae0\u6709\uf978\u500b\u76ee\u7684\uff1a\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u4e00\u518d\u4fee\u6b63\u7684\uf9e4\uf941\u548c\u65b9\u6cd5\u4f5c\u500b\u968e\u6bb5\u6027 \u7684\u7e3d\u7d50\uff0c\u4f5c\u70ba\u65e5\u5f8c\u5206\u6790\u7684\uf96b\u8003\u4f9d\u64da\uff1b\u53e6\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u7684\uf9e4\uf9a3\u548c\u4f5c\u6cd5\u4ecb\u7d39\u7d66\u5927 \u5bb6\uff0c\u5e0c\u671b\u80fd\u5f97\u5230\u5404\u754c\u4eba\u58eb\u7684\u6279\u8a55\u6307\u6559\u3002",
+                "sec_num": null
+            },
+            {
+                "text": "2. \u7814\u7a76\u4e3b\u984c\uff1a \u6b63\u5982\u7814\u7a76\u80cc\u666f\u7684\uf96f\u660e\uff0c\u6211\u5011\u7684\u89c0\u5bdf\u548c\u7814\u7a76\u4e3b\u984c\u5305\u62ec\u4e09\u65b9\u9762\uff1a1)\u52d5\u8a5e\u8a5e\u5f59\u8a9e\u610f\u7684\u8868\u9054\u6a21\u5f0f\uff1b 2)\u52d5\u8a5e\u8a5e\u5f59\u8a9e\u610f\u7684\u5340\u5206\u548c\u8a5e\u5f59\u8a9e\u610f\u5ef6\u4f38\uff1b 3)\u52d5\u8a5e\u8a5e\u5f59\u8a9e\u610f\u548c\u5176\u4ed6\u8a5e\u5f59\u8a9e\u610f\u6216\u7d50\u69cb\u8a9e\u610f\u7684 \u7d50\u5408\u60c5\u6cc1\u3002 2.1 \u52d5\u8a5e\u8a5e\u5f59\u8a9e\u610f\u7684\u8868\u9054\u65b9\u5f0f \u70ba\uf9ba\u660e\u78ba\u638c\u63e1\u8a5e\u5f59\u8a9e\u610f\u7684\u5f71\u97ff\uf98a\uff0c\u6211\u5011\u9700\u8981\u4e00\u5957\u7531\u8a9e\u610f\u5c6c\u6027\u7d44\u6210\u7684\u67b6\u69cb\uf92d\u8868\u9054\u8a5e\u5f59\u8a9e \u610f\u3002\u52d5\u8a5e\u8868\u9054\u7684\u6982\uf9a3\u662f\u300c\u4e8b\u4ef6\u300d\uff0c\u5176\u5167\u5bb9\u6d89\u53ca\u4e8b\u4ef6\u7684\uf96b\u8207\u8005\u3001\u4e8b\u4ef6\u7684\u57f7\ufa08\u65b9\u5f0f\u3001\u4e8b\u4ef6\u9032 \ufa08\u7684\u6642\u9593\u9577\u77ed\u548c\u5730\u9ede\u2026\u7b49\u3002\u4f46\u662f\u901a\u5e38\u4e00\u500b\uf906\u5b50\u6240\u8868\u9054\u7684\u4e8b\u4ef6\u5167\u5bb9\uff0c\u53ef\u80fd\u53ea\u662f\u5b8c\u6574\u4e8b\u4ef6\u7684 \u4e00\u90e8\u4efd\uff0c\u7531\u52d5\u8a5e\u53ca\u5176\uf941\u5143\u53ca\u9644\u52a0\u6210\u5206\u6240\u8868\u793a\u7684\u8a9e\u610f\u6210\u5206\u7d50\u5408\u800c\u6210\u3002\uf906\u5b50\u6240\u80fd\u8868\u9054\u7684\u4e8b\u4ef6 \u5167 \u5bb9 \u53ca \u65b9 \u5f0f \u4e3b \u8981 \u662f \u53d7 \u6838 \u5fc3 \u52d5 \u8a5e \u9650 \u5236 \u7684 \uff0c \u6240 \u4ee5 \u6211 \u5011 \u5b9a \u51fa \u4e00 \u500b \u52d5 \u8a5e \u8a9e \u610f \u8868 \u9054 \u6a21 \u5f0f \"Module-Attribute Representation of Verbal Semantics (MARVS)\" \uff0c \u7c21 \u7a31 \u70ba \"MARVS",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u9019\u7bc7\u6587\u7ae0\u6709\uf978\u500b\u76ee\u7684\uff1a\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u4e00\u518d\u4fee\u6b63\u7684\uf9e4\uf941\u548c\u65b9\u6cd5\u4f5c\u500b\u968e\u6bb5\u6027 \u7684\u7e3d\u7d50\uff0c\u4f5c\u70ba\u65e5\u5f8c\u5206\u6790\u7684\uf96b\u8003\u4f9d\u64da\uff1b\u53e6\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u7684\uf9e4\uf9a3\u548c\u4f5c\u6cd5\u4ecb\u7d39\u7d66\u5927 \u5bb6\uff0c\u5e0c\u671b\u80fd\u5f97\u5230\u5404\u754c\u4eba\u58eb\u7684\u6279\u8a55\u6307\u6559\u3002",
+                "sec_num": null
+            },
+            {
+                "text": "Representation\"\uf92d\u8868\u9054\u6bcf\u4e00\u500b\u52d5\u8a5e\u7fa9\u9805\u7684\u4e8b\u4ef6\u8a0a\u606f\u7d50\u69cb [Huang and Ahrens, 1999] ",
+                "cite_spans": [
+                    {
+                        "start": 33,
+                        "end": 57,
+                        "text": "[Huang and Ahrens, 1999]",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u9019\u7bc7\u6587\u7ae0\u6709\uf978\u500b\u76ee\u7684\uff1a\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u4e00\u518d\u4fee\u6b63\u7684\uf9e4\uf941\u548c\u65b9\u6cd5\u4f5c\u500b\u968e\u6bb5\u6027 \u7684\u7e3d\u7d50\uff0c\u4f5c\u70ba\u65e5\u5f8c\u5206\u6790\u7684\uf96b\u8003\u4f9d\u64da\uff1b\u53e6\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u7684\uf9e4\uf9a3\u548c\u4f5c\u6cd5\u4ecb\u7d39\u7d66\u5927 \u5bb6\uff0c\u5e0c\u671b\u80fd\u5f97\u5230\u5404\u754c\u4eba\u58eb\u7684\u6279\u8a55\u6307\u6559\u3002",
+                "sec_num": null
+            },
+            {
+                "text": "\u2027//////\u2027 \u5efa\u3001\u88fd\u9020\u3001\u5403\u3001\u5403\u98ef\u2026 \u59cb\u7d42\u52d5\u4f5c(bounded activity) -----------------------------------------------------------------------------------------------------------",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u9019\u7bc7\u6587\u7ae0\u6709\uf978\u500b\u76ee\u7684\uff1a\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u4e00\u518d\u4fee\u6b63\u7684\uf9e4\uf941\u548c\u65b9\u6cd5\u4f5c\u500b\u968e\u6bb5\u6027 \u7684\u7e3d\u7d50\uff0c\u4f5c\u70ba\u65e5\u5f8c\u5206\u6790\u7684\uf96b\u8003\u4f9d\u64da\uff1b\u53e6\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u7684\uf9e4\uf9a3\u548c\u4f5c\u6cd5\u4ecb\u7d39\u7d66\u5927 \u5bb6\uff0c\u5e0c\u671b\u80fd\u5f97\u5230\u5404\u754c\u4eba\u58eb\u7684\u6279\u8a55\u6307\u6559\u3002",
+                "sec_num": null
+            },
+            {
+                "text": "------------------------------------------------------------------------------------------------------------ \u2027 \u6253\u6b7b\u3001\u6253\u7834\u3001\u9a0e\uf94f\u3001\u7b54\u5c0d\u2026 \u7d50\u679c(",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u9019\u7bc7\u6587\u7ae0\u6709\uf978\u500b\u76ee\u7684\uff1a\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u4e00\u518d\u4fee\u6b63\u7684\uf9e4\uf941\u548c\u65b9\u6cd5\u4f5c\u500b\u968e\u6bb5\u6027 \u7684\u7e3d\u7d50\uff0c\u4f5c\u70ba\u65e5\u5f8c\u5206\u6790\u7684\uf96b\u8003\u4f9d\u64da\uff1b\u53e6\u4e00\u65b9\u9762\u662f\u5c07\u6211\u5011\u7684\uf9e4\uf9a3\u548c\u4f5c\u6cd5\u4ecb\u7d39\u7d66\u5927 \u5bb6\uff0c\u5e0c\u671b\u80fd\u5f97\u5230\u5404\u754c\u4eba\u58eb\u7684\u6279\u8a55\u6307\u6559\u3002",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Alternation Across Semantic Fields: A Study of Mandarin Verbs of Emotion",
+                "authors": [
+                    {
+                        "first": "Li-Li",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Keh-Jiann",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of the 13 th Pacific Asia Conference on Language, Information and Computation",
+                "volume": "",
+                "issue": "",
+                "pages": "39--50",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chang, Li-li, Keh-jiann Chen and Chu-Ren Huang. 1999. \"Alternation Across Semantic Fields: A Study of Mandarin Verbs of Emotion.\" Proceedings of the 13 th Pacific Asia Conference on Language, Information and Computation, pp39-50, Taipei.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "What Can Near Synonyms Tell Us",
+                "authors": [
+                    {
+                        "first": "Lian-Cheng",
+                        "middle": [],
+                        "last": "Chief",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Keh-Jiann",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Mei-Chih",
+                        "middle": [],
+                        "last": "Tsai",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Li-Li",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the 1998 International Lexical-Functional Grammar Conference",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chief, Lian-Cheng, Chu-Ren Huang, Keh-Jiann Chen, Mei-Chih Tsai, and Li-li Chang. 1998. \"What Can Near Synonyms Tell Us.\" In Proceedings of the 1998 International Lexical-Functional Grammar Conference. Brisbane, Australia.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "A Description to the Sinica Corpus",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Ckip",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "CKIP. 1995. A Description to the Sinica Corpus. Technical Report 95-02. Academia Sinica. Taipei.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Semantic Fields, Prototypes, and the Lexicon",
+                "authors": [
+                    {
+                        "first": "Richard",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Grandy",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "Lehrer and Kittay Eds. Frames, Fields, and Contrasts: New Essays in Semantic and Lexical Organization. Pp103-122",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Grandy, Richard E. 1992. \"Semantic Fields, Prototypes, and the Lexicon.\" In Lehrer and Kittay Eds. Frames, Fields, and Contrasts: New Essays in Semantic and Lexical Organization. Pp103-122. Hillsdale: Lawrence Erlbaum.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Classifying Event Structure Attributes: A Verbal Semantic Perspective from Chinese",
+                "authors": [
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "The 1998 International Lexical-Functional Grammar Conference",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Huang, Chu-Ren. 1998. Classifying Event Structure Attributes: A Verbal Semantic Perspective from Chinese. Invited paper. Chinese Workshop. The 1998 International Lexical-Functional Grammar Conference. Brisbane, Australia.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "The Module-Attribute Representation of Verbal Semantics",
+                "authors": [
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Kathleen",
+                        "middle": [],
+                        "last": "Ahrens",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Academia Sinica, and Chinese Knowledge Processing Group",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Huang, Chu-Ren and Kathleen Ahrens. 1999. \"The Module-Attribute Representation of Verbal Semantics.\" In Working Papers on Chinese Verbal Semantics, Vol. I. Corpus Research Group, Institute of Linguistics, Academia Sinica, and Chinese Knowledge Processing Group, Institute of Information Science, Academia Sinica. Taipei.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "From Near Synonyms to Event Structure: Corpus-based Studies of Mandarin Verbal Semantics",
+                "authors": [
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Mei-Chih",
+                        "middle": [],
+                        "last": "Tsai",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Mini-Conference on Lexical Semantics. November",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Huang, Chu-Ren and Mei-Chih Tsai. 1997. From Near Synonyms to Event Structure: Corpus-based Studies of Mandarin Verbal Semantics. Mini-Conference on Lexical Semantics. November. National Chung-cheng University.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "From Lexical Meaning to Event Structure Attributes: Across Semantic Classes of Mandarin Verbs",
+                "authors": [
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Mei-Chun",
+                        "middle": [],
+                        "last": "Liu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Mei-Chih",
+                        "middle": [],
+                        "last": "Tsai",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "The 6th International Conference on Chinese Linguistics/The 10th North American Conference on Chinese Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Huang, Chu-Ren, Mei-chun Liu, and Mei-chih Tsai. 1998. \"From Lexical Meaning to Event Structure Attributes: Across Semantic Classes of Mandarin Verbs.\" The 6th International Conference on Chinese Linguistics/The 10th North American Conference on Chinese Linguistics. June 26-28. Stanford.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "English Verb Classes and Alternations: A Preliminary Investigation",
+                "authors": [
+                    {
+                        "first": "Beth",
+                        "middle": [],
+                        "last": "Levin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Levin, Beth. 1993. English Verb Classes and Alternations: A Preliminary Investigation. Chicago: University of Chicago Press.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Mandarin Chinese: A Functional Reference Grammar",
+                "authors": [
+                    {
+                        "first": "Charles & Sandra",
+                        "middle": [],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Thompson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1981,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Li, Charles & Sandra Thompson. 1981. Mandarin Chinese: A Functional Reference Grammar. California: University of California Press.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Lexical Meaning and Discourse Patterning -the three Mandarin cases of 'build'",
+                "authors": [
+                    {
+                        "first": "Mei-Chun",
+                        "middle": [],
+                        "last": "Liu",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Paper presented at the Third Conference on Conceptual Structure, Discourse, and Language",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Liu, Mei-chun. 1997. \"Lexical Meaning and Discourse Patterning -the three Mandarin cases of 'build'.\" Paper presented at the Third Conference on Conceptual Structure, Discourse, and Language. Boulder, Colorado.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "When Endpoint Meets Endpoint: A Corpus-based Semantic Study of Throwing Verbs",
+                "authors": [
+                    {
+                        "first": "Liu",
+                        "middle": [],
+                        "last": "Mei-Chun",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Charles",
+                        "middle": [
+                            "C L"
+                        ],
+                        "last": "Lee",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "The 6th International Conference on Chinese Linguistics/The 10th North American Conference on Chinese Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Liu Mei-chun, Chu-Ren Huang, and Charles C.L. Lee. 1998. \"When Endpoint Meets Endpoint: A Corpus-based Semantic Study of Throwing Verbs.\" The 6th International Conference on Chinese Linguistics/The 10th North American Conference on Chinese Linguistics. June 26-28. Stanford.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "Lexical Information and Beyond: Constructional Inferences in Semantic Representation",
+                "authors": [
+                    {
+                        "first": "Mei-Chun",
+                        "middle": [],
+                        "last": "Liu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ching-Yi",
+                        "middle": [],
+                        "last": "Lee",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of the 13 th Pacific Asia Conference on Language, Information and Computation",
+                "volume": "",
+                "issue": "",
+                "pages": "27--38",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Liu, Mei-Chun, Chu-Ren Huang and Ching-Yi Lee. 1999. \"Lexical Information and Beyond: Constructional Inferences in Semantic Representation.\" Proceedings of the 13 th Pacific Asia Conference on Language, Information and Computation, pp27-38, Taipei.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Lexical Semantic Techniques for Corpus Analysis",
+                "authors": [
+                    {
+                        "first": "James",
+                        "middle": [],
+                        "last": "Pustejovsky",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Bergler",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Anick",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Computational Linguistics",
+                "volume": "19",
+                "issue": "",
+                "pages": "331--358",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Pustejovsky, James, S. Bergler, and P. Anick. 1993. \"Lexical Semantic Techniques for Corpus Analysis.\" Computational Linguistics, 19.2, pp331-358.",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "The Generative Lexicon",
+                "authors": [
+                    {
+                        "first": "James",
+                        "middle": [],
+                        "last": "Pustejovsky",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Pustejovsky, James. 1995. The Generative Lexicon. Cambridge: MIT Press.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "The Parameter of Aspect",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Smith",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Smith, C. 1991. The Parameter of Aspect. Dordrecht: Kluwer.",
+                "links": null
+            },
+            "BIBREF16": {
+                "ref_id": "b16",
+                "title": "The Aspectual Interface Hypothesis",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Tenny",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Tenny, C. 1992. \"The Aspectual Interface Hypothesis.\" In I. Sag and A. Szabolcsi Eds. Lexical Matter. Standford: CSLI.",
+                "links": null
+            },
+            "BIBREF17": {
+                "ref_id": "b17",
+                "title": "Towards a Representation of Verbal Semantic: An Approach Based on Near-Synonyms",
+                "authors": [
+                    {
+                        "first": "Mei-Chi",
+                        "middle": [],
+                        "last": "Tsai",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Keh-Jiann",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Kathleen",
+                        "middle": [],
+                        "last": "Ahrens",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "International Journal of Computational Linguistics & Chinese Language Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "62--74",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Tsai, Mei-chi, Chu-Ren Huang, Keh-jiann Chen, and Kathleen Ahrens. 1998. \"Towards a Representation of Verbal Semantic: An Approach Based on Near-Synonyms.\" International Journal of Computational Linguistics & Chinese Language Processing, pp62-74.",
+                "links": null
+            },
+            "BIBREF18": {
+                "ref_id": "b18",
+                "title": "Verbs and Times",
+                "authors": [
+                    {
+                        "first": "Zeno",
+                        "middle": [],
+                        "last": "Vendler",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1957,
+                "venue": "Also in Z. Vendler. 1967. Linguistics in Philosophy",
+                "volume": "56",
+                "issue": "",
+                "pages": "97--121",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Vendler, Zeno 1957. \"Verbs and Times.\" Philosophical Review 56, 143-160. Also in Z. Vendler. 1967. Linguistics in Philosophy. Cornell University Press. Ithaca, NY, 97-121.",
+                "links": null
+            }
+        },
+        "ref_entries": {}
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-2002.json

@@ -0,0 +1,270 @@
+{
+    "paper_id": "O00-2002",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:10.480881Z"
+    },
+    "title": "",
+    "authors": [
+        {
+            "first": "Chu-Ren",
+            "middle": [],
+            "last": "Huang",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Kathleen",
+            "middle": [],
+            "last": "Ahrens",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Li-Li",
+            "middle": [],
+            "last": "Chang",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Keh-Jiann",
+            "middle": [],
+            "last": "Chen",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Mei-Chun",
+            "middle": [],
+            "last": "Liu",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Mei-Chi",
+            "middle": [],
+            "last": "Tsai",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-2002",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "& 5 +XDQJ HW DO",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "'RZW\\ 'DYLG 7KHPDWLF 3URWR5ROHV DQG $UJXPHQW 6HOHFWLRQ /DQJXDJH",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "'RZW\\ 'DYLG 7KHPDWLF 3URWR5ROHV DQG $UJXPHQW 6HOHFWLRQ /DQJXDJH",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "+XDQJ &KX5HQ &ODVVLI\\LQJ (YHQW 6WUXFWXUH $WWULEXWHV $ 9HUEDO 6HPDQWLF 3HUVSHFWLYH",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "+XDQJ &KX5HQ &ODVVLI\\LQJ (YHQW 6WUXFWXUH $WWULEXWHV $ 9HUEDO 6HPDQWLF 3HUVSHFWLYH",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "RUNVKRS 7KH ,QWHUQDWLRQDO /H[LFDO)XQFWLRQDO & 5 +XDQJ HW DO",
+                "authors": [
+                    {
+                        "first": "Qylwhg",
+                        "middle": [],
+                        "last": "Iurp &klqhvh",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Sdshu &klqhvh",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "IURP &KLQHVH ,QYLWHG SDSHU &KLQHVH :RUNVKRS 7KH ,QWHUQDWLRQDO /H[LFDO)XQFWLRQDO & 5 +XDQJ HW DO",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "&RQFHSWXDO 6WUXFWXUH 'LVFRXUVH DQG /DQJXDJH 6WDQIRUG &6/, DQG &DPEULGJH &DPEULGJH",
+                "authors": [
+                    {
+                        "first": "*udppdwlfdol]dwlrq $ 6wxg\\ Ri 0dqgdulq &klqhvh Tlodl $ *",
+                        "middle": [],
+                        "last": "Rogehuj",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "8",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "*UDPPDWLFDOL]DWLRQ $ 6WXG\\ RI 0DQGDULQ &KLQHVH TLODL $ *ROGEHUJ (G &RQFHSWXDO 6WUXFWXUH 'LVFRXUVH DQG /DQJXDJH 6WDQIRUG &6/, DQG &DPEULGJH &DPEULGJH 8 3UHVV",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "+XDQJ &KX5HQ DQG 0HL&KLK 7VDL )URP 1HDU 6\\QRQ\\PV WR",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "+XDQJ &KX5HQ DQG 0HL&KLK 7VDL )URP 1HDU 6\\QRQ\\PV WR (YHQW 6WUXFWXUH",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "&RUSXVEDVHG 6WXGLHV RI 0DQGDULQ 9HUEDO 6HPDQWLFV 0LQL&RQIHUHQFH RQ /H[LFDO 6HPDQWLFV 1RYHPEHU 1DWLRQDO &KXQJ&KHQJ 8QLYHUVLW\\",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "&RUSXVEDVHG 6WXGLHV RI 0DQGDULQ 9HUEDO 6HPDQWLFV 0LQL&RQIHUHQFH RQ /H[LFDO 6HPDQWLFV 1RYHPEHU 1DWLRQDO &KXQJ&KHQJ 8QLYHUVLW\\",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "LFDO 6HPDQWLFV DQG &RQVWUXFWLRQDO 0HDQLQJV ,Q &KLQHVH &KLQHVH /DQJDXJHV DQG /LQJXLVWLFV 9 7DLSHL ,QVWLWXWH RI /LQJXLVWLFV $FDGHPLD 6LQLFD",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "+xdqj &kx5hq &kdqj /L3lqj",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Qwhudfwlrq",
+                        "middle": [],
+                        "last": "Dwkohhq $kuhqv Dqg &kdrudq &khq 7kh",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Ri /H",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "+XDQJ &KX5HQ &KDQJ /L3LQJ .DWKOHHQ $KUHQV DQG &KDRUDQ &KHQ 7KH ,QWHUDFWLRQ RI /H[LFDO 6HPDQWLFV DQG &RQVWUXFWLRQDO 0HDQLQJV ,Q &KLQHVH &KLQHVH /DQJDXJHV DQG /LQJXLVWLFV 9 7DLSHL ,QVWLWXWH RI /LQJXLVWLFV $FDGHPLD 6LQLFD",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "QJOLVK 9HUE &ODVVHV DQG $OWHUQDWLRQV $ 3UHOLPLQDU\\ ,QYHVWLJDWLRQ &KLFDJR 8QLYHUVLW\\ RI &KLFDJR 3UHVV",
+                "authors": [
+                    {
+                        "first": "/",
+                        "middle": [],
+                        "last": "Hylq %hwk",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/HYLQ %HWK (QJOLVK 9HUE &ODVVHV DQG $OWHUQDWLRQV $ 3UHOLPLQDU\\ ,QYHVWLJDWLRQ &KLFDJR 8QLYHUVLW\\ RI &KLFDJR 3UHVV",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "9HQGOHU =HQR /LQJXLVWLFV LQ 3KLORVRSK\\ ,WKDFD &RUQHOO 8QLYHUVLW\\ 3UHVV",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "9HQGOHU =HQR /LQJXLVWLFV LQ 3KLORVRSK\\ ,WKDFD &RUQHOO 8QLYHUVLW\\ 3UHVV",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF0": {
+                "content": "<table><tr><td>#'$ 0$596 0$596 0$596 0$596 0$596</td><td>&amp; 5 +XDQJ HW DO &amp; 5 +XDQJ HW DO &amp; 5 +XDQJ HW DO &amp; 5 +XDQJ HW DO &amp; 5 +XDQJ HW DO</td></tr><tr><td colspan=\"2\">DOWKRXJK HDFK FRPSRQHQW LV OLQJXLVWLFDOO\\ DWWHVWHG 7KLV K\\SRWKHVLV LV PRWLYDWHG E\\ RXU E %RXQGHG SURFHVV E WD ]XR TLDQPLDQ E ]DL TLQGL VKHQVKDQJ NDQOH ZXVKLOLX GDR E PDPD IDQJFKX \\L ]XR FDL F EDL GLWDQ D ODRVKL JDL OH VDQ SLDQ ]XRZHQ D TLQJNXDQJ ELDQKDRKXDL OH</td></tr><tr><td colspan=\"2\">K\\SRWKHVLV WKDW V\\QWDFWLF YDULDWLRQV LQFOXGLQJ /HYLQV &gt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gt;+XDQJ DQG 7VDL +XDQJ @ $VSHFWXDO DWWULEXWHV DWWULEXWHV SHUWDLQLQJ WR WKH FRPSRVLWLRQ RI WKH HYHQWV VXFK DV 7HOLFLW\\ +RPRJHQHLW\\ HWF (YHQWLQWHUQDO DWWULEXWHV DWWULEXWHV UHIHUULQJ WR WKH VHPDQWLFV RI WKH HYHQW LWVHOI VXFK DV &amp;RQWURO (IIHFW HWF 5ROH DWWULEXWHV DWWULEXWHV UHIHUULQJ WR WKH IRFXVVHG UROHV RI WKH HYHQW VXFK DV $JHQW 7KHPH ,QVWUXPHQW 0DQQHU HWF 5ROH,QWHUQDO DWWULEXWHV DWWULEXWHV UHIHUULQJ WR WKH LQWHUQDO VHPDQWLFV RI D SDUWLFXODU IRFXVHG UROH RI WKH HYHQW VXFK DV VHQWLHQFH YROLWLRQ DIIHFWHGQHVV HWF +RZHYHU DV WKH 9HUE 6HQVH L (YHQWLYH ,QIRUPDWLRQ (YHQW 0RGXOHV (YHQW,QWHUQDO $WWULEXWHV 5ROH 0RGXOHV 5ROH,QWHUQDO $WWULEXWHV \u2022 \u2022 \u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amp;KLQHVH DW OHDVW WKH DFWLYLW\\ RU SXWWLQJ RQ DQG WKH VWDWH RI ZHDULQJ &amp;RPSOHWH HYHQW YV RWKHU HYHQWV D 6KHPH VKLKRX 9 OH :KHQ 9 $63 E 6KHPH VKLKRX NDLKXL OH\" :KHQ PHHWLQJ :KHQ GRHV WKH PHHWLQJ VWDUW\" F 6KHPH VKLKRX GDVXDQ OH\" :KHQ SODQ 6HFRQG VLQFH SURFHVV HQFRGHV D WLPH FRXUVH D GXUDWLRQDO SKUDVH QDWXUDOO\\ PHDVXUHV WKH OHQJWK RI WKH WLPH FRXUVH DQG FDQ GLVWLQJXLVK EHWZHHQ SURFHVV HYHQWV DQG ERXQGDU\\FRPSOHWH HYHQWV DV DQG VKRZ 3URFHVV YV &amp;RPSOHWH (YHQW%RXQGDU\\ 9 OH 'XUDWLRQ 9 $63 'XUDWLRQ D \\L]KL VL DOZD\\V GLH E \\L]KL SDR DOZD\\V UXQ 6KH KDV EHHQ UXQQLQJ FRQWLQXRXVO\\ F \\L]KL VL VL OH VDQ JH [LDRVKL DOZD\\V GLH GLH $63 WKUHH &amp;/ KRXUV +HV EHHQ GHDG IRU WKUHH KRXUV G \\L]KL SDR SDR OH VDQ JH [LDRVKL DOZD\\V UXQ UXQ $63 WKUHH &amp;/ KRXUV +H KDV NHSW RQ UXQQLQJ IRU WKUHH KRXUV HYHQWV 6WDJH YV $FWLYLW\\ D WD SDROH VDQ JH ]KRQJWRX VKH UXQ/( WKUHH &amp;/6 KRXU 6+H KDV EHHQ UXQQLQJ IRU WKUHH KRXUV E KXD GLDR[LHOH VDQ WLDQ IORZHU ZLWKHU/( WKUHH GD\\ FRPS7KHVH IORZHUV KDYH EHHQ ZLWKHULQJ RQ WKH YLQH IRU WKH SDVW WKUHH GD\\V F VKLTL PLPDQOH VDQ WLDQ KXPLGLW\\ SHUPHDWH WKUHH GD\\ 7KH KXPLG DLU KDV EHHQ SHUPHDWLQJ IRU WKUHH GD\\V 7KLUG D $WRPLF (YHQW 5HSUHVHQWDWLRQ 7KH YHUEV OLVWHG EHORZ LQ KDYH VWDQGDORQH HYHQW PRGXOHV D &amp;RPSOHWLRQ DFKLHYHPHQW \u2022 VL WR GLH SR WR EUHDN E 3XQFWXDOLW\\ GDVXDQ WR SODQ WR F 3URFHVV ]RX ZDON SDR UXQ G +RPRJHQHRXV 6WDWH FLDWHG QRQLQVWDQWDQHRXV HYHQW D ,QFKRDWLYH 3URFHVV \u2022 [LD\\X WR UDLQ NDLKXL WR FRQYHQH D PHHWLQJ VKH VLW IURQW DW ORYHIRH ERG\\WRS .$13(5) NQLIH IDFW DQG WKH XVXDO LPSOLFDWLYH UHODWLRQVKLS EHWZHHQ &gt;FRQWURO@ DQG &gt;YROLWLRQ@ ZLOO EH PRWKHU SXWRXW RQH WDEOH GLVK VHWJURXQGVSUHDG WHDFKHU UHYLVH 3(5) WKUHH &amp;/6 ZULWLQJ VLWXDWLRQ FKDQJJRRGEDG 3(5) \u2022 \u2022 JDL WR EXLOG F 5HVXOWDWLYH \u2022 GDVL WR KLW DQG NLOO G &amp;RPSOHWLYH 3XQFWXDOLW\\ \u2022 FKXID VHW IRUWK EL\\H JUDGXDWH OLNDL JR DZD\\ H ,QFKRDWLYH 6WDWH (IIHFW 6WDWH \u2022 BBBB JDR[LQJ WR EH JODG I ,QFKRDWLYH 6WDJH \u2022 AAAA VKDQJVKHQJ WR ULVH J %RXQGHG 6WDJH D &amp;RPSOHWLYH 5HVXOWDWLYH BBBBB ]XR WR VLW WDQJ WR OLH &gt;GRZQ@ EDRZHL WR VXUURXQG E 'XDO 3URFHVV6WDWH \u2022 \u2022 D ]XR VLW 6LW &gt;GRZQ@ %H VHDWHG 6+H LV VHDWHG LQ WKH IURQW F WD ]XR OH VDQ JH ]KRQJWRX VKH VLW $63 WKUHH &amp;/$66 KRXU 6+H KDV EHHQ VLWWLQJ IRU WKUHH KRXUV G KDRKDR ]XR ZHOO VLW 6LW VWUDLJKW (YHQWLQWHUQDO $WWULEXWHV &gt;FRQWURO@ ELH JDR[LQJELH NXDLOH 1(* KDSS\\ 1(* KDSS\\ 'RQW EH KDSS\\ 5ROH 0RGXOHV WKH IUHTXHQF\\ RI WKH DFWLYLW\\ VKH %$ DUP *(3(5) WHQSOXVNQLIH VR VKRZ UHVROXWLRQ 6+H PDGH PRUH WKDQ WHQ FXWV RQ KLVKHU DUP WR VKRZ KLVKHU UHVROXWLRQ D PDPD EDLFKX \\L ]XR FDL PRWKHU VHWRXW RQH WDEOH GLVK D VLQJOH WKHPH VXEMHFW ,OOXVWUDWLYH H[DPSOHV DUH JLYHQ EHORZ VXQ WRZDUG ZHVW PRYH 3(5) RQH &amp;/6 KRXU OLQNHG HYHQWV VXFK DV FDXVDWLYH SXUSRVLYH HWF DUH VWLOO EHLQJ GHYHORSHG :H ZLOO QRW JLYH D 0$596 UHSUHVHQWDWLRQ IRU WKH YHUEV LQ WKLV VHFWLRQ 7KH 0$596 UHSUHVHQWDWLRQ RI 0RWKHU FRRNHG DQG VHW D WDEOHIXO RI GLVKHV VHW 6% RQH &amp;/6 ZHDWKHU WUDQVIRUP 3(5) WKUHH KRXUV E WDL\\DQJ [LDQJ [L \\L OH \\L JH ]KRQJWRX LQKHUHQW WLPHGXUDWLRQ ZKLOH TLH GHQRWHV D PRYHPHQW ZKRVH PDQQHU LV QRW VSHFLILHG WR VHW VRPHRQH XS RQFH EHWZHHQ WKH WZR YHUEV 7KDW LV JH HQWDLOV D FDUHIXO WUDFHDEOH PRYHPHQW WKDW KDV DQ GXUDWLRQ RI DFWLYLW\\@ D WD ED VKRXEL JHOH VKLMLGDR \\L VKL MXH[LQ &gt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gt;FRYHULQJ WHUP@ %XW ZKHQ JH LV FRQWUDVWHG ZLWK FL WR VWDE WKH SURSRVHG FRQWUDVWLQJ UHODWLRQ RI &gt;HIIHFW@ LV FOHDUO\\ HYLGHQW 5HVHDUFK 0HWKRGRORJ\\ DQG &amp;DVH 6WXGLHV HYHQWVWUXFWXUH DWWULEXWHV 5HVHDUFK 0HWKRGRORJ\\ H[FKDQJHDEOH LQ FHUWDLQ FRQWH[WV D EDLIDQJ TL]L VHWSXW FKHVVSLHFH WR SXW GRZQ FKHVV SLHFHV E EDLIDQJ \\L]L VHWSXW FKDLU WR SXW GRZQ FKDLUV +RZHYHU WKHUH DGMXQFW EXW IDQJ FDQQRW D WD ]KHQJ]DL EDL VKX VKH '85 VHW ERRN E \"WD ]KHQJ]DL IDQJ VKX E EDL 620(%2'&lt; \\L GDR D WLDQTL ELDQ OH VHPDQWLF H[SODQDWLRQ :H DOVR REVHUYH WKDW WKHUH LV D IXQGDPHQWDO GLIIHUHQFH LQ PDQQHU VDQ [LDRVKL &gt;ZLWK WKH LQWHQGHG LQWHUSUHWDWLRQ RI +RZHYHU WKH &gt;HIIHFW@ DFFRXQW PD\\ QRW RIIHU D FRPSOHWH DQG VXIILFLHQW OH[LFDO 6+H LV SXWWLQJ GRZQ WKH ERRNV QRZ D QD JH WDLVKL\\L EDL GRQJELDQFKDR GRQJ EDL WKDW &amp;/6 HDV\\FKDLU VHW HDVWVLGHWRZDUGV HDVW VHW 3XW WKDW HDV\\FKDLU VR WKDW LW IDFHV HDVW E QD JH WDLVKL\\L IDQJ GRQJELDQFKDR GRQJ IDQJ WKDW &amp;/6 HDV\\FKDLU SXW HDVWVLGHWRZDUGV HDVW SXW 7KH DERYH WKUHH D EDL MLD]L VHWVKHOI WR SXW RQ DLUV WR VHW XS D VWUHHW YHQGLQJ SRVLWLRQ E\\ VSUHDGLQJ D SLHFH RI FORWK RQ WKH JURXQG G EDL NXR VHWULFK WR VKRZ RII RQHV ZHDOWK 7KH DERYH LGLRPFRPSRXQG HYLGHQFH QRW RQO\\ RIIHUV DGGLWLRQDO VXSSRUW IRU WKH &gt;GHVLJQ@ 'LDJUDP 0$596 IRU EDL DQG IDQJ EDL \u2022 BBBB $JHQW 7KHPH /RFDWLRQ! _ &gt;GHVLJQ@ IDQJ \u2022 BBBB $JHQW 7KHPH /RFDWLRQ! D NXDLGLDQ EDQ KXUU\\XS PRYH 0RYH &gt;WKH WKLQJV@ IDVWHU E PDQPDQ JDL VORZO\\ UHYLVH 5HYLVHFRUUHFW VORZO\\ DQG FDUHIXOO\\ D WD ED VKRXEL JHOH VKLMLGDR \\L VKL MXH[LQ VKH %$ DUP *(3(5) WHQSOXVNQLIH VR VKRZ UHVROXWLRQ 6+H FXW PRUH WKDQ WHQ FXWV RQ KLVKHU DUP WR VKRZ KLVKHU UHVROXWLRQ YV E ]DL TLQJGL VKHQVKDQJ NDQOH ZXVKLOLX GDR DW ORYHIRH ERG\\WRS .$13(5) NQLIH &gt;7KH SHUVRQ@ KDFNHG KLVKHU ULYDO LQ ORYH DIIDLU WLPHV ,Q D [LDRKDL PR]KH EL]L FKLOG WRXFK'85 QRVH 7KH FKLOG LV WRXFKLQJ KLVKHU RZQ QRVH E [LDRKDL SHQJ]KH EL]L FKLOG WRXFK'85 QRVH D 7D PR OH EDQWLDQ VKHPH \\H PHL PR GDR VKH WRXFK 3(5) KDOIGD\\ ZKDW &lt;( 1(* WRXFK UHDFK 6+H JURSHG IRU D ORQJ WLPH EXW GLG QRW WRXFK DQ\\WKLQJ E 7D SHQJ OH EDQWLDQ VKHPH \\H PHL SHQJ GDR VKH WRXFK 3(5) KDOIGD\\ ZKDW &lt;( 1(* WRXFK UHDFK VDQJH EDR VKH KHDG EXPS3(5) WKUHH&amp;/6 EXPSV 6+H EXPSHG WKUHH EXPSV LQ WKH KHDG E FKH]L SHQJOH \\LJH GD GRQJ FDU EXPS3(5) RQH&amp;/6 ELJ KROH 7KHUH ZDV D ELJ KROH LQ WKH FDU DV D UHVXOW RI EXPSLQJ LQWR VRPHWKLQJ 7KH WHDFKHU FRUUHFWHG WKUHH ZULWLQJ DVVLJQPHQWV E WLDQTL ELDQ OH ZHDWKHU WUDQVIRUP 3(5) 7KH ZHDWKHU FKDQJHG D EDQFKX OLDQJ ]KDQJ \\L]L PRYHRXW WZR &amp;/6 FKDLU 7KH VLWXDWLRQ KDV LPSURYHGZRUVHQHG E 1L ]KHJH PDRELQJ \\LGLQJ \\DR JDL \\RX WKLV VKRUWFRPLQJ PXVW ZDQW FKDQJH &lt;RX PXVW LPSURYH E\\ JHWWLQJ ULG RI WKLV VKRUWFRPLQJ 7KH DERYH FRQWUDVWV FOHDUO\\ VKRZ WKDW WKH OH[LFDO VHPDQWLF VSHFLILFDWLRQ RI FDXVDWLYH HYHQWWUDQVLWLRQ KDV PDQ\\ PRUH LPSOLFDWLRQV WKHQ GR WKH VLPSOH DUJXPHQW VWUXFWXUH 0$596 UHSUHVHQWDWLRQ LQ 'LDJUDP 1RWH WKDW WKH V\\QWDFWLF UHDOL]DWLRQ FDQ KDYH HLWKHU D FRPSOHWH VHW RU D VXEVHW RI UROHV IRXQG LQ WKH OH[LFDO VHPDQWLF UHSUHVHQWDWLRQ FI ([DPSOH ZKLFK RQO\\ KDV WZR UROHV 7KH WZR VHWV D WD WRX SHQJOH &gt;VRPHRQH@ PRYHG WZR FKDLUV RXW E VKLWRX \\LGRQJ OH VWRQH PRYHPRYH 3(5) 7KH VWRQH PRYHG FKDQJHV SUHYLRXVO\\ VWXGLHG )RU LQVWDQFH WKH FXUUHQW H[SODQDWLRQ DOORZV OH[LFDOO\\ VSHF LILHG GLUHFWLRQ RI FKDQJHRIVWDWH ZKHUH JDL VSHFLILHV D FKDQJH RI VWDWH IRU WKH EHWWHU ZKLOH ELDQ KDV QR VXFK VSHFLILFDWLRQ 2XU VWXG\\ ZLOO VKRZ DJDLQ KRZ D OH[LFDO VHPDQWLF DWWULEXWH FDQ EH D SRZHUIXO H[SODQDWRU\\ WRRO 'LDJUDP 0$596 UHSUHVHQWDWLRQ RI JH DQG TLH JH \u2022 \u2022 $JHQW 7KHPH 0DQQHU! _ ,Q D UHVHQWHG DV IROORZV 6LQFH FDXVDWLYH DOWHUQDWLRQ KDV EHHQ WKRURXJKO\\ VWXGLHG LQ WKH OLWHUDWXUH ZH ZLOO IROORZ SUHYLRXV ZRUNV DQG FKDUDFWHUL]H WKH FRQWUDVW DV GLUHFWO\\ LQYROYLQJ &gt;HIIHFW@ &amp;DVH 6WXG\\ TLH YV JH 0DQQHU 'LDJUDP 0$596 UHSUHVHQWDWLRQ RI SHQJ DQG PR 3HQJ $JHQW *RDO! _ &gt;GHILQLWH@ PR \u2022 $JHQW *RDO! WD JHOH \\L NXDL URX VKH *(3(5) &amp;/6 PHDW 6+H PDGH D VOLFH RI PHDW TLH \u2022 \u2022 $JHQW 7KHPH! HYHQWVWUXFWXUHV ,Q YHUEV KDYH QR VXFK LPSOLFDWLRQ ,Q</td></tr></table>",
+                "type_str": "table",
+                "html": null,
+                "text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
+                "num": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-2003.json

@@ -0,0 +1,104 @@
+{
+    "paper_id": "O00-2003",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:11.130155Z"
+    },
+    "title": "",
+    "authors": [
+        {
+            "first": "Lian-Cheng",
+            "middle": [],
+            "last": "Chief",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Chu-Ren",
+            "middle": [],
+            "last": "Huang",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Keh-Jiann",
+            "middle": [],
+            "last": "Chen",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Mei-Chih",
+            "middle": [],
+            "last": "Tsai",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Li-Li",
+            "middle": [],
+            "last": "Chang",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-2003",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": ":KDW &DQ 1HDU 6\\QRQ\\PV 7HOO 8V\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "/ & &KLHI HW DO",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "7VDL 0HL&KLK &KX5HQ +XDQJ .HKMLDQQ &KHQ DQG .DWKOHHQ $KUHQV 7RZDUGV D 5HSUHVHQWDWLRQ RI 9HUEDO 6HPDQWLFV $Q $SSURDFK %DVHG RQ 1HDU 6\\QRQ\\PV &RPSX WDWLRQDO /LQJXLVWLFV DQG &KLQHVH /DQJXDJH 3URFHVVLQJ SS :KDW &DQ 1HDU 6\\QRQ\\PV 7HOO 8V\" / & &KLHI HW DO",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "7VDL 0HL&KLK &KX5HQ +XDQJ .HKMLDQQ &KHQ DQG .DWKOHHQ $KUHQV 7RZDUGV D 5HSUHVHQWDWLRQ RI 9HUEDO 6HPDQWLFV $Q $SSURDFK %DVHG RQ 1HDU 6\\QRQ\\PV &RPSX WDWLRQDO /LQJXLVWLFV DQG &KLQHVH /DQJXDJH 3URFHVVLQJ SS :KDW &DQ 1HDU 6\\QRQ\\PV 7HOO 8V\" / & &KLHI HW DO",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF0": {
+                "type_str": "table",
+                "text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
+                "html": null,
+                "num": null,
+                "content": "<table><tr><td colspan=\"2\">/ &amp; &amp;KLHI HW DO :KDW &amp;DQ 1HDU 6\\QRQ\\PV 7HOO 8V\" / &amp; &amp;KLHI HW DO :KDW &amp;DQ 1HDU 6\\QRQ\\PV 7HOO 8V\" / &amp; &amp;KLHI HW DO :KDW &amp;DQ 1HDU 6\\QRQ\\PV 7HOO 8V\" / &amp; &amp;KLHI HW DO</td></tr><tr><td colspan=\"2\">IDQJELDQ DQG HQWULHV RI ELDQOL :H ZLOO ILUVW SUHVHQW WKHLU V\\QWDFWLF EHKDYLRUV LQ VLWLRQDO REMHFW RI IDQJELDQ FDQ XQGHUJR LQYHUVLRQ DV LQ D DQG E ZKLOH ELDQOL GRHV QRW Q \u00b1\u00a4 7DEOH &amp;RRFFXUUHQFH ZLWK 1HJDWLYH 0DUNHU EX QRW D { ,&lt; \u00b1Q ' \u00e0 V $Q $GGLWLRQDO 3HUVSHFWLYH</td></tr><tr><td>VHFWLRQ DQG WKHQ WKHLU GLVWULEXWLRQDO GLIIHUHQFHV LQ VHFWLRQ DOORZ VXFK DOWHUQDWLRQ ELDQOL GH IDQJVKL Negation (preceded by bu 'not') Total instances KH]KL EDQVKLFKX IDQJELDQ PLQ]KRQJ FKXJXR $Q</td><td>JXDQJXDQJ</td></tr><tr><td colspan=\"2\">7KH 1HDU 6\\QRQ\\P 3DLU )DQJELDQ DQG %LDQOL 7KH PHPEHUV RI WKH QHDU V\\QRQ\\P SDLU IDQJELDQ DQG ELDQOL DUH XVHG WR GHILQH HDFK RWKHU FRQYHQLHQW GH ZD\\ fangbian 44 445 HVWDEOLVK RIILFH FRQYHQLHQW SHRSOH JRDEURDG YLVLW D \u00fb\u00e3 \u00f1V D &lt; V^\u00b1Q 1&lt; \u00f6\u00e3 FRQYHQLHQW ZD\\ bianli 0 125 (VWDEOLVKLQJ DQ RIILFH PDNHV LW FRQYHQLHQW IRU SHRSOH WR WUDYHO DEURDG OL[LDQJ GH FKDQJGL VKL OLQMLQ JRQJ]XR GLGLDQ IDQJELDQ \\XDQJRQJ FDQMLD LGHDO '( SODFH EH QHDU ZRUN SODFH FRQYHQLHQW ZRUNHU MRLQ ([DPSOHV DQG VKRZ WKDW WKLV SDLU RI QHDU V\\QRQ\\PV FDQ IXQFWLRQ DV QRPLQDO E z u Q^' |l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u00d6\u00eb \u00b1Q WLQJFKH IDQJELDQ SDUNLQJ FRQYHQLHQW 3DUNLQJ KHUH LV FRQYHQLHQW L Q MLDRWRQJ ELDQOL WUDIILF FRQYHQLHQW 7UDQVSRUWDWLRQ LV FRQYHQLHQW ,Q DGGLWLRQ WR WKHLU LQWUDQVLWLYH XVDJHV WKH\\ DOVR KDYH WUDQVLWLYH XVDJHV DV VKRZQ LQ VHQWHQFH DQG WKH PHHWLQJ E \u00fb\u00e3 1 \u00f1V h` V^1`\u00f6\u00e3 \u00b1Q OL[LDQJ GH FKDQJGL VKL OLQMLQ JRQJ]XR GLGLDQ \\XDQJRQJ FDQMLD IDQJELDQ LGHDO '( SODFH EH QHDU ZRUN SODFH ZRUNHUV MRLQ FRQYHQLHQW $Q LGHDO ORFDWLRQ LV QHDU WKH ZRUN SODFH DQG FRQYHQLHQW IRU ZRUNHUV WR MRLQ WKH PHHWLQJ D  \u00b6 z \u00b4^ \u00e3,C /\\ \\RX JH]KRQJ FKDQJSLQ ELDQOL [LDRIHL]KH [XDQJRX KDYH YDULRXV SURGXFW FRQYHQLHQW FRQVXPHU FKRRVHEX\\ 7KH YDULHW\\ RI SURGXFWV PDNHV VHOHFWLRQ FRQYHQLHQW IRU FRQVXPHUV E  \u00b6 z P PRGLI\\ PDQ\\ UXOH FRQYHQLHQW PRXQWDLQSHRSOH FXOWLYDWH B \u00b1Q OLDQ[L VKDQJ GH 0RGLI\\LQJ PDQ\\ UXOHV PDNHV LW FRQYHQLHQW IRU WKH DERULJLQHV WR FXOWLYDWH &gt;ODQG@ IDQJELDQ FRPPXQLFDWH LQ GH FRQYHQLHQFH FRQYHQLHQFH LQ FRPPXQLFDWLQJ !4 Verbal Predicates Nominal Modifiers Verbal Modifiers Nominalization Fangbian 445 77% 7% 5% 10% Bianli 125 44% 34% 0% 22% DQG UHSHDWHG KHUH IRU FRQYHQLHQW UHIHUHQFH ,Q VHQWHQFH D WKH PDLQ YHUE LV IDQJELDQ DQG WKH YHUEDO PHDQLQJ SURILOHV WKH ZKROH HPEHGGHG HYHQW SHRSOH JR DEURDG DQG YLVLW 7KH V\\QWDFWLF HYLGHQFH DV VKRZQ E\\ WKH FRQVWUXFWHG VHQWHQFHV D DQG E VXSSRUW WKLV DUJXPHQW EHFDXVH LQ D WKH D \u00fb\u00e3 \u00f1V D &lt; V^\u00b1Q 1&lt; \u00f6\u00e3 OL[LDQJ GH FKDQJGL VKL OLQMLQ JRQJ]XR GLGLDQ IDQJELDQ \\XDQJRQJ FDQMLD Q VKHQJKXR GH ELDQOL OLYLQJ GH FRQYHQLHQFH FRQYHQLHQFH LQ OLYLQJ SRVWYHUEDO HOHPHQW WKH SURSRVLWLRQDO HYHQW FDQ EH LQYHUWHG WR WKH SUHYHUEDO SRVLWLRQ ZKHUHDV LQ VHQWHQFH E VXFK DQ LQYHUVLRQ LV QRW DOORZHG D { ,&lt; ' \u00e0 `\u00b1Q VKH]KL EDQVKLFKX PLQ]KRQJ FKXJXR JXDQJXDQJ IDQJELDQ HVWDEOLVK RIILFH SHRSOH JRDEURDG YLVLW FRQYHQLHQW LGHDO '( SODFH EH QHDU ZRUN SODFH FRQYHQLHQW ZRUNHU MRLQ $Q LGHDO ORFDWLRQ LV QHDU WKH ZRUNLQJ SODFH DQG FRQYHQLHQW IRU ZRUNHUV WR MRLQ WKH PHHWLQJ E \u00fb\u00e3 1 \u00f1V h` V^1`\u00f6\u00e3 \u00b1Q OL[LDQJ GH FKDQJGL VKL OLQMLQ JRQJ]XR GLGLDQ \\XDQJRQJ FDQMLD IDQJELDQ ,Q YHQLHQW\" (VWDEOLVKLQJ DQ RIILFH PDNHV LW FRQYHQLHQW IRU SHRSOH WR WUDYHO DEURDG LGHDO '( SODFH EH QHDU ZRUN SODFH ZRUNHUV MRLQ FRQYHQLHQW \u00b4^\u00e3,C /\\ \\RX JH]KRQJ FKDQJSLQ [LDRIHL]KH [XDQJRX ELDQOL KDYH YDULRXV SURGXFW FRQYHQLHQW FRQVXPHU FKRRVHEX\\ :H ZLOO DFFRXQW IRU WKLV SKHQRPHQRQ LQ VHFWLRQ 'LVWULEXWLRQDO 'LIIHUHQFHV LQ WHUPV RI WKH 7UDQVLWLYH ,QWUDQVLWLYH E z u^' $Q LGHDO ORFDWLRQ LV QHDU WKH ZRUNLQJ SODFH DQG FRQYHQLHQW IRU ZRUNHUV WR MRLQ |l Q $OWHUQDWLRQ [LXJDL VKXGXR IDJXL VKDQPLQ NHQ]KL WKH PHHWLQJ ELDQOL PRGLI\\ PDQ\\ UXOH PRXQWDLQSHRSOH FXOWLYDWH FRQYHQLHQW 6\\QWDFWLF 3DWWHUQV D  \u00b6 z \u00b4^ \u00e3,C /\\ \\RX JH]KRQJ FKDQJSLQ ELDQOL [LDRIHL]KH [XDQJRX %DVHG RQ WKH WZR VHPDQWLF IHDWXUHV WKH EHQHILFLDO UROH DQG WKH OH[LFDO FRQFHSWXDO SURILOH ZH SURSRVH WKDW IDQJELDQ DQG ELDQOL KDYH GLIIHUHQW HYHQW VWUXFWXUHV DQG DUJXPHQW KDYH YDULRXV SURGXFW FRQYHQLHQW FRQVXPHU FKRRVHEX\\ 2WKHU 6\\QWDFWLF )XQFWLRQV RI IDQJELDQ DQG ELDQOL 7KH YDULHW\\ RI SURGXFWV PDNHV VHOHFWLRQ FRQYHQLHQW IRU FRQVXPHUV VWUXFWXUH IUDPHV { ,`\u00b1Q ' \u00e0 V KH]KL EDQVKLFKX IDQJELDQ PLQ]KRQJ FKXJXR Transitive IDQJELDQ &gt;$*(17 *2$/ 3URSRVLWLRQ@ E  \u00b6 z \u00b4^\u00e3,C /\\ Intransitive JXDQJXDQJ ELDQOL DV QRPLQDO PRGLILHUV Fangbian 342 31% 69% \\RX JH]KRQJ FKDQJSLQ [LDRIHL]KH [XDQJRX ELDQOL</td></tr><tr><td colspan=\"2\">SRLQW WKDW WKH SUHVHQW VWXG\\ DLPV WR VXSSRUW HVWDEOLVK RIILFH FRQYHQLHQW SHRSOH (VWDEOLVKLQJ DQ RIILFH PDNHV LW FRQYHQLHQW IRU SHRSOH WR WUDYHO DEURDG JRDEURDG YLVLW z u Q^' |l \u00b1Q Bianli 55 53% 47% 68%-;&amp;203! KDYH YDULRXV SURGXFW FRQYHQLHQW FRQVXPHU FKRRVHEX\\ \u00f7 IDQJELDQ GH ]L[XQ FRQYHQLHQW GH 6HQWHQFHV GHPRQVWUDWH WKDW SRVWYHUEDO HOHPHQWV RI IDQJELDQ FDQ XQGHUJR 7DEOH 'LVWULEXWLRQDO 'LIIHUHQFH LQ WHUPV RI WKH 7\\SH RI 2EMHFW ELDQOL &gt;$*(17 %(1 *2$/ 3URSRVLWLRQ @ LQYHUVLRQ ZKHUHDV WKRVH RI ELDQOL FDQQRW 6LQFH ELDQOL KDV WZR SRVWYHUEDO HOHPHQWV RQH LQIRUPDWLRQ [LXJDL VKXGXR IDJXL ELDQOL VKDQPLQ NHQ]KL HDVLO\\DFFHVVLEOH LQIRUPDWLRQ Sentential or Verbal Object Complex Nominal Object RI WKH JUDPPDWLFDO IXQFWLRQV FDQQRW EH LQYHUWHG E\\ LWVHOI 2Q WKH FRQWUDU\\ IDQJELDQ KDV 7KH 'DWD 7KH GDWD XVHG LQ WKLV VWXG\\ ZHUH WDNHQ IURP WKH 6LQLFD &amp;RUSXV YHUVLRQ ZKLFK FRQWDLQV PLOOLRQ WDJJHG &amp;KLQHVH ZRUGV ,Q WKLV FRUSXV ZH IRXQG HQWULHV RI PRGLI\\ PDQ\\ UXOH FRQYHQLHQW PRXQWDLQSHRSOH FXOWLYDWH 0RGLI\\LQJ PDQ\\ UXOHV PDNHV LW FRQYHQLHQW IRU WKH DERULJLQHV WR FXOWLYDWH &gt;ODQG@ ,Q WKHLU LQWUDQVLWLYH XVDJHV ERWK IDQJELDQ DQG ELDQOL WDNH D SURSRVLWLRQ DV D VXEMHFW ,Q WKHLU WUDQVLWLYH XVDJHV WKH\\ WDNH D SURSRVLWLRQDO REMHFW 8VXDOO\\ WKH SURSRVLWLRQDO VXE MHFW RU SURSRVLWLRQDO REMHFW LV UHSUHVHQWHG E\\ D FODXVH D YHUE SKUDVH RU D FRPSOH[ QRPLQDO HOHPHQW 7KH SURSRVLWLRQ GHVFULEHV ZKDW LV FRQYHQLHQW +RZHYHU WKH SURSR +RZHYHU ZH RQO\\ IRXQG H[DPSOHV RI ELDQOL EXW QRW IDQJELDQ XVHG LQ QRPLQDO FRPSRXQGV LQ WKH 6LQLFD &amp;RUSXV DV VKRZQ EHORZ :H GR QRW DFFRXQW IRU WKLV GLIIHUHQFH LQ WKLV SDSHU Q \u00f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u00f1\u00d0 ] \u00b1Q &lt;\u00fb ? Fangbian 107 90% 10% Bianli 29 62% RQO\\ RQH SRVWYHUEDO HOHPHQW ,Q EULHI WKH V\\QWDFWLF SURILOH FDQQRW FRQWUDGLFW WKH OH[LFD 37% FRQFHSWXDO SURILOH</td></tr></table>"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-2004.json

@@ -0,0 +1,39 @@
+{
+    "paper_id": "O00-2004",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:02.384973Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-2004",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "$ 6WXG\\ RQ 0DQGDULQ 9HUEV RI (PRWLRQ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {},
+        "ref_entries": {
+            "TABREF0": {
+                "content": "<table><tr><td>/ / &amp;KDQJ HW DO / / &amp;KDQJ HW DO</td></tr><tr><td>-897,.9 \u00e9 \u00f8 \u00f5 \u00f9 \u00d8 \u00fc \u00f4\u00f7 \u00fb\u00f8\u00fc \u00d9\u00f8\u00f8\u00f6 \u00f9 \u00e2\u00f6\u00f6\u00f8\u00f6\u00f8 \u00fc \u00fb\u00f8 ae\u00fc\u00fc\u00f6\u00f4 \u00d6 \u00fe\u00f4\u00fc\u00c7\u00ff\u00f8\u00c7 v9\u00bb\u00cc\u00c7\u00c5\u00bc\u00bf \u00fa\u00f4\u00c4\u00fc\u00fa\u00c7 \u00cf\u00bb\u00c9\u00c9\u00cc\u00bc\u00bf \u00c5\u00fe\u00f4\u00fc\u00c7 -v\u00bb\u00c5\u00ca\u00c4\u00bc\u00bf \u00ff\u00f8\u00c7 9\u00bb\u00c5\u00c9\u00c7\u00bc\u00bf \u00fcAE\u00f8\u00c7 \u00dbc\u00bb\u00c4\u00c8\u00c9\u00bc\u00bf \u00fe\u00f4\u00fc\u00c4\u00fc\u00c4 r\u00a8\u00bb\u00c4\u00c8\u00c5\u00bc\u00bf \u00fb\u00f4\u00c4\u00ff\u00f8\u00c7 9\u00bb\u00c4\u00c7\u00c4\u00bc\u00bf \u00fb\u00f4\u00c4\u00fcAE \u00db\u00bb\u00c4\u00c3\u00ca\u00bc\u00bf \u00fe\u00f4\u00fc\u00c7\u00fb\u00c5 vR\u00bb\u00c7\u00cb\u00bc\u00bf \u00fa\u00c7\u00fe\u00f4\u00fc\u00c7 \u00c0v\u00bb\u00c7\u00c3\u00bc \u00fa\u00c7\u00feAE \u00c0\u00c7\u00bb\u00c7\u00c7AE\u00bc\u00bf \u00f4\u00fa\u00c5\u00fa\u00c7 \u00c4\u00bb\u00c5AE\u00c5\u00bc\u00bf \u00f6\u00fb\u00f8\u00c5\u00fb\u00fa\u00c7  \u00a7\u00f9\u00bb\u00cbAE\u00bc\u00bf \u00fdAE\u00f4\u00fa\u00c7 \u00dc \u00bb\u00c9\u00c5\u00bc\u00bf \u00fa\u00c7\u00fc\u00c4 \u00c0\u00a8\u00bb\u00c7\u00cb\u00bc \u00fb\u00f4\u00fa\u00c4\u00fc\u00c4 \u00a8\u00bb\u00c4AE\u00c7\u00bc\u00bf \u00f5\u00f8\u00fc\u00c4\u00fb\u00f4\u00fa\u00c4 \u00bb\u00c8\u00c5\u00bc \u00fc\u00c5\u00fb\u00f4\u00c7 J\u00bb\u00c4\u00cc\u00cb\u00bc\u00bf \u00fb\u00c7\u00fb\u00fcAE \u00f5a\u00bb\u00c4\u00c3\u00c5\u00bc \u00fb\u00f8\u00fa\u00c4\u00fc\u00c7 !\u00db\u00bb\u00c5\u00cc\u00c8\u00bc\u00bf \u00fc\u00c7 \u00db\u00bb\u00c4\u00c5\u00c9\u00bc\u00bf \u00f9\u00f8\u00c7\u00c7 ,\u00f7\u00bb\u00c4\u00c4\u00c5\u00bc\u00bf \u00fc\u00c7\u00f9\u00f8\u00c7 \u00db,\u00bb\u00c7\u00cc\u00bc \u00f4\u00c7 \u00a2\u00bb\u00c8\u00c7\u00cb\u00bc\u00bf \u00fb\u00f4\u00fc\u00c7\u00f4\u00c7 m\u00a2\u00bb\u00c5\u00c9\u00c4\u00bc\u00bf \u00fe\u00faAE\u00fd\u00c7 0\u00bb\u00c4\u00c7\u00cc\u00bc\u00bf \u00f8\u00fc\u00c7\u00fd\u00c7 y0\u00bb\u00c7\u00c3\u00bc \u00f7\u00f4\u00c4\u00fc\u00c4 \u00a8\u00bb\u00c9\u00c3\u00cc\u00bc\u00bf \u00f9\u00f4\u00c5\u00f4AE (*\u00bb\u00c4\u00cc\u00cc\u00bc\u00bf \u00f7\u00f4\u00c4\u00c4 !\u00bb\u00c9\u00c7\u00bc\u00bf\u00f9\u00f4\u00c5(\u00bb\u00c8\u00c7\u00bc\u00bf \u00c4\u00fc\u00c4 !\u00a8\u00bb\u00c7\u00c9\u00bc\u00bf \u00feAE\u00f4AE \u00c7*\u00bb\u00c7\u00c8\u00bc ,QLWLDO REVHUYDWLRQV DQG WKHRUHWLFDO DVVXPSWLRQV ae\u00f5\u00f8 \u00db\u00f4\u00fc\u00f8 \u00d7\u00f8\u00f8\u00fc ae\u00f4\u00f7\u00f8 \u00e5\u00f8\u00fa\u00f8 \u00d4\u00fa\u00f8 \u00d9\u00f8\u00f4 \u00ea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ae\u00f5\u00f8 \u00e7 \u00f8 \u00d4 \u00e7\u00f8 \u00d5 \u00db\u00f4\u00fc\u00f8 \u00fa\u00f4\u00c4\u00fc\u00fa\u00c7 \u00cf\u00bb\u00c9\u00c9\u00cc\u00bc \u00fe\u00f4\u00fc\u00c4\u00fc\u00c4 H\u00a8\u00bb\u00c4\u00c8\u00c5\u00bc \u00fa\u00c7\u00fe\u00f4\u00fc\u00c7 \u00c0L\u00bb\u00c7\u00c3\u00bc \u00fe\u00f4\u00fc\u00c7\u00ff\u00f8\u00c7 L9\u00bb\u00cc\u00c7\u00c5\u00bc \u00c5\u00fe\u00f4\u00fc\u00c7 -L\u00bb\u00c5\u00ca\u00c4\u00bc \u00fcAE\u00f8\u00c7 \u00dbc\u00bb\u00c4\u00c8\u00c9\u00bc \u00fb\u00f4\u00c4\u00ff\u00f8\u00c7 9\u00bb\u00c4\u00c7\u00c4\u00bc \u00fb\u00f4\u00c4\u00fcAE \u00db\u00bb\u00c4\u00c3\u00ca\u00bc \u00fe\u00f4\u00fc\u00c7\u00fb\u00c5 L4\u00bb\u00c7\u00cb\u00bc \u00d7\u00f8\u00f8\u00fc \u00f4\u00c5\u00fa\u00c7 \u00c4\u00bb\u00c5AE\u00c5\u00bc \u00fa\u00c7\u00fc\u00c4 \u00c0\u00a8\u00bb\u00c7\u00cb\u00bc \u00e7\u00fa\u00c7\u00feAE \u00c0\u00c7\u00bb\u00c7\u00c7AE\u00bc \u00f6\u00fb\u00f8\u00c5\u00fb\u00fa\u00c7  \u00a7\u00f9\u00bb\u00cbAE\u00bc \u00fdAE\u00f4\u00fa\u00c7 \u00dc\u00bb\u00c9\u00c5\u00bc ae\u00f4\u00f7\u00f8 \u00fb\u00f4\u00fa\u00c4\u00fc\u00c4 \u00a8\u00bb\u00c4AE\u00c7\u00bc \u00f5\u00f8\u00fc\u00c4\u00fb\u00f4\u00fa\u00c4 \u00bb\u00c8\u00c5\u00bc \u00e5\u00f8\u00fa\u00f8 \u00fb\u00c7\u00fb\u00fcAE \u00f5\u00bb\u00c4\u00c3\u00c5\u00bc \u00fc\u00c5\u00fb\u00f4\u00c7 J\u00bb\u00c4\u00cc\u00cb\u00bc \u00d4\u00fa\u00f8 \u00fb\u00f8\u00fa\u00c4\u00fc\u00c7 !\u00db\u00bbAE\u00c3\u00ca\u00bc \u00f9\u00f8\u00c7\u00c7 ,\u00f7\u00bb\u00c4\u00c4\u00c5\u00bc \u00fc\u00c7\u00f9\u00f8\u00c7 \u00db,\u00bb\u00c7\u00cc\u00bc \u00d9\u00f8\u00f4 \u00fb\u00f4\u00fc\u00c7\u00f4\u00c7 m\u00a2\u00bb\u00c5\u00c9\u00c4\u00bc \u00fe\u00faAE\u00fd\u00c7 0\u00bb\u00c4\u00c7\u00cc\u00bc \u00f8\u00fc\u00c7\u00fd\u00c7 y0\u00bb\u00c7\u00c3\u00bc \u00ea \u00f7\u00f4\u00c4\u00fc\u00c4 \u00a8\u00bb\u00c9\u00c3\u00cc\u00bc \u00f7\u00f4\u00c4\u00c4 !\u00bb\u00c9\u00c7\u00bc \u00c4\u00fc\u00c4 !\u00a8\u00bb\u00c7\u00c9\u00bc \u00f9\u00f4\u00c5\u00f4AE (*\u00bb\u00c4\u00cc\u00cc\u00bc \u00feAE\u00f4AE \u00c7*\u00bb\u00c7\u00c8\u00bc 7KH GHILQLWLRQ RI D 6HPDQWLF )LHOG DFFRUGLQJ WR *UDQG\\ &gt;@ LV DV IROORZV L &gt;$ VHPDQWLF ILHOG@ LV D VHW LQFOXGLQJ RQH RU PRUH FRQWUDVW VHWV DQG SRVVLEO\\ DOVR LQFOXGLQJ SHUPXWDWLRQ UHODWLRQV VXFK WKDW DW PRVW RQH FRYHULQJ WHUP GRHV QRW RFFXU DV DQ HOHPHQW RI D FRQWUDVW VHW LQ WKH VHPDQWLF ILHOG H[FHSW IRU WKH FRYHULQJ WHUP DQ\\ H[SUHVVLRQ WKDW RFFXUV LQ D FRQWUDVW VHW ZLWK DQ HOHPHQW RI WKH VHPDQWLF ILHOG LV DOVR LQ WKH ILHOG ,Q DGGLWLRQ ZH UHLQWHUSUHW *UDQG\\V &gt;@ IRUPDO GHILQLWLRQ RI D &amp;RQWUDVW 6HW EHORZ LL $ FRQWUDVW VHW ZLOO FRQVLVW PLQLPDOO\\ RI D FRYHULQJ WHUP 7 D VHW RI IXQGDPHQWDO FRQWUDVW UHODWLRQV DQG D VHW RI OLQJXLVWLF WHUPV VXFK WKDW WKHUH DUH FRPPRQ OLQJXLVWLF EHOLHIV WKDW HDFK OLQJXLVWLF WHUP LQ WKH VHW LV D NLQG RI 7 WKDW LV WKH UHODWLRQ EHWZHHQ DQ\\ WHUP DQG 7 FDQ EH GHILQHG E\\ WKH LVD UHODWLRQ IRU DQ\\ WZR GLIIHUHQW WHUPV LQ WKH VHW LW LV D FRPPRQ OLQJXLVWLF EHOLHI WKDW WKH\\ FRQWUDVW LQ WHUPV RI D VLQJOH UHODWLRQ ZKLFK LV GHILQHG E\\ WKH VHW RU LV GHULYDEOH IURP WKH UHODWLRQV GHILQHG LQ WKH VHW 7KHRUHWLFDO 3UHPLVH &amp;RQWUDVWEDVHG VHPDQWLF ILHOGV 7KH IDFW WKDW LQ HDFK RI WKH ILHOGV RI HPRWLRQV ZH KDYH H[DPLQHG WKH WZR PRVW IUHTXHQW DQG WKHUHIRUH PRVW GRPLQDQW WHUPV IRUP D FRQWUDVW SDLU OHDGV XV WR DGRSW D UHYLVLRQ RI *UDQG\\V &gt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u00ef ln 1 \u00b4^\u00f0 J^~\u00a3 \u00a4 v\u00e0] KHZHL \\LVKXMLD GH ]XRSLQ KHQ \\LKDQGL MLQQLDQ ZXID ]KDQFKX WKLV DUWLVW V ZRUNV YHU\\ UHJUHWIXOO\\ WKLV \\HDU FRXOGQW H[KLELW ,7\\SH $ 7RWDO 3UHG 1RP 10 $GMXQFW &amp;RPS (OVH JDR[LQJ \u00cf QDQJXR \u00c4 VKDQJ[LQ \u00a8 KRXKXL \u00f5 VKHQJTL !\u00db KDLSD m\u00a2 GDQ[LQ \u00a8 $YHUDJH \u00e7\u00f8 \u00d5 \u00e7\u00f4\u00ff \u00e3\u00f8\u00f7\u00c1 \u00e1 \u00c1 \u00e1\u00c1\u00e0\u00c1 \u00d4\u00f7\u00fd\u00f6 \u00d6 \u00c1 \u00d8\u00ff\u00f8 \u00fe\u00f4\u00fc\u00c7\u00ff\u00f8\u00c7 LW \u00cc\u00c7\u00c5 AE\u00ca\u00c1\u00ca\u00cc\u00b8\u00c5\u00c9\u00c1\u00c7AE\u00b8\u00c5\u00c7\u00c1\u00cb\u00c7\u00b8\u00c8\u00c1\u00caAE\u00b8\u00c8\u00c1\u00c5\u00c3\u00b8\u00c3\u00c1\u00c3\u00c3 \u00fa\u00c7\u00feAE \u00c0\u00c7 \u00c7\u00c7AE \u00c5\u00c8\u00c1\u00caAE\u00b8\u00c7\u00c8\u00c1\u00c9\u00c3\u00b8\u00c5\u00c3\u00c1\u00c8\u00c7\u00b8\u00c9\u00c1\u00c3\u00cc\u00b8\u00c5\u00c1\u00c3AE\u00b8\u00c3\u00c1\u00c3\u00c3\u00f5 \u00f8\u00fc\u00c4\u00fb\u00f4\u00fa\u00c4 \u00c8\u00c5 \u00c7\u00c3\u00c1AE\u00cb\u00b8\u00c5\u00cb\u00c1\u00cb\u00c8\u00b8\u00c4\u00cc\u00c1\u00c5AE\u00b8\u00cc\u00c1\u00c9\u00c5\u00b8\u00c4\u00c1\u00cc\u00c5\u00b8\u00c3\u00c1\u00c3\u00c3 \u00fc\u00c5\u00fb\u00f4\u00c7 J \u00c9 \u00c4\u00cc\u00cb AE\u00c7\u00c1\u00cb\u00c8\u00b8AEAE\u00c1\u00cb\u00c7\u00b8AE\u00c1\u00c8\u00c7\u00b8\u00c7\u00c1\u00c3\u00c7\u00b8\u00c3\u00c1\u00c3\u00c3\u00b8\u00c5AE\u00c1\u00ca\u00c7\u00f9 \u00f8\u00c7\u00c7 ,\u00f7 \u00c4\u00c4\u00c5 \u00c5\u00cb\u00c1\u00c8\u00ca\u00b8AE\u00ca\u00c1\u00c8\u00c3\u00b8\u00c4\u00ca\u00c1\u00cb\u00c9\u00b8\u00c4\u00c9\u00c1\u00c3\u00ca\u00b8\u00c3\u00c1\u00c3\u00c3\u00b8\u00c3\u00c1\u00c3\u00c3\u00fe \u00faAE\u00fd\u00c7 0 \u00c4\u00c7\u00cc \u00c5AE\u00c1\u00c7\u00cc\u00b8\u00c9\u00cb\u00c1\u00c7\u00c9\u00b8\u00ca\u00c1AE\u00cb\u00b8\u00c5\u00c1\u00c3\u00c7\u00b8\u00c3\u00c1\u00c3\u00c3\u00b8\u00c3\u00c1\u00c3\u00c3\u00f9 \u00f4\u00c5\u00f4AE (* \u00c4\u00cc\u00cc \u00c5\u00c7\u00c1\u00c4\u00c5\u00b8\u00c9\u00cc\u00c1\u00cb\u00c8\u00b8\u00c9\u00c1\u00c3AE\u00b8\u00c3\u00c1\u00c3\u00c3\u00b8\u00c3\u00c1\u00c3\u00c3\u00b8\u00c3\u00c1\u00c3\u00c3\u00d4 \u00f8\u00f4\u00fa\u00f8 \u00c5\u00cc\u00cc AE\u00c3\u00c1\u00ca\u00c3\u00b8\u00c7\u00c7\u00c1AE\u00c9\u00b8\u00c4\u00c7\u00c1\u00c5\u00c4\u00b8\u00c9\u00c1\u00c5AE\u00b8\u00c4\u00c1AE\u00c4\u00b8AE\u00c1AE\u00cc /LNHOLKRRG UDWLR PHDVXULQJ WKH HQFRGLQJ SUHIHUHQFH 7DEOH 7KH /LNHOLKRRG 5DWLR RI 'RPLQDQW 7\\SH RYHU 1RQGRPLQDQW 7\\SH LQ WHUPV RI 3UHGLFDWLYH DQG 1RPLQDO )XQFWLRQV 7\\SH $7\\SH % YHUEV 3UHGLFDWH )UHTXHQF\\ 5DWLR RI $ RYHU % 1RPLQDO )UHTXHQF\\ 5DWLR RI % RYHU $ JDR[LQJ \u00cfNXDLOH LW QDQJXR \u00c4WRQJNX \u00c0\u00c7 6KDQJ[LQ \u00a8EHLVKDQJ KRXKXL \u00f5\\LKDQ J VKHQJTL !\u00dbIHQQX ,\u00f7 KDLSD m\u00a2NRQJMX 0 GDQ[LQ \u00a8IDQQDR(* $YHUDJH UDWLR 7DEOH 9HUEV RI (PRWLRQ 6RUWHG $FFRUGLQJ WR 'HYHUEDO 8VHV \u00e7\u00f8 \u00d4 \u00e9\u00f8\u00f5 \u00e1 \u00c1 \u00e1\u00c1\u00e0\u00c1 \u00f7\u00f8 \u00f8\u00f5\u00f4\u00ff \u00e7\u00f8 \u00d5 \u00e9\u00f8\u00f5 \u00e1 \u00c1 \u00e1\u00c1\u00e0\u00c1 \u00f7\u00f8 \u00f8\u00f5\u00f4\u00ff \u00fa\u00c7\u00fe \u00f4\u00fc\u00c7\u00c0 v \u00c3\u00c1\u00c3\u00c3\u00b8\u00c3\u00c1\u00c3\u00c3\u00b8\u00c3\u00c1\u00c3\u00c3\u00b8\u00fc\u00c7\u00f9\u00f8\u00c7\u00db , \u00c5\u00c3\u00c1\u00c7\u00c4\u00b8\u00c7\u00c1\u00c3\u00cb\u00b8\u00c5\u00c7\u00c1\u00c7\u00cc\u00fa \u00f4 \u00c4 \u00fc\u00fa\u00c7\u00cf \u00c3\u00c1AE\u00c3\u00b8\u00c4\u00c1AE\u00c8\u00b8\u00c4\u00c1\u00c9\u00c8\u00b8\u00f8\u00fc\u00c7\u00fd\u00c7y H \u00c5\u00c5\u00c1\u00c8\u00c3\u00b8\u00c5\u00c1\u00c8\u00c3\u00b8\u00c5\u00c8\u00c1\u00c3\u00c3\u00fb \u00c7\u00fb\u00fcAE\u00f5 \u00c3\u00c1\u00cc\u00cb\u00b8\u00c5\u00c1\u00cc\u00c7\u00b8\u00c5\u00c1\u00cc\u00c7\u00b8\u00c5\u00fe\u00f4\u00fc\u00c7-v \u00ca\u00c1\u00ca\u00c8\u00b8\u00c5\u00c5\u00c1\u00c4\u00c7\u00b8\u00c5\u00cc\u00c1\u00cb\u00cc\u00f7 \u00f4 \u00c4 \u00fc\u00c4\u00a8\u00c4\u00c1\u00cc\u00ca\u00b8\u00c4\u00c1AE\u00c4\u00b8AE\u00c1\u00c5\u00cb\u00b8\u00fb\u00f4\u00c4\u00fc\u00c4 \u00db \u00c5\u00c4\u00c1\u00c8\u00c3\u00b8\u00cc\u00c1AE\u00c8\u00b8AE\u00c3\u00c1\u00cb\u00c7 \u00fb \u00f8 \u00fa\u00c4 \u00fc\u00c7! \u00db \u00c3\u00c1\u00c3\u00c3\u00b8AE\u00c1\u00c8\u00cb\u00b8AE\u00c1\u00c8\u00cb\u00b8\u00fe\u00f4\u00fc\u00c7\u00fb\u00c5v R \u00c9\u00c1\u00c5\u00c8\u00b8\u00c5\u00ca\u00c1\u00c3\u00cb\u00b8AEAE\u00c1AEAE \u00fa\u00c7\u00fc\u00c4\u00c0\u00a8\u00c5\u00c1\u00c3\u00cb\u00b8\u00c5\u00c1\u00c3\u00cb\u00b8\u00c7\u00c1\u00c4\u00ca\u00b8\u00fdAE\u00f4\u00fa\u00c7 \u00dc \u00c5\u00c3\u00c1\u00cc\u00ca\u00b8\u00c4\u00c5\u00c1\u00cc\u00c3\u00b8AEAE\u00c1\u00cb\u00ca \u00f4 \u00c5\u00fa\u00c7\u00c4 \u00c5\u00c1\u00c4\u00c9\u00b8\u00c5\u00c1\u00c8\u00cc\u00b8\u00c7\u00c1\u00ca\u00c8\u00b8\u00fc\u00c5\u00fb\u00f4\u00c7J AEAE\u00c1\u00cb\u00c7\u00b8AE\u00c1\u00c8\u00c7\u00b8AE\u00ca\u00c1AE\u00cb\u00fb \u00f4\u00fc\u00c7 \u00f4 \u00c7m \u00a2 AE\u00c1\u00c3\u00ca\u00b8\u00c5\u00c1\u00c9\u00cb\u00b8\u00c8\u00c1\u00ca\u00c8\u00b8\u00feAE\u00f4AE\u00c7 * AE\u00c8\u00c1\u00c8\u00c9\u00b8\u00c4\u00c4\u00c1\u00c4\u00c4\u00b8\u00c7\u00c9\u00c1\u00c9\u00ca \u00c4\u00fc\u00c4!\u00a8\u00c9\u00c1\u00c8\u00c5\u00b8\u00c3\u00c1\u00c3\u00c3\u00b8\u00c9\u00c1\u00c8\u00c5\u00b8\u00f5\u00f8\u00fc\u00c4\u00fb\u00f4\u00fa\u00c4 \u00c5\u00cb\u00c1\u00cb\u00c8\u00b8\u00c4\u00cc\u00c1\u00c5AE\u00b8\u00c7\u00cb\u00c1\u00c3\u00cb\u00fe \u00f4\u00fc\u00c4 \u00fc \u00c4 r\u00a8\u00c4\u00c1\u00cc\u00ca\u00b8\u00c8\u00c1\u00cc\u00c5\u00b8\u00ca\u00c1\u00cb\u00cc\u00b8\u00f6\u00fb\u00f8\u00c4\u00fb\u00fa\u00c7 \u00a7\u00f9 \u00c3\u00c1\u00c3\u00c3\u00b8\u00c7\u00cb\u00c1\u00c4\u00cc\u00b8\u00c7\u00cb\u00c1\u00c4\u00cc\u00f7 \u00f4 \u00c4 \u00c4 ! \u00cc\u00c1AE\u00cb\u00b8\u00c3\u00c1\u00c3\u00c3\u00b8\u00cc\u00c1AE\u00cb\u00b8\u00fe\u00f4\u00fc\u00c7\u00ff\u00f8\u00c7v W \u00c5\u00c9\u00c1\u00c7AE\u00b8\u00c5\u00c7\u00c1\u00cb\u00c7\u00b8\u00c8\u00c4\u00c1\u00c5\u00ca \u00fb \u00f4 \u00fa\u00c4 \u00fc\u00c4\u00a8\u00c5\u00c1\u00cc\u00cc\u00b8\u00c4\u00c4\u00c1\u00c4\u00cc\u00b8\u00c4\u00c7\u00c1\u00c4\u00cb\u00b8\u00f9\u00f8\u00c7\u00c7, \u00f7 AE\u00ca\u00c1\u00c8\u00c3\u00b8\u00c4\u00ca\u00c1\u00cb\u00c9\u00b8\u00c8\u00c8\u00c1AE\u00c9 \u00fa\u00c7\u00feAE\u00c0 \u00c7 \u00c7\u00c8\u00c1\u00c9\u00c3\u00b8\u00c5\u00c3\u00c1\u00c8\u00c7\u00b8\u00c9\u00c9\u00c1\u00c4\u00c7\u00fe \u00faAE\u00fd\u00c7 H \u00c9\u00cb\u00c1\u00c7\u00c9\u00b8\u00ca\u00c1AE\u00cb\u00b8\u00ca\u00c8\u00c1\u00cb\u00c7\u00f9 \u00f4 \u00c5\u00f4AE( * \u00c9\u00cc\u00c1\u00cb\u00c8\u00b8\u00c9\u00c1\u00c3AE\u00b8\u00ca\u00c8\u00c1\u00cb\u00cb \u00fc\u00c4\u00f8 \u00c7\u00db \u00cc\u00c3\u00c1AE\u00cb\u00b8\u00c4\u00c1\u00cc\u00c5\u00b8\u00cc\u00c5\u00c1\u00c5\u00c3\u00fb \u00f4\u00c4\u00ff\u00f8\u00c7 W 7\\SH $ \" \u00cf1 \u00d9\u00a3 \" \u00cf1 &amp;\u00d7 \" \u00cf1 * \" \u00cf1 \u00e9| JDR[LQJGH WRQJQLDQ JDR[LQJGH KXQ\\LQ JDR[LQJGH VKDQJEDQ]X JDR[LQJGH KXDQMLQJ JODG FKLOGKRRG JODG PDUULDJHJODG ZRUNHUV JODG HQYLURQPHQW 7\\SH % L91 \u00d9\u00a3 L91 &amp;\u00d7 L91 * L91 \u00e9| NXDLOHGH WRQJQLDQ NXDLOHGH KXQ\\LQ NXDLOHGH VKDQJEDQ]X NXDLOHGH KXDQMLQJ KDSS\\ FKLOGKRRG KDSS\\ PDUULDJH KDSS\\ ZRUNHUV KDSS\\ HQYLURQPHQW KDSS\\ FKLOGKRRG KDSS\\ PDUULDJH KDSS\\ ZRUNHUV KDSS\\ HQYLURQPHQW :FRPPDQG RU HYDOXDWLRQ PHDQLQJ 3UDJPDWLFDOO\\ VSHDNLQJ ERWK FRQVWUXFWLRQV ZLWK YHUEV RI HPRWLRQ KDYH WKH VDPH GLVVXDGLQJ IXQFWLRQ \u00a8 7 \u00a8 l\u00de \u00a8 ELH VKDQJ[LQ PR VKDQJ[LQ EX\\DR VKDQJ[LQ GRQW VDG GRQW VDG GRQW VDG 3OHDVH GRQW IHHO VDG l 5 \u00a8 \u00b2 y\u00d6 ] \u00a8 EX ]KLGH VKDQJ[LQ PHL VKHPH KDR VKDQJ[LQ GH 1(* ZRUWK VDG ZLWKRXW DQ\\WKLQJ ZRUWK VDG 3$57,&amp;/( ,W LV QRW ZRUWKZKLOH WR IHHO VDG 7KHUHV QRWKLQJ WR EH VDG DERXW 3OHDVH GRQW IHHO VDG %DVHG RQ WKH 6LQLFD &amp;RUSXV ZH ILQG WKDW DOO W\\SH $ YHUEV DSSHDU LQ WKH LPSHUDWLYH RU WKH HYDOXDWLYH FRQVWUXFWLRQ DQG ZLWK RQO\\ RQH H[FHSWLRQ LH 1,33,4 W\\SH % 7DEOH ,PSHUDWLYH DQG (YDOXDWLYH 8VHV RI WKH 6HYHQ 3DLUV 9HUE 7\\SHV 9HUEV ,PS (YD 7RWDO 9HUEV ,PS (YD 7RWDO +DSS\\ JDR[LQJ \u00cf NXDLOH L9 'HSUHVVLRQ QDQJXR \u00c4 WRQJNX \u00c0\u00c7 6DGQHVV VKDQJ[LQ \u00a8 EHLVKDQJ 5HJUHW KRXKXL \u00f5a \\LKDQ J $QJHU VKHQJTL !\u00db IHQQX ,\u00f7 )HDU KDLSD m\u00a2 NRQJMX 0 :RUU\\ GDQ[LQ \u00a8 IDQQDR (* 9HUEDO DVSHFW RU DNWLRQVDUW 9HUEV RI HPRWLRQ H[SUHVV PHQWDO VWDWHV 7KH\\ FDQ UHSUHVHQW HLWKHU D KRPRJHQHRXV VWDWH DV LQ RU DQ LQFKRDWLYH VWDWH DV LQ \u00d4 = \u00bc \u00a8lb\u1e80 D ZHL FL VKL VKDQJ[LQ EX\\L KH IRU WKLV PDWWHU VDG FRQWLQXRXV +H KDV EHHQ VDG DERXW WKLV IRU D ORQJ WLPH \u00d4+ \u00e3\u00a6 &lt;W b`\u00bd 1^ \u00a81 \u00a6\u1e80 D \\L [$ % 9HUE 7\\SHV 9HUEV )UHT 9HUEV )UHT +DSSLQHVV JDR[LQJ \u00cf NXDLOH LW 'HSUHVVLRQ QDQJXR \u00c4 WRQJNX \u00c0\u00c7 6DGQHVV VKDQJ[LQ \u00a8 EHLVKDQJ 5HJUHW KRXKXL \u00f5 \\LKDQ J $QJHU VKHQJTL !\u00db IHQQX ,\u00f7 )HDU KDLSD m\u00a2 NRQJMX 0 :RUU\\ GDQ[LQ \u00a8 IDQQDR (* 7UDQVLWLYLW\\ $ \u00d4 \u00f0 \u00cf M? \u00b2 \u00e8`&gt;7VDL @ WDPHQ KHQ JDR[LQJ ]KDQJVDQ PHL ]RX WKH\\ YHU\\ JODG -RKQ GRHVQW JR 7KH\\ ZHUH JODG WKDW -RKQ GLGQW JR $GRSWLQJ 7HQJV &gt;@ L a1 ; z : \u00a81 ] F ZHLOH ]KH MLDQ VKL ZR FHQJ VKDQJ[LQ OH KDR MLX IRU WKLV SLHFH PDWWHU , HYHU VDG /( TXLWH ORQJ WLPH ,YH IHOW VDG DERXW WKLV PDWWHU IRU TXLWH D ORQJ WLPH LL W  \u00b6 l5 \u00e0\u00ff \u00fb\u00d6 D l \u00a8K PX ]L MLQJ EXGH MLDQPLDQ ]HQPH QHQJ EX VKDQJ[LQ QH PRWKHU VRQ GDUH FRXOGQW PHHW KRZ FDQ QRW VDG 1( +RZ FDQ WKH\\ QRW IHHO VDG WKDW WKH PRWKHU DQG VRQ FDQW PHHW HDFK RWKHU \u00d4 \u00f0 L9 M? \u00b2 \u00e8`&gt;7VDL @ WDPHQ KHQ NXDLOH ]KDQJVDQ PHL ]RX WKH\\ YHU\\ JODG -RKQ GRHVQW JR 7KH\\ ZHUH KDSS\\ WKDW -RKQ GLGQW JR $V IRU JRDO RQO\\ WKH YHUEV RI $QJU\\ $IUDLG DQG :RUULHG VHPDQWLFDOO\\ WDNH WKLV NLQG RI DUJXPHQW DQG WKXV V\\QWDFWLFDOO\\ WDNH WKHP DV GLUHFW REMHFWV +RZHYHU LQ WKH 6LQLFD &amp;RUSXV RQO\\ *URXS $ YHUEV RI WKRVH W\\SHV FDQ WDNH D JRDO DV D GLUHFW REMHFW ZKLOH *URXS % YHUEV DV D UXOH GR QRW WDNH D JRDO DV D GLUHFW REMHFW DV VKRZQ LQ 7DEOH 7DEOH 7KH 7UDQVLWLYH 8VHV RI )RXU 5HSUHVHQWDWLYH 3DLUV &amp;DXVH (YHQW *RDO &amp;DXVH (YHQW *RDO 7\\SH $ 936 (YHQW 1 6LPSOH 1 7\\SH % 936 HYHQW 1 6LPSOH 1 JDR[LQJ \u00cf NXDLOH LW VKHQJTL !\u00db IHQQX ,\u00f7 KDLSD m\u00a2 NRQJMX 0 GDQ[LQ \u00a8 IDQQDR (* 6HPDQWLF H[SODQDWLRQ ,Q WKLV VHFWLRQ ZH ZLOO ILUVW VXPPDUL]H WKH FRQWUDVWV DQG WKHQ SURSRVH D OH[LFDO VHPDQWLF H[SODQDWLRQ IRU DOO WKH FRQWUDVWV 7KH V\\QWDFWLF FRQWUDVWV ,Q D FDQRQLFDO REMHFW SRVLWLRQ VXFK DV VKDQJ WD GH TL !\u00d4\u00db &gt;+XDQJ @ FDQ UHSUHVHQW LQFKRDWLYH VWDWHV DQG FDQ WDNH FDXVHV RU JRDOV DV WKHLU GLUHFW REMHFWV 7\\SH % YHUEV DUH WKH SUHGRPLQDQW FKRLFH LQ D GHYHUEDO FRQWH[W KDYH ORRVHU VHOHFWLRQDO UHVWULFWLRQV RQ WKH KHDG ZKHQ WKH\\ IXQFWLRQ DV DGMXQFWV DUH VHOGRP XVHG LQ LPSHUDWLYH RU HYDOXDWLYH FRQVWUXFWLRQV UDUHO\\ UHSUHVHQW LQFKRDWLYH VWDWH DQG VHOGRP WDNH FDXVHV RU JRDOV DV WKHLU GLUHFW REMHFWV 7KH VHPDQWLF EDVLV IRU WKH ELSDUWLWH FODVVLILFDWLRQ %HDU LQ PLQG WKDW WKH YHUEV ZH DUH VWXG\\LQJ KHUH IRUP VHYHQ FRQWUDVW SDLUV :KLOH HDFK 7\\SH$ 7\\SH % 3UHGLFDWLYH XVH 6WURQJ :HDN ,QFKRDWLYH VWDWHV 3UHGRPLQDWH 5DUH 7UDQVLWLYLW\\ 6WURQJ :HDN ,PSHUDWLYH RU HYDOXDWLYH FRQVWUXFWLRQV 3UHGRPLQDWH 5DUH $GMXQFWV WR QRQWUDQVLHQW DFWLYLWLHV 5DUH 3UHGRPLQDWH $GMXQFWV WR QRXQV :HDN 6WURQJ 1RPLQDOL]DWLRQ 5DUH 3UHGRPLQDWH )URP WKH DERYH FRQWUDVW ZH JHQHUDOL]H WKDW DOO WKH GLVWLQFWLYH OLQJXLVWLF LQVWDQWLDWLRQV DUH UHODWHG WR HYHQW VWUXFWXUH SURSHUWLHV *HQHUDOO\\ VSHDNLQJ W\\SH $ YHUEV DUH SUHIHUUHG IRU LQGLFDWH WUDQVLWLRQ ZKLOH W\\SH % YHUEV DUH SUHIHUUHG IRU KRPRJHQHLW\\ ,Q SDUWLFXODU ZKHQ ZH ZDQW WR LQGLFDWH D FKDQJH RI VWDWH VXFK DV ZLWK WKH FKDQJHRIVWDWH 7\\SHS % NXDLOH L9 99 [L\\XH \u00db 99 KXDQOH 9 99 IDQQDR (* 99 NRQJMX H 99 WRQJNX \u00c0\u00c7 99 IHQQX ,\u00f7 99 FKHQ]KRQJ  \u00a7\u00f9 99 EHLVKDQJ 99 NXQDR \u00c7* 99 \\LKDQ J $1 RU 92 MXVDQJ \u00dc 99 NXDLKXR vR 99 RU $9 KXDQ[L \u00db 99 \\XNXDL -v 99 ZHLMX y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td></tr></table>",
+                "num": null,
+                "text": "$OWHUQDWLRQ $FURVV 6HPDQWLF )LHOGV $ 6WXG\\ RQ 0DQGDULQ 9HUEV RI (PRWLRQ 7KLV SDSHU H[SORUHV SRVVLEOH FRUHODWLRQV EHWZHHQ OH[LFDO VHPDQWLFV DQG PRUSKRV\\QWDFWLF VWUXFWXUHV :H ILUVW H[DPLQH D FRQVLVWHQW GLFKRWRP\\ DPRQJ YHUEV RI HPRWLRQ ZKLFK ZDV ILUVW REVHUYHG IRU YHUEV RI KDSSLQHVV E\\ 7VDL HW DO >@ ,W LV VKRZQ WKDW WKH GLFKRWRP\\ FDQ EH GHWHUPLQHG EDVHG RQ WKH FULWHULRQ RI ZKHWKHU D YHUE LV D 99 FRPSRXQG RU QRW 7KH OLQJXLVWLF FRQWUDVWV REVHUYHG LQFOXGH WKH JUDPPDWLFDO IXQFWLRQV RI D YHUE DV ZHOO DV WKHLU GLVWULEXWLRQ WKH VHOHFWLRQDO UHVWULFWLRQV WKH YHUEV LPSRVH DV DQ DGMXQFW D YHUEV RFFXUUHQFHV LQ LPSHUDWLYH DQG HYDOXDWLYH FRQVWUXFWLRQV LWV DNWLRQVDUW DQG LWV WUDQVLWLYLW\\ :H ZLOO VKRZ WKDW WKH RYHUW PRUSKRV\\QWDFWLF FRQ WUDVWV DUH GXH WR OH[LFDO HYHQW VWUXFWXUH SURSHUWLHV 7KH GHVFULSWLRQ RI D VWDWH RI HPRWLRQ FDQ IRFXV RQ KRZ WKH VWDWH FRPHV WR EH LH WKH LQFKRDWLYH VWDWH RU RQ WKH EHLQJ RI WKH VWDWH LH WKH KRPRJHQHRXV VWDWH 6LQFH 99 FRPSRXQGLQJ KDV WKH VHPDQWLF IXQFWLRQ RI UHIHUULQJ WR WKH JHQHULF SURSHUWLHV RI WKH VHW RI HYHQW WRNHQV LW LV QDWXUDO IRU 99 FRPSRXQGLQJ WR EH FKRVHQ DV WKH PRUSKRV\\QWDFWLF UHSUHVHQWDWLRQ RI KRPRJHQHLW\\ ,QWURGXFWLRQ 5HFHQW OH[LFDO VHPDQWLF VWXGLHV VXFK DV WKRVH RI /HYLQ >@ DQG 3XVWHMRYVN\\ >@ KDYH WULHG WR H[SODLQ KRZ OH[LFDO PHDQLQJ SUHGLFWV V\\QWDFWLF UHJXODULWLHV 2QH DSSURDFK LV WR VWXG\\ WKH GLIIHUHQFHV EHWZHHQ QHDU V\\QRQ\\PV WR LGHQWLI\\ WKH PLQLPDO VHPDQWLF 1RWH WKDW WKH\\ KDYH WKH XQLIRUP FDWHJRULFDO DVVLJQPHQW RI VWDWLYH YHUEV LQ &KLQHVH ZKLOH WKH WUDQVODWLRQ HTXLYDOHQWV PD\\ SUHGRPLQDQWO\\ EH DGMHFWLYHV LQ ODQJXDJHV VXFK DV (QJOLVK 2XU FODVVLILFDWLRQ RI WKH YHUEV RI HPRWLRQ ODUJHO\\ IROORZV WKDW RI WKH 7RQJ\\LFL &LOLQ DWWULEXWHV WKDW FUHDWH WKH GLIIHUHQFHV >7VDL HW DO /LX DQG /LX HW DO @ ,Q WKLV FXUUHQW VWXG\\ ZH H[WHQG WKH UDQJH RI VWXG\\ WR VHPDQWLF ILHOGV ZKLFK FRQWDLQ PRUH WKDQ RQH V\\QRQ\\P SDLU 7KXV ZH FDQ DWWHVW WR WKH SULPDU\\ VWDWXV RI WKH VHPDQWLF DWWULEXWHV SURSRVHG LQ SUHYLRXV VWXGLHV E\\ VKRZLQJ WKDW VLPLODU JHQHUDOL]DWLRQV FDQ EH H[WHQGHG WR WKH RWKHU V\\QRQ\\P SDLUV LQ WKH VDPH VHPDQWLF ILHOG 7KLV VWXG\\ DWWHPSWV WR HODERUDWH RQ 7VDL HW DOV >@ ZRUN RQ WKH GLIIHUHQFHV EHWZHHQ WKH V\\QRQ\\PV NXDLOH L9 WR EH KDSS\\ DQG JDR[LQJ \u00cf WR EH JODG :H UHH[DPLQH WKH GLIIHUHQFHV RYHU D EURDGHU UDQJH LH WKH YHUEV RI HPRWLRQ 2XU VWXG\\ ZLOO OHDG WR WKH IROORZLQJ IRXU LPSRUWDQW JHQHUDOL]DWLRQV 7KHVH GLIIHUHQFHV DUH QRW VSHFLILF WR NXDLOH DQG JDR[LQJ EXW WR WKH ZKROH VHPDQWLF ILHOG RI YHUEV RI HPRWLRQ 7KHVH GLIIHUHQFHV FDQ EH PRUH VXFFLQFWO\\ GHILQHG 7KHVH GLIIHUHQFHV DUH PRWLYDWHG E\\ GLIIHUHQW OH[LFDO HYHQW W\\SHV 7KH VRXUFH RI WKHVH GLIIHUHQFHV FDQ DFFRXQW IRU WKH FKRLFH RI FRPSRXQG VWUXFWXUH 9HUEV LQYROYLQJ VHYHQ VXEILHOGV RI HPRWLRQ ZLOO EH H[DPLQHG LQ WKLV SDSHU LH +DSSLQHVV 'HSUHVVLRQ 6DGQHVV 5HJUHW $QJHU )HDU DQG :RUU\\ 7KLV DOORZV XV WR REWDLQ JHQHUDOL]DWLRQV DERXW WKH FRPSOHWH VHPDQWLF ILHOG RI HPRWLRQ DV ZHOO DV WR REVHUYH LI WKHUH DUH DQ\\ YDULDWLRQV DPRQJ WKH VXEILHOGV $OO WKH REVHUYDWLRQV DQG VWDWLVWLFV LQ WKLV SDSHU DUH EDVHG RQ WKH $FDGHPLD 6LQLFD %DODQFHG &RUSXV RI 0RGHUQ 0DQGDULQ &KLQHVH DEEUHYLDWHG DV WKH 6LQLFD &RUSXV LQ WKH IROORZLQJ WH[W ZKLFK LV D WDJJHG 0DQGDULQ FRUSXV FRQWDLQLQJ D WRWDO RI ILYH PLOOLRQ ZRUGV >&.,3 @ :H WDNH LQWR DFFRXQW RQO\\ YHUEV ZLWK RFFXUUHQFH IUHTXHQF\\ UDWHV RI RYHU LQ WKH 6LQLFD &RUSXV VR WKDW WKHUH ZLOO EH VXIILFLHQW HPSLULFDO HYLGHQFH IRU DQ\\ REVHUYHG FRQWUDVWV 7KH YHUEV VWXGLHG KHUH DUH OLVWHG LQ 7DEOH ZLWK WKHLU IUHTXHQF\\ RI RFFXUUHQFH LQ WKH 6LQLFD &RUSXV OLVWHG LQ SDUHQWKHVHV 7KHUH DUH WKLUW\\WKUHH YHUEV LQ WRWDO )RXU RI WKHP DUH PRQR V\\OODELF DQG WZHQW\\QLQH RI WKHP DUH GLV\\OODELF ,Q WKLV VWXG\\ ZH ZLOO IRFXV RQ WKH GLV\\OODELF YHUEV LQ RUGHU WR H[SORUH WKH FRUHODWLRQ EHWZHHQ PRUSKROH[LFDO FRPSRVLWLRQ DQG OH[LFDO VHPDQWLFV ,Q DGGLWLRQ WKH H[FOXVLRQ RI PRQRV\\OODELF YHUEV DOORZV XV DYRLG GHDOLQJ ZLWK SRWHQWLDO FRPSOLFDWLRQV LQYROYLQJ SRO\\VHP\\ DQG ERXQGHGQHVV / / &KDQJ HW DO 7DEOH 9HUEV RI HPRWLRQ ZLWK D IUHTXHQF\\ RI RYHU LQ WKH 6LQLFD &RUSXV FOHDU GLFKRWRP\\ DPRQJ YHUEV RI +DSSLQHVV 7KH\\ DUH WKH WZR PRVW IUHTXHQWO\\ XVHG YHUEV RI WKH VXEILHOG PRUHRYHU DOO RWKHU YHUEV RI +DSSLQHVV DOVR IDOO RQWR HLWKHU VLGH RI WKH GLFKRWRP\\ +HQFH LW LV QDWXUDO WR DVN LI VXFK JHQHUDOL]DWLRQV FDQ EH FDUULHG RYHU WR RWKHU VXEILHOGV RI HPRWLRQ ,Q RWKHU ZRUGV DUH WKHVH FRQWUDVWV LGLRV\\QFUDWLF WR YHUEV RI +DSSLQHVV RU GR WKH\\ UHSUHVHQW VRPH FRPPRQ FRQFHSWXDO PRWLYDWLRQ RI WKH YHUEV RI HPRWLRQ\" 2XU WKRURXJK H[DPLQDWLRQ RI WKH YHUEV RI HPRWLRQ KDV UHYHDOHG WKDW WKHVH GLIIHUHQFHV DUH UHSHDWHG LQ HDFK RI WKH VHYHQ VXEILHOGV RI HPRWLRQ ,Q DGGLWLRQ ZH KDYH DOVR IRXQG DGGLWLRQDO UHSUHVHQWDWLRQDO FOXHV DV WR WKLV GLFKRWRP\\ %DVHG RQ RXU VWXG\\ ILYH GLVWULEXWLRQDO FULWHULD FDQ EH XVHG WR FUHDWH D ELSDUWLWH FODVVLILFDWLRQ RI WKH YHUEV RI HPRWLRQ 7KH ILUVW WZR FULWHULD DUH QHZO\\ SURSRVHG KHUH ZKLOH WKH IROORZLQJ WKUHH ZHUH SURSRVHG E\\ 7VDL HW DO >@ :LVK VHQWHQFHV UHIHU WR WKH JUHHWLQJ FRQVWUXFWLRQV VXFK DV ]KX QL NXDLOH \"L9 0D\\ \\ RX EH 6WXG\\ RQ 0DQGDULQ 9HUEV RI (PRWLRQ FRQWUDVW SDLUV LQ WKH IROORZLQJ VHFWLRQV 2XU SUHGLFWLRQ LV WKDW WKH RWKHU YHUEV DV PHPEHUV RI WKH ILHOG ZLOO EHKDYH OLNH HLWKHU FRYHULQJ WHUP LQ WKH SDLU 7KH UHSUHVHQWDWLRQDO GLVWLQFWLRQV WV UHJUHWIXO WKDW WKH ZRUNV RI WKLV DUWLVW FRXOGQW EH H[KLELWHG WKLV \\HDU ,Q WKLV VWXG\\ WKH FKRLFH RI WKH FRQWUDVW SDLUV KDV LPSRUWDQW PHWKRGRORJLFDO FRQVLGHUDWLRQV $V H[SODLQHG LQ RXU SUHYLRXV GLVFXVVLRQ WKHUH DUH XVXDOO\\ PRUH WKDQ WZR WHUPV LQ HDFK VHPDQWLF ILHOG +HQFH IRU HLWKHU W\\SH $ RU W\\SH % WKH FRYHULQJ WHUP LQ WKH FRQWUDVW SDLU PD\\ QRW EH WKH RQO\\ FKRLFH +RZHYHU ZH FDQ VHH IURP WKH IUHTXHQF\\ VWDWLVWLFV LQ 7DEOH WKDW D FRYHULQJ WHUP KDV D IUHTXHQF\\ PXFK KLJKHU WKDQ HYHQ WKH QH[W PRVW IUHTXHQW WHUP RI WKH VDPH W\\SH 7KXV RXU VWXG\\ XVHV D VLPSOLILHG PRGHO ZKHUH RQO\\ WKH WZR GRPLQDQW WHUPV LQ WKH FRQWUDVW SDLU DUH FRPSDUHG DVVXPLQJ WKDW WKH OHVV IUHTXHQW WHUPV FDQFHO WKHPVHOYHV RXW DQG GR QRW FRQWULEXWH WR VLJQLILFDQW GLIIHUHQFHV LI WDNHQ LQWR FRQVLGHUDWLRQ 7DEOH 'LVWULEXWLRQ RI *UDPPDWLFDO )XQFWLRQV RI WKH 6HYHQ &RQWUDVW 3DLUV DERYH GDWD VKRZ WKH FOHDU GLVWULEXWLRQDO GLVSDULW\\ EHWZHHQ W\\SH $ DQG W\\SH % YHUEV LQ WHUPV RI SUHGLFDWLYH DQG QRPLQDO XVHV +RZHYHU LQ RUGHU WR REWDLQ D OLQJXLVWLFDOO\\ VLJQLILFDQW DFFRXQW DQG LWV DVVRFLDWHG LPSOLFDWLRQV IURP WKH GLVWULEXWLRQDO GLVSDULW\\ ILQHU VWDWLVWLFDO FRQWUDVWV PXVW EH XWLOL]HG ,Q SDUWLFXODU IURP D IXQFWLRQDO SRLQW RI YLHZ WKH WZR QHDU V\\QRQ\\PV LQ D FRQWUDVW SDLU DUH FRPSHWLQJ ZLWK HDFK RWKHU WR UHSUHVHQW WKH VDP H FRQFHSW ,Q RWKHU ZRUGV ZKHQ D FHUWDLQ JUDPPDWLFDO IXQFWLRQ LV H[SUHVVHG WKH FKRLFH LV EHWZHHQ WKH WZR YHUEV RI D FRQWUDVW SDLU *LYHQ WKLV IXQFWLRQDO SHUVSHFWLYH WKH TXDQWLWDWLYH PHDVXUHPHQW WKDW GLUHFWO\\ FKDUDFWHUL]HV WKH FRQWUDVW EHWZHHQ W\\SH $ DQG W\\SH % YHUEV LV WKH UDWLR RI KRZ RIWHQ RQH W\\SH LV FKRVHQ RYHU DQRWKHU W\\SH IRU D FHUWDLQ JUDPPDWLFDO IXQFWLRQ 7R REWDLQ WKLV VLJQLILFDQW PHDVXUHPHQW ZH FRPSLOHG WZR OLNHOL KRRG UDWLRV IRU HDFK SDLU ZLWK WKH GRPLQDQW W\\SH DV WKH GLYLGHH LQ HLWKHU FDVH IRU WKH OLNHOLKRRG UDWLR LQ SUHGLFDWLYH XVHV W\\SH $ WHUP IUHTXHQF\\ ZDV PHDVXUHG DJDLQVW WKH W\\SH % WHUP IUHTXHQF\\ DQG YLFH YHUVD IRU QRPLQDO XVHV )RU LQVWDQFH WKH IUHTXHQF\\ RI SUHG LFDWLYH XVH RI W\\SH $ VKHQJTL LV ZKLOH WKDW RI LWV QHDU V\\QRQ\\PRXV FRQWUDVWLQJ WHUP IHQQX LV +HQFH WKH W\\SH $ YHUE $QJHU LV WLPHV PRUH OLNHO\\ WR EH FKRVHQ WR H[SUHVV D SUHGLFDWLYH PHDQLQJ WKDQ LV LWV W\\SH % FRXQWHUSDUW 2Q WKH RWKHU KDQG IRU WKH VDPH SDLU RI YHUEV WKH W\\SH % YHUE LV WLPHV PRUH OLNHO\\ WR EH FKRVHQ WR H[SUHVV D QRPLQDO PHDQLQJ WKDQ LV LWV W\\SH $ FRXQWHUSDUW %RWK OLNHOLKRRG UDWLRV ZHUH FDOFXODWHG IRU HDFK RI WKH VHYHQ SDLUV DQG JLYHQ EHORZ WKDW WKHUH FDQ EH DQ HPSLULFDO GHPDUFDWLRQ EHWZHHQ WKH WZR W\\SHV RI YHUEV )ROORZLQJ WKH GLVFXVVLRQ LQ WKH SUHYLRXV VHFWLRQ WKH WZR W\\SHV RI QRPLQDO XVHV LH WKDW RI D QRPLQDOL]HG HYHQW DQG WKDW RI D QRPLQDO PRGLILHU ZLOO EH PHUJHG :H ZLOO UHIHU WR WKH PHUJHG IUHTXHQF\\ DV WKH TXDQWLWDWLYH LQGH[ RI EHLQJ GHYHUEDO ,Q DGGLWLRQ VLQFH ZKDW LV VWXGLHG KHUH DPRXQWV WR TXDQWLWDWLYH FULWHULD XVHG WR GHWHUPLQH ZKHWKHU D WHUP LV D W\\SH $ RU W\\SH % YHUE DOO UHODYHQW YHUEV RI HPRWLRQ DUH WDNHQ LQWR FRQVLGHUDWLRQ RU W\\SH % DFFRUGLQJ WR WKH VLPSOH TXDQWLWDWLYH PHDVXUHPHQW RI WKH IUHTXHQF\\ RI WKHLU GHYHUEDO XVHV )RU H[DPSOH LQ 7DEOH ZH ILQG WKDW ZLWKRXW H[FHSWLRQ DOO W\\SH $ YHUEV KDYH D QRPLQDO XVH IUHTXHQF\\ RI RU ORZHU ZKLOH DOO W\\SH % YHUEV KDYH D QRPLQDO XVH IUHTXHQF\\ RI RU KLJKHU %HWZHHQ WKH WZR JURXSV WKHUH LV DQ REYLRXV JDS LQ WZR FUXFLDO VHQVHV )LUVW WKH OHDVW IUHTXHQW GHYHUEDO XVH RI D W\\SH % YHUE LV PRUH WKDQ KLJKHU WKDQ WKDW RI WKH PRVW IUHTXHQW W\\SH $ YHUE ,Q RWKHU ZRUGV WKH GLVWULEXWLRQV RI WKH WZR JURXSV RI YHUEV DUH FOHDUO\\ GLVFUHWH DQG QRW FRQWLQXRXV 6HFRQG WKH FRQWUDVW LQ HDFK FRQWUDVW SDLU LV HYHQ PRUH SURPLQHQW WKDQ WKH DERYH JDS 7KH VPDOOHVW JDS EHWZHHQ D SDLU LV VKDQJ[LQ YV EHLVKDQJ 6XPPDU\\ ,Q WKLV VHFWLRQ ZH ZLOO VXPPDUL]H WKH TXDQWLWDWLYH PHDVXUHPHQWV ZH FKRVH WR FKDUDF WHUL]H WKH GLVWULEXWLRQ RI WKH JUDPPDWLFDO IXQFWLRQV RI WKH WZR W\\SHV RI YHUEV DQG WKHLU LQWHUSUHWDWLRQV )LUVW ZH REVHUYHG WKH RYHUDOO GLVWULEXWLRQV RI DOO WKH UHSUHVHQWHG JUDPPDWLFDO IXQFWLRQV RI WKH VHYHQ FRQWUDVW SDLUV DQG HVWDEOLVKHG WKDW W\\SH $ YHUEV DUH XVHG SUHGRPLQDQWO\\ IRU SUHGLFDWLYH XVHV DQG WKDW W\\SH % YHUEV DUH XVHG SUHGRPLQDQWO\\ IRU QRPLQDO XVHV 6HFRQG WKH OLNHOLKRRG UDWLR PHDVXUHPHQWV IRU ERWK SUHGLFDWLYH DQG QRPLQDO XVHV ZHUH FDOFXODWHG IRU HDFK FRQWUDVW SDLU 7KLV PHDVXUHPHQW ZDV XVHG WR WHVW WKH IXQFWLRQGULYHQ K\\SRWKHVLV WKDW WKH WZR FRQWUDVW SDLU PHPEHUV DUH FDQGLGDWHV ZKLFK FRPSHWH WR UHSUHVHQW WKH VDPH FRQFHSW LQ DQ\\ JLYHQ FRQWH[W 7KH TXDQWLWDWLYH PHDVXUHPHQWV ZHUH SUHGLFWHG EDVHG RQ WKH K\\SRWKHVLV DQG WKXV VXSSRUWHG D IXQFWLRQDO DFFRXQW /DVW WR YHULI\\ WKDW RXU TXDQWLWDWLYH PHDVXUHPHQWV UHSUHVHQWHG D WUXH FODVVLILFDWRU\\ FULWHULRQ LQVWHDG RI D UDQGRP GHPDUFDWLRQ SRLQW LQ FRQWLQXRXV GLVWULEXWLRQ ZH VKRZHG WKDW WKH W\\SH $ DQG W\\SH % YHUEV DFWXDOO\\ IRUP WZR GLVFUHWH JURXSV VHSDUDWHG E\\ D VLJQLILFDQW JDS DFFRUGLQJ WR WKHLU IUHTXHQF\\ RI GHYHUEDO XVHV 6HOHFWLRQDO UHVWULFWLRQV WKH YHUEV LPSRVH DV DGMXQFWV 7KH VHFRQG LPSRUWDQW REVHUYDWLRQ UHJDUGLQJ WKH GLVWULEXWLRQ RI WKH WZR W\\SHV RI YHUEV RI HPRWLRQ LV WKDW DV DGMXQFWV WKH\\ LPSRVH YHU\\ GLIIHUHQW VHOHFWLRQDO UHVWULFWLRQV RQ WKHLU KHDGV 7\\SH $ YHUEV FDQ RQO\\ PRGLI\\ D YHU\\ UHVWULFWHG VHW RI QRXQV RU YHUEV ZKLOH W\\SH % YHUEV VHHP WR EH PXFK IUHHU ,Q WKH 6LQLFD &RUSXV W\\SH $ YHUEV VXFK DV JDR[LQJ FDQ RQO\\ PRGLI\\ VL[ W\\SHV RI QRXQV WLPH ZKHQ HJ VKLKRX\u00b4\" VKL\u00b4 HYHQWVWRU\\ PRRG IDFLDO H[SUHVVLRQV SHUVRQ DQG XWWHUDQFH ,Q FRQWUDVW W\\SH % YHUEV VXFK DV NXDLOH FDQ EH DGMXQFWV IRU PDQ\\ DGGLWLRQDO QRXQ FODVVHV 7KH FRQWUDVW LV VKRZQ LQ DQG LWK UHJDUG WR SRVWYHUEDO DGMXQFWV ERWK JURXSV FDQ PRGLI\\ WUDQVLHQW DFWLYLWLHV VXFK DV ZDQ GH KHQ JDR[LQJ (5\u00f0\u00cf SOD\\ KDSSLO\\ DQG ZDQ GH KHQ NXDLOH (5\u00f0L9 SOD\\ KDSSLO\\ +RZHYHU RQO\\ W\\SH % YHUEV FDQ EH DGMXQFWV RI QRQWUDQVLHQW VWDWHOLNH DFWLYLWLHV VXFK DVKXR GH NXDLOH R5L9 OLYH KDSSLO\\ JXR GH NXDLOH 5L9 OLYH KDSSLO\\ DQG DR GH KHQ WRQJNX L5\u00f0\u00c0\u00c7 HQGXUH WHUULEO\\ 7KH LPSHUDWLYH DQG HYDOXDWLYH FRQVWUXFWLRQV 6RPH YHUEV RI HPRWLRQ DUH XVHG LQ LPSHUDWLYH VHQWHQFHV FRQWDLQLQJ GHRQWLF PRGDO YHUEV DV LQ 0DQ\\ RI WKHP FDQ DOVR RFFXU LQ HYDOXDWLYH VHQWHQFHV ZKLFK FRQWDLQ WKH YHUE ]KLGH 5 EH ZRUWKZKLOH WR RU WKH SKUDVH PHL VKHPH KDR /0 \u00b2y\u00d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u00cf QRQ99 QDQJXR \u00c4 QRQ99 KRXKXL \u00f5 QRQ99WRQJNX \u00c0L QRQ99 GDQ\\RX QRQ99VKHQJTL 3\u00db QRQ99 FKLMLQ~\u00eb QRQ99 GDQ[LQ \u00a8 QRQ99 VKDQJ[LQ \u00a8 QRQ99 NDL[LQ H\u00a8 QRQ99 \\RX[LQ QRQ99 WRQJ[LQ \u00c0\u00a8 QRQ99 KDLSD m\u00a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
+                "html": null,
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-2005.json

@@ -0,0 +1,410 @@
+{
+    "paper_id": "O00-2005",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:03.915736Z"
+    },
+    "title": "",
+    "authors": [
+        {
+            "first": "Mei-Chun",
+            "middle": [],
+            "last": "Liu",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Chu-Ren",
+            "middle": [],
+            "last": "Huang",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Charles",
+            "middle": [],
+            "last": "Lee",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Ching-Yi",
+            "middle": [],
+            "last": "Lee",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O00-2005",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "& /LX HW DO",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "& /LX HW DO",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "& /LX HW DO",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "'HUEV RI 7KURZLQJ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "& /LX HW DO",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "& /LX HW DO",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "& /LX HW DO",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "16 %7 DQG",
+                "authors": [
+                    {
+                        "first": "% $gplwwlqj Lpshglphqwv Lq /H[lfdo $",
+                        "middle": [],
+                        "last": "Ftxlvlwlrq",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "$7.,16 %7 DQG /(9,1 % $GPLWWLQJ LPSHGLPHQWV LQ /H[LFDO $FTXLVLWLRQ ([SORLWLQJ 2QOLQH 5HVRXUFHV WR %XLOG D /H[LFRQ HG E\\ 8 =HUQLN +LOOVGDOH 1-/DZUHQFH (UOEDXP $VVRFLDWHV",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "QYHVWLJDWLQJ /DQJXDJH 8VH 7KURXJK &RUSXVEDVHG $QDO\\VHV RI $VVRFLDWLRQ 3DWWHUQV ,QWHUQDWLRQDO -RXUQDO RI &RUSXV /LQJXLVWLFV 9RO",
+                "authors": [
+                    {
+                        "first": "%",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "%",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "%,%(5 ' ,QYHVWLJDWLQJ /DQJXDJH 8VH 7KURXJK &RUSXVEDVHG $QDO\\VHV RI $VVRFLDWLRQ 3DWWHUQV ,QWHUQDWLRQDO -RXUQDO RI &RUSXV /LQJXLVWLFV 9RO",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "&KDR <8(15(1 $ *UDPPDU RI 6SRNHQ &KLQHVH %HUNHOH\\ 8QLYHUVLW\\ RI &DOLIRUQLD 3UHVV",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "&KDR <8(15(1 $ *UDPPDU RI 6SRNHQ &KLQHVH %HUNHOH\\ 8QLYHUVLW\\ RI &DOLIRUQLD 3UHVV",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "*5$1'< 5,&+$5' ( 6HPDQWLF )LHOGV 3URWRW\\SHV DQG WKH /H[LFRQ LQ )UDPHV )LHOGV DQG &RQWUDVWV 1HZ",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "*5$1'< 5,&+$5' ( 6HPDQWLF )LHOGV 3URWRW\\SHV DQG WKH /H[LFRQ LQ )UDPHV )LHOGV DQG &RQWUDVWV 1HZ (VVD\\V LQ 6HPDQWLF DQG /H[LFDO 2UJDQL]DWLRQ HG E\\ /HKUHU DQG .LWWD\\ SS +LOOVGDOH /DZUHQFH (UOEDXP",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "*UDPPDU $ )XQFWLRQEDVHG ,QWURGXFWLRQ 9RO $PVWHUGDP3KLODGHOSKLD -RKQ %HQMDPLQV 3XEOLVKLQJ &R",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "*,921 7 (QJOLVK *UDPPDU $ )XQFWLRQEDVHG ,QWURGXFWLRQ 9RO $PVWHUGDP3KLODGHOSKLD -RKQ %HQMDPLQV 3XEOLVKLQJ &R",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "7KH 0RGXOH$WWULEXWH 5HSUHVHQWDWLRQ RI 9HUEDO 6HPDQWLFV LQ :RUNLQJ 3DSHUV RQ &KLQHVH 9HUEDO 6HPDQWLFV 9RO SS HG E\\ . $KUHQV &5 +XDQJ DQG 0& 7VDL 7DLSHL $FDGHPLD 6LQLFD",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "+8$1* &+85(1 DQG $+5(16 .$7+/((1 7KH 0RGXOH$WWULEXWH 5HSUHVHQWDWLRQ RI 9HUEDO 6HPDQWLFV LQ :RUNLQJ 3DSHUV RQ &KLQHVH 9HUEDO 6HPDQWLFV 9RO SS HG E\\ . $KUHQV &5 +XDQJ DQG 0& 7VDL 7DLSHL $FDGHPLD 6LQLFD",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "&+81 DQG 76$, 0(,&+,+ )URP /H[LFDO 0HDQLQJ WR (YHQW 6WUXFWXUH $WWULEXWHV $FURVV 6HPDQWLF &ODVVHV RI 0DQGDULQ 9HUEV LQ :RUNLQJ 3DSHUV RQ &KLQHVH 9HUEDO 6HPDQWLFV 9RO SS HG E\\ . $KUHQV &5 +XDQJ DQG 0& 7VDL",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "+8$1* &+85(1 /LX 0(,&+81 DQG 76$, 0(,&+,+ )URP /H[LFDO 0HDQLQJ WR (YHQW 6WUXFWXUH $WWULEXWHV $FURVV 6HPDQWLF &ODVVHV RI 0DQGDULQ 9HUEV LQ :RUNLQJ 3DSHUV RQ &KLQHVH 9HUEDO 6HPDQWLFV 9RO SS HG E\\ . $KUHQV &5 +XDQJ DQG 0& 7VDL",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "5 )RXQGDWLRQV RI &RJQLWLYH *UDPPDU 9RO 6WDQIRUG 6WDQIRUG 8QLY 3UHVV",
+                "authors": [
+                    {
+                        "first": "*",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/$1*$&.(5 5 )RXQGDWLRQV RI &RJQLWLYH *UDPPDU 9RO 6WDQIRUG 6WDQIRUG 8QLY 3UHVV",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "5 )RXQGDWLRQV RI &RJQLWLYH *UDPPDU 9RO 6WDQIRUG",
+                "authors": [
+                    {
+                        "first": "*",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "8",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/$1*$&.(5 5 )RXQGDWLRQV RI &RJQLWLYH *UDPPDU 9RO 6WDQIRUG 8 3UHVV",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "7+ 9HUE &ODVVHV DQG $OWHUQDWLRQ &KLFDJR 8QLY RI &KLFDJR 3UHVV",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/(9,1 %(7+ 9HUE &ODVVHV DQG $OWHUQDWLRQ &KLFDJR 8QLY RI &KLFDJR 3UHVV",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "&+81 /H[LFDO 0HDQLQJ DQG 'LVFRXUVH 3DWWHUQLQJ WKH 7KUHH 0DQGDULQ &DVHV RI %XLOG LQ &RJQLWLRQ DQG )XQFWLRQ LQ /DQJXDJH HG E\\ % )R[ ' -XUDIVN\\ DQG / 0LFKDHOLV SS 6WDQIRUG &6/, 'HUEV RI 7KURZLQJ",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/,8 0(,&+81 /H[LFDO 0HDQLQJ DQG 'LVFRXUVH 3DWWHUQLQJ WKH 7KUHH 0DQGDULQ &DVHV RI %XLOG LQ &RJQLWLRQ DQG )XQFWLRQ LQ /DQJXDJH HG E\\ % )R[ ' -XUDIVN\\ DQG / 0LFKDHOLV SS 6WDQIRUG &6/, 'HUEV RI 7KURZLQJ",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "&+81 &RQFHSWXDO %DVLV DQG &DWHJRULDO 6WUXFWXUH D 6WXG\\ RI 0DQGDULQ 9 5",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/,8 0(,&+81 &RQFHSWXDO %DVLV DQG &DWHJRULDO 6WUXFWXUH D 6WXG\\ RI 0DQGDULQ 9 5",
+                "links": null
+            },
+            "BIBREF17": {
+                "ref_id": "b17",
+                "title": "&+81 $ 3LORW 6WXG\\ RQ &KLQHVH 9HUE &ODVVHV DQG $OWHUQDWLRQV 16& 3URMHFW 5HSRUW 16&+ E",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "3867",
+                "issue": "",
+                "pages": "3--3867",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/,8 0(,&+81 $ 3LORW 6WXG\\ RQ &KLQHVH 9HUE &ODVVHV DQG $OWHUQDWLRQV 16& 3URMHFW 5HSRUW 16&+ E 3867(-2<6.< -$0(6 7KH *HQHUDWLYH /H[LFRQ &DPEULGJH 7KH 0,7 3UHVV 3867(-2<6.< -$0(6 7KH 6\\QWD[ RI (YHQW 6WUXFWXUH &RJQLWLRQ SS 60,7+ & 6 7KH 3DUDPHWHU RI $VSHFW 'RUGUHFKW .OXZHU $FDGHPLF 3XEOLVKHUV",
+                "links": null
+            },
+            "BIBREF19": {
+                "ref_id": "b19",
+                "title": "* 6+28+6,1 HG &KLQHVH V\\QRQ\\PV XVDJH GLFWLRQDU\\ 7DLSHL &UDQH 3XEOLVKLQJ &R",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "* 6+28+6,1 HG &KLQHVH V\\QRQ\\PV XVDJH GLFWLRQDU\\ 7DLSHL &UDQH 3XEOLVKLQJ &R",
+                "links": null
+            },
+            "BIBREF21": {
+                "ref_id": "b21",
+                "title": "$+5(16 7RZDUGV D 5HSUHVHQWDWLRQ RI 9HUEDO 6HPDQWLFV DQ $SSURDFK %DVHG RQ 1HDU 6\\QRQ\\PV",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "$, 0(,&+,+ &+85(1 +8$1* .(+-,$11 &+(1 .$7+/((1 $+5(16 7RZDUGV D 5HSUHVHQWDWLRQ RI 9HUEDO 6HPDQWLFV DQ $SSURDFK %DVHG RQ 1HDU 6\\QRQ\\PV",
+                "links": null
+            },
+            "BIBREF22": {
+                "ref_id": "b22",
+                "title": "3URFHHGLQJV RI 52&/,1* ;SS +VLQFKX 7DLZDQ",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "3URFHHGLQJV RI 52&/,1* ;SS +VLQFKX 7DLZDQ",
+                "links": null
+            },
+            "BIBREF23": {
+                "ref_id": "b23",
+                "title": "$2 )(1*)8 2Q 9HUE &ODVVLILFDWLRQ LQ &KLQHVH -RXUQDO RI &KLQHVH /LQJXLVWLFV SS",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "$2 )(1*)8 2Q 9HUE &ODVVLILFDWLRQ LQ &KLQHVH -RXUQDO RI &KLQHVH /LQJXLVWLFV SS",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF0": {
+                "text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z %RWK 728 DQG =+, PD\\ WDNH D *RDO DV WKH GLUHFW REMHFW EXW ',8 DQG 5(1* GR QRW HJ 728 ODQ WR VKRRW WKH EDVNHW 728KX WR WKURZ RQHVHOI LQWR WKH ODNH =+, GL\\RXVKHQJ \u00aa \u00b6 WR WKURZ VRPHWKLQJ WR WKH JURXQG 3DWKHQGSRLQW 728 VHOHFWV D VSDWLDOO\\ ERXQGHG 3DWKHQGSRLQW EXW =+, GRHV QRW 7KLV LV HYLGHQW IURP WKH IDFW WKDW ZKHQ RFFXUULQJ ZLWK D ORFDWLYH SKUDVH DERXW RI WKH RFFXUUHQFHV RI 728 WDNH UX # RU MLQ ^ LQWR DV WKH ORFDWLYH WKDW LV 728 FROORFDWHV PRVW IUHTXHQWO\\ ZLWK D &217$,1(5JRDO ZKLOH WKH PDMRULW\\ RI WKH RFFXUUHQFHV RI =+, LV IROORZHG E\\ [LDQJFKRXZDQJ {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{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{ FKDR [ RU ZDQJ ] DOO PHDQLQJ WRZDUG ZKLFK VLPSO\\ LQGLFDWHV D GLUHFWHG SDWK ZLWK QR IXUWKHU VSHFLILFDWLRQ RI WKH VKDSH RI WKH LV SRO\\VHPLF ZLWK WZR PHDQLQJ IDFHWV %HVLGHV LWV XVH DV DQ DFWLYLW\\ YHUE LW FDQ DOVR EH XVHG DV DQ DFKLHYHPHQW YHUE 7KH PDLQ UHDVRQ LV WKDW ',8 OH[LFDOO\\ VSHFLILHV DQ (YHQWHQGSRLQW WKXV DOORZLQJ WKH IRFXV WR EH RQ WKH HQGLQJ VWDWH RI WKH HYHQW :H QRZ GUDZ WKH FRQFOXVLRQ WKDW ',8 GLIIHUV IURP 5(1* LQ WKDW LW DOORZV ZH KDYH PHQWLRQHG WZR W\\SHV RI HQGSRLQWV 3DWKHQGSRLQW YV (YHQWHQGSRLQW 3DWKHQGSRLQW PDUNV WKH ILQDO SRLQW RI D WUDMHFWRU\\SDWK LQ EDOOLVWLF PRWLRQ ZKLFK FRLQFLGHV ZLWK WKH VHPDQWLF UROH *RDO (YHQWHQGSRLQW RQ WKH RWKHU KDQG LV UHOHYDQW WR WKH ILQDO SRLQW RI DQ HYHQW FRQWRXU XVXDOO\\ LQGLFDWLQJ D UHVXOWDWLYH VWDWH 0 & /LX HW DO 7KHVH WZR W\\SHV RI HQGSRLQW DUH FUXFLDO IRU ILQHWXQLQJ WKH OH[LFDO VHPDQWLFV RI WKH IRXU LV FOHDU IURP WKDW ZKLOH ERWK 728 DQG =+, LV OH[LFDOO\\ VSHFLILHG ZLWK D 3DWKHQGSRLQW RQO\\ 728 UHTXLUHV D VSDWLDOO\\ ERXQGHG SDWKHQGSRLQW $V IRU ',8 DQG 5(1* WKHLU OH[LFDO PHDQLQJV DUH QRW VHQVLWLYH WR WKH HQFRGLQJ RI D 3DWKHQGSRLQW LQVWHDG WKH\\ FDQ EH IXUWKHU GLVWLQJXLVKHG LQ WHUPV RI WKHLU OH[LFDO VSHFLILFDWLRQ RI DQ (YHQWHQGSRLQW 9HUEDO 6HPDQWLFV DV (YHQWLYH ,QIRUPDWLRQ 7KH REVHUYHG GLIIHUHQFHV DV RXWOLQHG LQ DERYH FDQ EH YLHZHG IURP D PRUH JHQHUDO SHUVSHFWLYH SURSRVHG LQ +XDQJ DQG $KUHQV >@ LQ ZKLFK YHUE PHDQLQJV DUH GHVFULEHG LQ WHUPV RI VWUXFWXUDO DQG DWWULEXWLYH GLVWLQFWLRQV 7KH\\ DUJXH WKDW DOO JUDPPDWLFDO LQIRUPDWLRQ LV HQFRGHG LQ WKH OH[LFRQ DQG WKDW YHUEV H[SUHVV HYHQWLYH LQIRUPDWLRQ (DFK YHUEDO VHQVH LV WKHQ WDNHQ WR EH D XQLTXH HYHQW VWUXFWXUH VHH EHORZ IRU GHWDLOV 7KH IUDPHZRUN PDNHV XVH RI WKH FRQFHSW RI DQ HYHQW IRFXV WR LGHQWLI\\ GLIIHUHQW HYHQW W\\SHV DV H[SODLQHG DQG LOOXVWUDWHG LQ EHORZ (YHQW )RFXV $ SURWRW\\SLFDO YHUE LV XVHG WR GHVFULEH DQ HYHQW DQG LWV OH[LFDO PHDQLQJ VSHFLILHV WKH SRVVLEOH VFRSH RI HYHQWV LW FDQ GHVFULEH )ROORZLQJ 6PLWKV >@ SURSRVDO RI YLHZSRLQW IRFXV LQ KHU DFFRXQW RI YHUEDO DVSHFWV DQ HYHQW IRFXV LV WDNHQ WR EH D FRQFHSWXDO DQG FRJQLWLYH SURILOH WKDW DOORZV PHDQLQJ H[WHQVLRQV ZLWKLQ WKH VFRSH RI OH[LFDO VSHFLILFDWLRQ 7KH QRWLRQ RI HYHQW IRFXV LV DV LPSRUWDQW DV WKDW RI HYHQW FRPSRQHQWV $ W\\SLFDO H[DPSOH FDQ EH IRXQG LQ WKH IROORZLQJ FDVH RI EXLOGLQJ YHUEV MLDQ \u00ec YV JDL > 7KH WZR V\\QRQRPRXV YHUEV VHHP WR KDYH WKH VDPH HYHQW FRPSRQHQWV \\HW WKH\\ KDYH GLIIHUHQW HYHQW IRFXVHV 7KH YHUE MLDQ DOORZV DQ LQWUDQVLWLYH XVH ZLWK WKH 7KHPH EHLQJ WKH VXEMHFW WKXV KLJKOLJKWLQJ WKH (YHQWHQGSRLQW >FI /LX +XDQJ /LX DQG 7VDL @ 7KXV LQ",
+                "html": null,
+                "type_str": "table",
+                "content": "<table><tr><td>9HUEV RI 7KURZLQJ 'HUEV RI 7KURZLQJ 'HUEV RI 7KURZLQJ</td><td>0 &amp; /LX HW DO</td></tr><tr><td colspan=\"2\">0DQGDULQ YHUEDO PHDQLQJV &amp;RUSXVEDVHG 6WXG\\ RI 1HDU6\\QRQ\\PV ,Q UHVSRQVH WR WKH QHHG RI ILQHWXQLQJ YHUEDO VHPDQWLFV 7VDL +XDQJ DQG &amp;KHQ &gt;@ SUHVHQWHG DQ LQWHUHVWLQJ ZRUN RQ GLIIHUHQWLDWLQJ D SDLU RI QHDUV\\QRQ\\PV JDR[LQJ \u00cf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gt;@ H[DPLQHG DQRWKHU LQWHUHVWLQJ VHW RI QHDUV\\QRQ\\PRXV YHUEV MLDQ \u00ec JDL &gt; DQG ]DR a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u00aa ',8 G DQG 5(1* DOO JORVVHG DV WR WKURZ ,W LV ZLWK D WKXPS z ',8 DQG 5(1* IRUP W\\SLFDO 99 FRPSRXQGV ZLWK 9 0DQQHU RU 9 5HVXOW ZKLOH 728 DQG =+, GR QRW VHHP WR IRUP WKHVH FRPSRXQGV HJ OXDQ ',85(1* G WR UHFNOHVVO\\ WKURZ VRPHWKLQJ ',85(1* GLDR G ~ WR WKURZ DZD\\ =+, ',8 5(1* )ROORZLQJ WKH DERYH EDFNJURXQG LQWURGXFWLRQ VHFWLRQ LQ WKLV SDSHU RXWOLQHV WKH SUHOLPLQDU\\ FRQWUDVW WKDW H[LVWV DPRQJ WKH IRXU YHUEV 6HFWLRQ WKHQ GHWDLOV WKHLU GLV WULEXWLRQDO GLIIHUHQFHV 6HFWLRQ HVWDEOLVKHV D V\\VWHPDWLF UHSUHVHQWDWLRQ RI WKH VHPDQWLF GLVWLQFWLRQV )LQDOO\\ VHFWLRQ FRQFOXGHV ZLWK D GLVFXVVLRQ RI WKH VLJQLILFDQFH RI WKLV ZRUN 3UHOLPLQDU\\ 2EVHUYDWLRQ 728 YV ',8 WHUPV RI PDQQHU DQG GLUHFWLRQDOLW\\ ,QWHUSUHWDWLRQDO 'LIIHUHQFHV EHWZHHQ 7284,8 \u00fa DQG ',84,8 / \u00fa 0$11(5 ',5(&amp;7,21$/,7&lt; 7284,8 FDUHIXOO\\ WDUJHWLQJ WRZDUG D VLQJOH DQG SUHFLVH GLUHFWLRQ ',84,8 UDQGRPO\\ WKURZLQJ QR VSHFLILF GLUHFWLRQ 'LVWLQFWLRQ LQ 3DWK(QGSRLQW 7KH VHFRQG REVHUYDWLRQ FRQFHUQV WKH VHPDQWLF UROH RI WKH GLUHFW REMHFW IROORZLQJ 728 RU ',8 ZKLFK LV WHUPHG WKH 3DWKHQGSRLQW %\\ 3DWKHQGSRLQW ZH UHIHU VSHFLILFDOO\\ WR WKH VHPDQWLF UROH JHQHUDOO\\ DQG ORRVHO\\ WHUPHG WKH *RDO ZKLFK PDUNV WKH ILQDO SRLQW RI D WUDMHFWRU\\ LQKHUHQW LQ D GLUHFWHG PRWLRQ &gt;FI WKH FDVH VWXG\\ RI (QJOLVK RYHU GLVFXVVHG E\\ /DNRII @ 7KHUH DUH WZR VHWV RI HYLGHQFH WKDW VKRZ WKDW 728 LV OH[LFDOO\\ VSHFLILHG ZLWK D 3DWKHQGSRLQW )LUVW LQ WHUP RI FRPSRXQGLQJ H[DPSOHV LQ EHORZ LOOXVWUDWH WKDW RQO\\ 728 PD\\ WDNH D 3DWKHQGSRLQW DV LWV GLUHFW REMHFW QRW ',8 728 ZLWK 3DWKHQGSRLQW D WRXODQ WR VKRRW D EDVNHW WRXKX]KLMLQ \u00de WR WKURZ RQHVHOI LQWR D ODNH WRXJRQJ A WR GHIHFW WR &amp;RPPXQLVW &amp;KLQD WRXTLVXRKDR !\u00ac WR SOHDVH VRPHRQH E\\ VDWLVI\\LQJ KLV ZLVKHV E GLXODQ G WR VKRRW D EDVNHW 7KH SRVVLEOH 7\\SLFDO 0DQQHUPRGLILHU RU 5HVXOWDWLYH&amp;RPSOHPHQW ZLWK ',8 D OXDQGLX`G WR PLQGOHVVO\\ WKURZ VRPHWKLQJ VRPHZKHUH E GLXGLDR G~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amp;;07-8 3DWK(QGSRLQW 6SHFLILHG 3DWKHQGSRLQW ',8YHUEV 3DWK(QGSRLQW 8QVSHFLILHG SDWKHQGSRLQW ,Q WKH QH[W VHFWLRQ ZH ZLOO JURXS WKH RWKHU WZR YHUEV =+, DQG 5(1* DFFRUGLQJ WR WKH EHKDYLRU RI 728 YV ',8 2EVHUYDWLRQ RQ =+, DQG 5(1* 3URSHUWLHV 6KDUHG E\\ =+, DQG 728 /LNH 728 =+, PD\\ DOVR WDNH D 3DWKHQGSRLQW DV LWV GLUHFW REMHFW =+, ZLWK D 3DWK(QGSRLQW D ]KLGL\\RXVKHQJ \u00aa \u00b6\" =+,JURXQGKDYHVRXQG WKURZLQJ VRPHWKLQJ WR WKH JURXQG ZLWK D WKXPS E OHLTLX]KL\\XDQ \u00a4\u00fa\u00aa VRIWEDOO=+,GLVWDQW SODFH VRIWEDOOWKURZLQJ 7KH IHDWXUH FRQWURO FRQFHUQV YROLWLRQDOLW\\ RI WKH VXEMHFW FRPSRXQGHG ZLWK D 3DWKHQGSRLQW EXW PD\\ IRUP D W\\SLFDO 99 FRPSRXQG ZLWK 9 LW LV FRPSOHWHG 7KH VWDWLYH YV DFWLYH GLVWLQFWLRQ FRQFHUQV NHQHVLV LQ JHQHUDO DV H[SODLQHG E\\ &amp;KDR &gt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gt;@ WKH GLIIHUHQFH EHWZHHQ LQFKRDWLYH DQG FRPSOHWLYH LV PDLQO\\ DVSHFWXDO LQFKRDWLYH UHIHUV WR D FKDQJH RI VWDWH RU WKH VWDUWLQJ SRLQW RI D QHZ HYHQW FRPSOHWLYH GHVFULEHV DQ HYHQW DV ,Q 3URSHUWLHV 6KDUHG E\\ ',8 DQG 5(1* 0DQQHU RU 9 5HVXOW ZKLFK LPSOLHV D ODFN RI GLUHFWLRQDOLW\\ 5(1* ZLWK PRGLILHUV WKDW ODFN GLUHFWLRQDOLW\\ D OXDQUHQJ` WR PLQGOHVVO\\ WKURZ VRPHWKLQJ LQ DOO GLUHFWLRQV E UHQJGLDR ~ WR WKURZ VRPHWKLQJ DZD\\ )XUWKHUPRUH ZKHQ IROORZHG E\\ D ORFDWLYH 728=+, RFFXU SUHGRPLQDQWO\\ ZLWK UXMLQ[LDQJFKDRZDQJ # ^ { [ ] ZKLFK DUH VWURQJO\\ GLUHFWLRQRULHQWHG EXW ',85(1* RFFXU PRUH FRPPRQO\\ ZLWK ]DLGDR ZKLFK DUH OHVV VSHFLILF LQ GLUHFWLRQDOLW\\ $V VKRZQ LQ WKH KLJKOLJKWHG SRUWLRQV LQ EHORZ WDNHQ WRJHWKHU RYHU RI WKH XVHV RI 728=+, WDNH D GLUHFWLRQDO ORFDWLYH /RFDWLYH 0DUNHUV 3UHIDFLQJ WKH 3DWK LQ 728=+, YV WKDW RI ',85(1* 'LUHFWLRQDO /RFDWLYHV UXMLQ[LDQJFKDRZDQJ #^ { [ 1RQGLUHFWLRQDO /RFDWLYHV ]DL\\XGDR \u00cd ) 728 =+, ',8 5(1* 7KHUHIRUH VXPPLQJ XS WKH DERYH GLVFXVVLRQ ZH FRQFOXGH WKDW =+, EHORQJV WR WKH 728JURXS VLQFH ERWK DUH &gt; 3DWKHQGSRLQW@ 5(1* EHORQJV WR WKH ',8JURXS VLQFH ERWK DUH &gt; 3DWKHQGSRLQW@ 7HQWDWLYH &amp;RQFOXVLRQ 728=+, YV ',85(1* 728=+, 3DWKHQGSRLQW VSHFLILHG VWURQJO\\ GLUHFWLRQDO ',85(1* 3DWKHQGSRLQW XQVSHFLILHG QRQGLUHFWLRQDO HQGSRLQW DV VKRZQ LQ /RFDWLYH 0DUNHUV 7\\SLFDOO\\ )ROORZLQJ 728 YV =+, UXMLQ #^\u00b1 \u00b5LQWR \u00b6 [LDQJFKDRZDQJ {[\u00b1 \u00b5WRZDUG \u00b6 ]DL\\X \u00cd\u00b1 \u00b5DW \u00b6 GDR ) \u00b5WR \u00b6 728 =+, $QRWKHU LQWHUHVWLQJ GLIIHUHQFH EHWZHHQ 728 DQG =+, LV WKDW =+, RIWHQ RFFXUV DV WKH VHFRQG YHUE LQ D FRJQDWH 99 FRPSRXQG LQGLFDWLQJ WKDW WKH HYHQW RI =+, LV FDWHJRULDOO\\ OHVVPDUNHG DQG OH[LFDOO\\ OHVVVSHFLILHG ZLWK PDQQHU VLQFH WKH ILUVW YHUE LQ WKH FRJQDWH 99 FRPSRXQG LV PRUH PDQQHUVSHFLILF DV VKRZQ LQ =+, DV WKH GHIDXOW 9 LQ FRJQDWH 99 FRPSRXQGV DOO PHDQLQJ WR WKURZ D WRX]KL \u00aa E UHQJ]KL \u00aa F GLX]KL G\u00aa G SDR]KL \u00b9\u00aa 7KH DERYH REVHUYDWLRQ FRQFHUQLQJ WKH PRUSKRV\\QWDFWLF GLIIHUHQFHV EHWZHHQ 728 DQG =+, VHHPV WR SRLQW WR D ILQHU GLVWLQFWLRQ 728 LV VHPDQWLFDOO\\ PRUH ORDGHG ZLWK D IXUWKHU VSHFLILFDWLRQ RI WKH VSDWLDO ERXQGHGQHVV RI LWV 3DWKHQGSRLQW ZKLOH =+, LV OH[L FDOO\\ OHVV LQIRUPDWLYH DV VXPPDUL]HG LQ 7HQWDWLYH &amp;RQFOXVLRQ /H[LFDOVHPDQWLF 'LVWLQFWLRQ EHWZHHQ 728 DQG =+, 728 3DWKHQGSRLQW 6SDWLDOO\\ ERXQGHG =+, 3DWKHQGSRLQW &amp;DXVDWLYHLQWUDQVLWLYH 8VH RI ',8 ZRGH JDQJEL GLXUHQJ OH z5\u00dd/ 1 0\\ SHQ LV ORVW z5\u00dd~1 7KH SRVVLEOH LQFOXVLRQ RI DQ (YHQWHQGSRLQW LQ WKH XVH RI ',8 JLYHV ULVH WR WKH SRWHQWLDO DPELJXLW\\ RI D ,QWHUSUHWDWLRQDO 'LIIHUHQFHV D ZR GLX OH \\L]KL JDQJEL zG1+5\u00dd T ORVW LQFKRDWLYH VWDWLYH UHVXOW FRQWURO T WKURZQ DZD\\ FRPSOHWLYH DFWLYH UHVXOW FRQWURO E ZR UHQJ OH \\L]KL JDQJEL z1+5\u00dd T WKURZQ DZD\\ FRPSOHWLYH DFWLYH UHVXOW FRQWURO *LYHQ LWV VWDWLYH XVH WKH YHUE ',8 PD\\ RFFFXU DV WKH UHVXOWDWLYH FRPSOHPHQW LQ D 9HUE5HVXOWDWLYH FRPSRXQG ZRGH JDQJEL JDR',8JDR5(1* OH z5\u00dd\u00f6G \u00f61 0\\ SHQ JRW ORVW :H VHH WKDW ',8 DVSHFWXDO HPSKDVLV WR EH SODFHG RQ WKH (YHQWHQGSRLQW 'LVWLQFWLRQ EHWZHHQ ',8 DQG 5(1* ',8 (YHQWHQGSRLQW 5(1* (YHQWHQGSRLQW 'LVWLQFWLRQV %DVHG RQ WKH 7ZR 7\\SHV RI (QGSRLQW $V D QHDUV\\QRQ\\P VHW WKH IRXU YHUEV 728 =+, ',8 5(1* GHPRQVWUDWH D WZRZD\\ FRQWUDVW LQ WHUPV RI WKHLU VSHFLILFDWLRQ RI 3DWKHQGSRLQW DQG (YHQWHQGSRLQW 7KH 'LVWLQFWLRQ EDVHG RQ 3DWK(QGSRLQW YV (YHQW(QGSRLQW 728 =+, ',8 5(1* 3DWK(QGSRLQW ERXQGHG (YHQW(QGSRLQW 6R IDU YHUEV VWXGLHG KHUH ,W</td></tr></table>",
+                "num": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-3002.json

@@ -0,0 +1,1553 @@
+{
+    "paper_id": "O00-3002",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:07.489738Z"
+    },
+    "title": "Design and Evaluation of Approaches to Automatic Chinese Text Categorization",
+    "authors": [
+        {
+            "first": "Jyh-Jong",
+            "middle": [
+                "J"
+            ],
+            "last": "Tsay",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Chung Cheng University",
+                "location": {
+                    "postCode": "62107",
+                    "settlement": "Chiayi",
+                    "country": "Taiwan, ROC"
+                }
+            },
+            "email": "tsay@cs.ccu.edu.tw"
+        },
+        {
+            "first": "Jing-Doo",
+            "middle": [
+                "D"
+            ],
+            "last": "Wang",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Chung Cheng University",
+                "location": {
+                    "postCode": "62107",
+                    "settlement": "Chiayi",
+                    "country": "Taiwan, ROC"
+                }
+            },
+            "email": "jdwang@cs.ccu.edu.tw"
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "In this paper, we propose and evaluate approaches to categorizing Chinese texts, which consist of term extraction, term selection, term clustering and text classification. We propose a scalable approach which uses frequency counts to identify left and right boundaries of possibly significant terms. We used the combination of term selection and term clustering to reduce the dimension of the vector space to a practical level. While the huge number of possible Chinese terms makes most of the machine learning algorithms impractical, results obtained in an experiment on a CAN news collection show that the dimension could be dramatically reduced to 1200 while approximately the same level of classification accuracy was maintained using our approach. We also studied and compared the performance of three well known classifiers, the Rocchio linear classifier, naive Bayes probabilistic classifier and k-nearest neighbors(kNN) classifier, when they were applied to categorize Chinese texts. Overall, kNN achieved the best accuracy, about 78.3%, but required large amounts of computation time and memory when used to classify new texts. Rocchio was very time and memory efficient, and achieved a high level of accuracy, about 75.4%. In practical implementation, Rocchio may be a good choice.",
+    "pdf_parse": {
+        "paper_id": "O00-3002",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "In this paper, we propose and evaluate approaches to categorizing Chinese texts, which consist of term extraction, term selection, term clustering and text classification. We propose a scalable approach which uses frequency counts to identify left and right boundaries of possibly significant terms. We used the combination of term selection and term clustering to reduce the dimension of the vector space to a practical level. While the huge number of possible Chinese terms makes most of the machine learning algorithms impractical, results obtained in an experiment on a CAN news collection show that the dimension could be dramatically reduced to 1200 while approximately the same level of classification accuracy was maintained using our approach. We also studied and compared the performance of three well known classifiers, the Rocchio linear classifier, naive Bayes probabilistic classifier and k-nearest neighbors(kNN) classifier, when they were applied to categorize Chinese texts. Overall, kNN achieved the best accuracy, about 78.3%, but required large amounts of computation time and memory when used to classify new texts. Rocchio was very time and memory efficient, and achieved a high level of accuracy, about 75.4%. In practical implementation, Rocchio may be a good choice.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "In recent years, we have seen a tremendous growth in the number of online text documents available on the Internet, in digital libraries and news sources. Effective location of information in these huge resources is difficult without good indexing as well as organization of text collections. Automatic text categorization, which is defined as the task of assigning predefined class (category) labels to free text documents, is one of The objective of this study was to design and evaluate approaches to categorizing Chinese texts. In particular, we implemented and evaluated approaches which consist of the following processes: term extraction, term selection, term clustering and text classification. Note that in Chinese texts, although a sentence is composed of a sequence of terms, no white spaces are inserted to separate terms from each other. Term extraction which segments sentences into term sequences is a difficult task [5] . Several approaches have been proposed to extract terms from Chinese texts [4, 13] . In this paper, we propose a scalable approach [17] which is based on String B-trees proposed in [7] and is capable of handling huge numbers of text documents. Our approach uses frequency counts to identify possible term boundaries as proposed in [13] and is able to identify new terms which occur very often in Chinese texts.",
+                "cite_spans": [
+                    {
+                        "start": 932,
+                        "end": 935,
+                        "text": "[5]",
+                        "ref_id": "BIBREF4"
+                    },
+                    {
+                        "start": 1012,
+                        "end": 1015,
+                        "text": "[4,",
+                        "ref_id": "BIBREF3"
+                    },
+                    {
+                        "start": 1016,
+                        "end": 1019,
+                        "text": "13]",
+                        "ref_id": "BIBREF12"
+                    },
+                    {
+                        "start": 1068,
+                        "end": 1072,
+                        "text": "[17]",
+                        "ref_id": "BIBREF16"
+                    },
+                    {
+                        "start": 1118,
+                        "end": 1121,
+                        "text": "[7]",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 1268,
+                        "end": 1272,
+                        "text": "[13]",
+                        "ref_id": "BIBREF12"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "However, the number of terms in Chinese can be very large. It is very easy to encounter 10 6 or even more terms in moderately-sized collections. The huge number of possible terms results in very high dimensionality when documents was presented in a vector space model and makes many machine learning algorithms impractical. To reduce the dimension to a practical level, we propose to perform term selection and term clustering on extracted terms. In particular, we use the \u03c7 2 statistic [16] to select terms that are highly correlated to class categories. In [16] , we presented an extensive comparison of several measures for term selection in Chinese text categorization, such as the odds ratio, information gain, mutual information, and \u03c7 2 statistic. Experimental results shows that the \u03c7 2 statistic approach achieves the best performance. Notice that in term selection, if only a small number of terms is selected, a document may contain very few or even none of the selected terms, and thus will be classified into the default class. On the other hand, a large number of selected terms make automatic categorization computationally impractical. We thus allow a large number of terms to be selected and then perform term clustering to group similar terms into clusters.",
+                "cite_spans": [
+                    {
+                        "start": 91,
+                        "end": 92,
+                        "text": "6",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 487,
+                        "end": 491,
+                        "text": "[16]",
+                        "ref_id": "BIBREF15"
+                    },
+                    {
+                        "start": 559,
+                        "end": 563,
+                        "text": "[16]",
+                        "ref_id": "BIBREF15"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "A large number of algorithms for clustering are Available [11] . Most of them are unsupervised and ignore any class labels that are given. In this study, we used distributional clustering [2] , which explicitly takes advantage of the class labels to group terms with similar class distributions into the same cluster. In an experiment on a collection of CNA news [1] articles, the number of terms extracted was 548363.",
+                "cite_spans": [
+                    {
+                        "start": 58,
+                        "end": 62,
+                        "text": "[11]",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 188,
+                        "end": 191,
+                        "text": "[2]",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 363,
+                        "end": 366,
+                        "text": "[1]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Experimental results show that the level of classification accuracy could be maintained while the dimension was reduced to 1200 by selecting 90000 terms first and then clustering them into 1200 clusters. Notice that term selection and term clustering also can compeensate for imprecision in term extraction as erroneous terms can be dropped out during term selection or grouped with more significant terms through term clustering. In addition to term selection and term clustering algorithms, there are others which can be applied to reduce the level of dimensionality, such as Principle Component Analysis (PCA) [6] . PCA is an unsupervised dimensional reduction technique, whereas distributional clustering is supervised and can take advantage of class labels to concentrate effort on the specific task of categorization. We expect distributional clustering to perform well in the context of text categorization.",
+                "cite_spans": [
+                    {
+                        "start": 613,
+                        "end": 616,
+                        "text": "[6]",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we also compare extensively three well-known classifiers, including the Rocchio linear classifier [12] , naive Bayes probabilistic classifier, and k-nearest neighbor (kNN) classifier [20] . We observed in an experiment that the classification accuracy of Rocchio and kNN improved slightly as the dimension was reduced to 1200 by means of term selection and term clustering but that the accuracy of the naive Bayes classifier dropped slightly. This might have been due to the fact that term clustering refines the shapes of each cluster but distorts the distribution of each term. Overall, kNN achieved the best accuracy, about 78.3%, but required large amounts of computation time and memory when used to classify new texts. Rocchio is very time and memory efficient, and achieves accuracy of about 75.4%, which is slightly worse than kNN.",
+                "cite_spans": [
+                    {
+                        "start": 113,
+                        "end": 117,
+                        "text": "[12]",
+                        "ref_id": "BIBREF11"
+                    },
+                    {
+                        "start": 198,
+                        "end": 202,
+                        "text": "[20]",
+                        "ref_id": "BIBREF19"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Recently, Huang et al. [10] evaluated the weight matrix approach, which estimates the relative importance of the keywords in each class and classifies a test news to the class that maximizes the sum of the weights of keywords appearing in that news. Although they achieved about 88% classification accuracy, their experiment was different from ours as well as those used in much related research [3, 22] . First, the training news did not come from the same news source as the test news, but come from a thesaurus [19] that was carefully built by linguistic specialists. Second, the test news was classified by readers who could employ logic that was close to that assumed by the classification algorithms but different from that employed by the editors. Third, a piece of test news could be assigned to multiple classes when it covered topics from different classes. In fact, for a collection of 1136 news items, 1380 class labels were assigned, which indicates that about 20% of the test news iteems had multiple class labels. However, in the CNA news collection used in this study, each news items had exactly one predefined class no matter how many topics it covered. It is not clear whether or not the weight matrix approach can achieve the same performance when all the differences are removed.",
+                "cite_spans": [
+                    {
+                        "start": 23,
+                        "end": 27,
+                        "text": "[10]",
+                        "ref_id": "BIBREF9"
+                    },
+                    {
+                        "start": 396,
+                        "end": 399,
+                        "text": "[3,",
+                        "ref_id": "BIBREF2"
+                    },
+                    {
+                        "start": 400,
+                        "end": 403,
+                        "text": "22]",
+                        "ref_id": "BIBREF21"
+                    },
+                    {
+                        "start": 514,
+                        "end": 518,
+                        "text": "[19]",
+                        "ref_id": "BIBREF18"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "The remainder of this paper is organized as follows. Section 2 sketches the String B-tree approach to term extraction. Section 3 describes the\u03c7 2 statistic approach to term selection. Section 4 describes distributional clustering. Section 5 reviews the classifiers compared in this paper. Section 6 gives experimental results. Section 7 gives conclusions.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we propose a scalable approach [18] to term extraction, which is based on String B-trees (SB-trees) [7] . This approach can handle large text collections and can identify newly created terms frequently found in Chinese. It does not use a dictionary but rather uses frequency counts to identify the boundaries of possible terms as in [13] . We will describe the term extraction method in the following.",
+                "cite_spans": [
+                    {
+                        "start": 46,
+                        "end": 50,
+                        "text": "[18]",
+                        "ref_id": "BIBREF17"
+                    },
+                    {
+                        "start": 115,
+                        "end": 118,
+                        "text": "[7]",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 348,
+                        "end": 352,
+                        "text": "[13]",
+                        "ref_id": "BIBREF12"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Extraction",
+                "sec_num": "2."
+            },
+            {
+                "text": "Let w be a string. For any character x, let P (wx|w) be the probability that w is followed by x, and let P(xw|w) be the probability that w is preceded by x. We say that w passes right boundary verification if ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Extraction",
+                "sec_num": "2."
+            },
+            {
+                "text": "= = \u03b8 \u03b8",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Extraction",
+                "sec_num": "2."
+            },
+            {
+                "text": ", which means that w will be identified as a significant term when it has at least two distinct successor and predecessor characters. For each class, we build two SB-trees, one for all the suffixes [8] of the original texts used for right boundary verification, and the other for the suffixes of the reversed texts which is used for left boundary verification. Notice that SB-trees are scalable; they can maintain dynamic collections and identify new terms as new articles are inserted.",
+                "cite_spans": [
+                    {
+                        "start": 198,
+                        "end": 201,
+                        "text": "[8]",
+                        "ref_id": "BIBREF7"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Extraction",
+                "sec_num": "2."
+            },
+            {
+                "text": "Term selection is performed to choose representative terms for each class such that these terms can distinguish one class from the others. After the term extraction process is completed, there are many terms remain that are not informative for categorization. In [16] , we extensively compared several measures used for term selection in Chinese text categorization, such as the odds ratio, information gain, mutual information, and \u03c7 2 statistic. Experimental results show that the \u03c7 2 statistic approach achieves the best performance when combined with the naive Bayes classifier. In this study, we used the \u03c7 2 statistic [21] approach to perform term selection.",
+                "cite_spans": [
+                    {
+                        "start": 263,
+                        "end": 267,
+                        "text": "[16]",
+                        "ref_id": "BIBREF15"
+                    },
+                    {
+                        "start": 624,
+                        "end": 628,
+                        "text": "[21]",
+                        "ref_id": "BIBREF20"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Selection",
+                "sec_num": "3."
+            },
+            {
+                "text": "For a term t and a class c, the \u03c7 2 statistic measures the correlation between t and c. Let A be the number of times t and c co-occur, let B be the number of times t occurs without c, let P be the number of times c occurs without t, let Q be the number of times neither t nor c occur, and let N be the total number of documents. The \u03c7 2 statistic is defined as",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Selection",
+                "sec_num": "3."
+            },
+            {
+                "text": ") ( ) ( ) ( ) ( ) ( ) , ( 2 2 Q P B A Q B P A BP AQ N c t + \u00d7 + \u00d7 + \u00d7 + \u2212 \u00d7 = \u03c7 .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Selection",
+                "sec_num": "3."
+            },
+            {
+                "text": "Notice that the \u03c7 2 statistic approach prefers terms that are highly correlated with a particular class. For each term, the \u03c7 2 statistic scores with regard to different classes can be different. In [21] , Yang used the",
+                "cite_spans": [
+                    {
+                        "start": 199,
+                        "end": 203,
+                        "text": "[21]",
+                        "ref_id": "BIBREF20"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Selection",
+                "sec_num": "3."
+            },
+            {
+                "text": "average or the maximum of the scores to select representative terms, which may result in a biased distribution of selected terms between classes. To avoid this situation, we select from each class the same number of terms having the largest\u03c7 2 statistic in that class.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Design and Evaluation of Approaches to Automatic Chinese Text Categorization 47",
+                "sec_num": null
+            },
+            {
+                "text": "We perform term clustering to further reduce the dimension of the vector space after the term selection process. In order to avoid the situation in which a document contains none of the selected terms, in term selection, we select a suitable large set of terms which may require a large amount of computation time and memory for classification. Term clustering groups similar terms into one cluster that no longer distinguishes between constituent terms. In this study, we used distributional clustering [2] , which groups terms with similar distributions over classes into the same cluster. Note that distributional clustering can compensate for the drawback of term extraction, where incomplete terms are clustered into the group containing their original terms. On the other hand, when training data is sparse, performance may be improved by averaging statistics of similar words together so that the resulting estimates are more robust. We describe distributional clustering [2] in more detail in the following.",
+                "cite_spans": [
+                    {
+                        "start": 504,
+                        "end": 507,
+                        "text": "[2]",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 979,
+                        "end": 982,
+                        "text": "[2]",
+                        "ref_id": "BIBREF1"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": "Term clustering algorithms define a similarity measure between terms and group similar terms into term clusters. In distributional clustering, the difference between two term distributions is measured by Kullback-Leibler (KL) divergence. For term i t and term j t , the KL divergence, denoted as",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": "( ( | ) || ( | )) i j D P C t P C t , is defined as ) | ( ) | ( log ) | ( | | 1 j k i k C k i k t C P t C P t C P \u2211 = \u2212",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": ", where |C| is the number of classes and",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": ") | ( i k t C P",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": "is the probability of class k C given term i t",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": ". To avoid the odd properties of KL divergence, such as asymmetry, we use the average KL divergence defined as",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": ")) | ( || ) | ( ( ) ( ) ( )) | ( || ) | ( ( ) ( ) ( j i j j i j j i i j i i t t C P t C P D t t P t P t t C P t C P D t t P t P \u2228 \u22c5 \u2228 + \u2228 \u22c5 \u2228 , where j i t t \u2228 represents",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": "clustering of term i t and term j t into one group. Based on the average KL divergence, we apply a simple greedy agglomerative algorithm to cluster terms as follows. Let M be the number of final clusters.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": "Initially, M terms are selected as seeds. Each term represents a singleton cluster. The following process is repeated until all the terms have been added: the two most similar clusters are merged into one cluster, and then the term that has the highest \u03c7 2 statistic measure among the remaining terms is added as a singleton cluster. The initial M seeding terms are uniformly selected from all classes. That is, from each class, the",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": "| | C M",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": "terms that have the highest \u03c7 2 statistic measure are selected as initial seeds. This avoids the problem of bias [2] , where the M initial clusters may prefer some classes.",
+                "cite_spans": [
+                    {
+                        "start": 113,
+                        "end": 116,
+                        "text": "[2]",
+                        "ref_id": "BIBREF1"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Term Clustering",
+                "sec_num": "4"
+            },
+            {
+                "text": "In this paper, we compare three wellknown classifiers, including the Rocchio linear classifier, naive Bayes (NB) probabilistic classifier and k-nearest neighbor (kNN) classifier, which are reviewed in the following sections.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Classifiers",
+                "sec_num": "5."
+            },
+            {
+                "text": "The Rocchio algorithm is a training algorithm [12] ",
+                "cite_spans": [
+                    {
+                        "start": 46,
+                        "end": 50,
+                        "text": "[12]",
+                        "ref_id": "BIBREF11"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rocchio Linear Classifier",
+                "sec_num": "5.1"
+            },
+            {
+                "text": "| | | | k C D D i k C D i i C D D C D G k i k i \u2212 \u2212 = \u2211 \u2211 \u2212 \u2208 \u2208 \u03b7 ,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rocchio Linear Classifier",
+                "sec_num": "5.1"
+            },
+            {
+                "text": "The Naive Bayes (NB) probabilistic classifiers have been studied for application to machine learning [14] . The basic idea in NB is to use the joint probabilities of terms and classes to estimate the probabilities of classes given a document. The naive part is the assumption of term independence, i.e., the conditional probability of a term, given a class, is assumed to be independent from the conditional probabilities of other words given that class. This assumption makes computation for NB classifiers far more efficient than that for the non-naive Bayes approaches [20] whose time complexity are exponential.",
+                "cite_spans": [
+                    {
+                        "start": 101,
+                        "end": 105,
+                        "text": "[14]",
+                        "ref_id": "BIBREF13"
+                    },
+                    {
+                        "start": 572,
+                        "end": 576,
+                        "text": "[20]",
+                        "ref_id": "BIBREF19"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": "Let X be a request document; NB assigns to X the most probable class NB",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": "C defined as arg max ( | ) k NB c c k C P C X \u2208 =",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": ". By Bayes' theorem,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": "\u2211 \u2208 = C c i i k k k i C P C X P C P C X P X C P ) ( ) | ( ) ( ) | ( ) | (",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": ". Due to the assumption of term independence,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": ") | ( ) | ( | | 1 k j X j k C t P C X P = \u03a0 = , where ) | ( k j C t P",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": "is the conditional probability of term j t given class k C . Notice that the above equation works well when every term appears in every document. However, the product becomes 0 when some terms do not appear in the given document. We use",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": "\u2211 + + = | | ) , ( | | ) , ( 1 ) | ( T j k j k j k j C t TF T C t TF C t P in order to approximate ) ( | j k P t C",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": "to avoid the possibility that the product will become 0, where ) , ( ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": ") | ( ) ( ) | ( ) ( ) | ( X t TF i j X t i i X t TF k j X t k k j j j j C t P C P C t P C P X C P \u2208 \u2208 \u03a0 \u03a0 = \u2211 .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Naive Bayes (NB) Classifier",
+                "sec_num": "5.2"
+            },
+            {
+                "text": "Given an arbitrary request document X, kNN ranks its nearest neighbors among the training documents and uses the classes of the k top-ranking neighbors to predict the classes of X. The similarity score of each neighbor document when it is compared to X is used as the weight of the class of the neighboring document, and the sum of the class weights over the k nearest neighbors is used to perform class ranking [20] . , respectively. To conduct categorization, the cosine similarity between each i D and X is calculated. The training documents are sorted using the cosine similarity metric in descending order. Then the k top-ranking documents are selected. The final score of the request document X when compared to each class is calculated by summing the cosine similarity metric of these k selected documents and their class association. The class with the highest score is assigned to X. We have performed an experiment using different values of k, including 5, 10, 15, 20, 30, 50, 100, 150, 200 and 300. The best choice of k in our experiment is 15 when n = 90000 and is 10 when n = 1200.",
+                "cite_spans": [
+                    {
+                        "start": 412,
+                        "end": 416,
+                        "text": "[20]",
+                        "ref_id": "BIBREF19"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "k-Nearest Neighbor (kNN) Classifier",
+                "sec_num": "5.3"
+            },
+            {
+                "text": "In our experiment, we used Chinese news articles from the Central News Agency (CNA) [1] . We used news articles spanning a period of one year, from 1/1/1991 to 12/31/1991, to extract terms. News articles from the six-month period 8/1/1991 to 1/31/1992 were used as training data to train classifiers. The testing data consisted of news articles from the one-month period 2/1/1992 to 2/28/1992. All the news articles were preclassified into 12 classes, listted in Figure 1 . Note that the number of texts used was far larger than that employed in previous related researches [10, 22] . As a result, the conclusions drawn based on our experimental results are believed to be more reliable.",
+                "cite_spans": [
+                    {
+                        "start": 84,
+                        "end": 87,
+                        "text": "[1]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 574,
+                        "end": 578,
+                        "text": "[10,",
+                        "ref_id": "BIBREF9"
+                    },
+                    {
+                        "start": 579,
+                        "end": 582,
+                        "text": "22]",
+                        "ref_id": "BIBREF21"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 463,
+                        "end": 471,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Experimental Results",
+                "sec_num": "6."
+            },
+            {
+                "text": "The news articles were not uniformly distributed over the classes, as shown in Figure 1 . We, thus, measure the classification accuracy at both micro and macro levels. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 79,
+                        "end": 87,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Figure 1 The distribution of CAN news articles.",
+                "sec_num": null
+            },
+            {
+                "text": "We performed term extraction, term selection and term clustering to reduce the dimension. Both the space and time required to classify new documents could be reduced as the dimension of the vector space was reduced. Figure 2 shows the time needed to classify new documents, measured on a PC with a Pentium II 233 CPU, 128MB RAM and an IDE HardDisk, for dimension n = 90000 and 1200, respectively.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 216,
+                        "end": 224,
+                        "text": "Figure 2",
+                        "ref_id": "FIGREF5"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Dimension Reduction",
+                "sec_num": "6.1"
+            },
+            {
+                "text": "In the term extraction process, terms that appeared fewer than 10 times or in only one document were dropped out. We then used frequency counts to identify significant terms. The number of significant terms extracted was 548363. Term selection was then performed to select a subset of most representative terms. In order to find an appropriate number p of selected terms, we experimented for different values of p, including 12000, 36000, 60000, 90000 and 120000. We choose a p value of 90000 because kNN and NB achieved the best MicroAccuracy results of 77.12% and 76.45%, respectively, when p was 90000, as indicated in Figure 3 . The selected terms were clustered using distributional clustering into term clusters. To choose a suitable number c of term clusters, we experimented with different values of c, including 120, 240, 360, 600, 900, 1200, 1800, 2400, 3600 and 4800. We choose a c value of 1200 because kNN and Rocchio achieved the best performance when c was 1200, as shown in Figure 4 . Figure 5 shows some examples of term groups. In addition to clustering similar terms to reduce the dimension, term clustering can also cluster redundant substrings that are erroneously identified during term extraction into the group that contains their original terms. For example, as shown in Figure6,\"\u4e8c\u5c46\u570b\" and \"\u4e8c\u5c46\u570b\u4ee3\" are clustered into group 12; \"\u8b49\u5238\u4ea4\uf9e0\u6240\" and \"\u5238\u4ea4\uf9e0\u6240\" are clustered into group 300. On the other hand, the averaging statistics of similar words may result in more robust estimates. For example, \"\uf983\ufa08\u696d\"(a travel agent) and \"\uf983\u904a\u5354\u6703\"(a travel agency association) are similar words and are clustered into group 100.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 622,
+                        "end": 630,
+                        "text": "Figure 3",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 990,
+                        "end": 998,
+                        "text": "Figure 4",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 1001,
+                        "end": 1009,
+                        "text": "Figure 5",
+                        "ref_id": "FIGREF7"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Dimension Reduction",
+                "sec_num": "6.1"
+            },
+            {
+                "text": "In [2] , Baker claimed that performance can be improved by means of term clustering when training data is sparse because by averaging statistics of similar words, more robust estimates can be obtained. This was confirmed by our experiment. Note that our training data was quite sparse as the average number of none-zero items in training vectors was 106 when n is 90000, and was 79 when c is 1200. The memory space could be reduced by 25% ",
+                "cite_spans": [
+                    {
+                        "start": 3,
+                        "end": 6,
+                        "text": "[2]",
+                        "ref_id": "BIBREF1"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Dimension Reduction",
+                "sec_num": "6.1"
+            },
+            {
+                "text": "Overall, kNN achieved the best MicroAccuracy results, and Rocchio achieved slightly worse results, as shown in Figure 7 and Figure 8 . Note that the MicroAccuracy results for Rocchio and kNN improved slightly from 75.24% and 77.12% to 75.39% and 78.33%, respectively, when the dimension of the vector space was reduced from 90000 to 1200 by means of distributional clustering. However, the performance of na\u00efve Bayes dropped when terms were clustered. This might have been due to the fact that naive Bayes is more sensitive to term distributions which might be distorted by term clustering. kNN prefered large classes as its MacroAccuracy result, 73.88%, was the lowest, but its MicroAccuracy result, 77.12%, was the best, as indicated in Figure 7 . For highly related classes, kNN may prefer a larger class as the probability that the k nearest neighbors will belong to the larger class is higher. kNN achieved much better recall results than Rocchio for the class Politics (\u653f\u6cbb), which was the largest class in our news collections. However, Rocchio achieved much better recall results than kNN did for the class Military (\u8ecd\u4e8b). Note that the class Politics (\u653f\u6cbb) and the class Military (\u8ecd\u4e8b) were highly correlated, as observed in [17] , and that the class Politics (\u653f\u6cbb) was 5 times larger than the class Military (\u8ecd\u4e8b).",
+                "cite_spans": [
+                    {
+                        "start": 1230,
+                        "end": 1234,
+                        "text": "[17]",
+                        "ref_id": "BIBREF16"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 111,
+                        "end": 119,
+                        "text": "Figure 7",
+                        "ref_id": "FIGREF10"
+                    },
+                    {
+                        "start": 124,
+                        "end": 132,
+                        "text": "Figure 8",
+                        "ref_id": "FIGREF11"
+                    },
+                    {
+                        "start": 739,
+                        "end": 747,
+                        "text": "Figure 7",
+                        "ref_id": "FIGREF10"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Classifiers Comparison",
+                "sec_num": "6.2"
+            },
+            {
+                "text": "In practical implementation, Rocchio could be a good choice. Rocchio is quite time and memory efficient because the time and memory requirements for the classification process are proportional to the number of classes. However, the time and memory requirements for kNN are proportional to the number of training documents. Rocchio is more noise tolerant than kNN and NB, as shown by the fact that the performance of kNN and NB worsened but the performance of Rocchio improved when n was changed from 90000 to 120000, as shown in Figure 3 . Rocchio produced slightly worse MicroAccuracy results than kNN did, but can be improved to produce results approaching the performance of kNN by taking more than one representative to represent each class in [17] .",
+                "cite_spans": [
+                    {
+                        "start": 748,
+                        "end": 752,
+                        "text": "[17]",
+                        "ref_id": "BIBREF16"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 529,
+                        "end": 537,
+                        "text": "Figure 3",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Classifiers Comparison",
+                "sec_num": "6.2"
+            },
+            {
+                "text": "In this paper, we have proposed and evaluated approaches to categorizing Chinese texts, which consist of term extraction, term selection, term clustering and text classification. For term extraction, we have proposed an approach based on String B-trees. It is scalable and is capable of handling very large numbers of text collections. We use the \u03c7 2 statistic to perform term selection and use distributional clustering to perform term clustering to reduce the dimension of the vector space. Although many redundant terms are identified as significant terms during the term extraction process, the combination of term selection and term clustering somehow can compensate for this drawback by either filtering them out or clustering them into the group containing their original terms. Results of an experiment on a CNA news collection shows that the dimension could be reduced from 90000 to 1200 while approximately the same level of classification accuracy was maintained. We have also studies and compared the performance of three well known classifiers, the Rocchio linear classifier (Rocchio), naive Bayes (NB) probabilistic classifier and k-nearest neighbors (kNN) classifier, when they were applied to categorize Chinese texts. Overall, kNN achieved the best accuracy, about 78.3%, but required large amounts of computation time and memory to classify new texts. Rocchio was very time and memory efficient, and achieved accuracy of about 75.4%. In practical implementation, Rocchio may be a good choice. In addition, we have recently shown [17] that the performance of the Rocchio linear classifier can be improved to approximate that of kNN by taking multiple representative vectors to represent one class.",
+                "cite_spans": [
+                    {
+                        "start": 1547,
+                        "end": 1551,
+                        "text": "[17]",
+                        "ref_id": "BIBREF16"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions",
+                "sec_num": "7."
+            }
+        ],
+        "back_matter": [
+            {
+                "text": "We would like to thank Dr. Chien, Lee-Feng and Mr. Lee, Min-Jer for kind help in gathering the CNA news articles.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Acknowledgements",
+                "sec_num": null
+            }
+        ],
+        "bib_entries": {
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Distributional clustering of words for text classification",
+                "authors": [
+                    {
+                        "first": "Douglas",
+                        "middle": [],
+                        "last": "Baker",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Kachites",
+                        "middle": [],
+                        "last": "Mccallum",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the 21th Ann Int ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR'98)",
+                "volume": "",
+                "issue": "",
+                "pages": "96--103",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Douglas Baker and Kachites McCallum. \"Distributional clustering of words for text classification.\" In Proceedings of the 21th Ann Int ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR'98), pages 96-103. 1998.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "PAT-tree-based online corpus collection and classification",
+                "authors": [
+                    {
+                        "first": "Chen",
+                        "middle": [],
+                        "last": "Chun",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "-",
+                        "middle": [],
+                        "last": "Liang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Lee-Feng",
+                        "middle": [],
+                        "last": "Chien",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "The Fourth International Workshop on Information Retrieval with Asian Languages(IRAL'99)",
+                "volume": "",
+                "issue": "",
+                "pages": "78--82",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen Chun-Liang and Lee-Feng Chien. \"PAT-tree-based online corpus collection and classification.\" In The Fourth International Workshop on Information Retrieval with Asian Languages(IRAL'99), pages 78-82. 1999.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "PAT-Tree-Based keyword extraction for Chinese information retrieval",
+                "authors": [
+                    {
+                        "first": "Chien",
+                        "middle": [],
+                        "last": "Lee-Feng",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the 20 th Ann Int ACM SIFIR Conference on Research and Development in Information Retrieval(SIGIR'97)",
+                "volume": "",
+                "issue": "",
+                "pages": "50--58",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chien Lee-Feng. \"PAT-Tree-Based keyword extraction for Chinese information retrieval.\" In Proceedings of the 20 th Ann Int ACM SIFIR Conference on Research and Development in Information Retrieval(SIGIR'97), pages 50-58. 1997.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Important issues on Chinese information retrieval",
+                "authors": [
+                    {
+                        "first": "Chien",
+                        "middle": [],
+                        "last": "Lee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "-Feng",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Hsiao-Tieh",
+                        "middle": [],
+                        "last": "Pu",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Computation Linguistics and Chinese Language Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "205--221",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chien Lee-Feng and Hsiao-Tieh Pu. \"Important issues on Chinese information retrieval.\" In Computation Linguistics and Chinese Language Processing, pages 205-221. 1996.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Indexing by latent semantic analysis",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Deerwester",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "T"
+                        ],
+                        "last": "Dumais",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Landauer",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Furnas",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Harshman",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Journal of the American Society for Information Science",
+                "volume": "41",
+                "issue": "6",
+                "pages": "391--407",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "S.C. Deerwester, S.T. Dumais, T.K. Landauer, G.W.Furnas, and R.A. Harshman. \"Indexing by latent semantic analysis.\" Journal of the American Society for Information Science, 41(6):391-407. 1990.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "The String B-tree: A new data structure for string search in external memory and its application",
+                "authors": [
+                    {
+                        "first": "Paolo",
+                        "middle": [],
+                        "last": "Ferragina",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Roberto",
+                        "middle": [],
+                        "last": "Grossi",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Journal of ACM",
+                "volume": "46",
+                "issue": "2",
+                "pages": "236--280",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Paolo Ferragina and Roberto Grossi. \"The String B-tree: A new data structure for string search in external memory and its application.\" Journal of ACM, 46(2):236-280. 1999.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Information Retrieval Data Structures Algorithm",
+                "authors": [
+                    {
+                        "first": "B",
+                        "middle": [],
+                        "last": "William",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Rick",
+                        "middle": [],
+                        "last": "Frakes",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Kazman",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "William B.Frakes and Rick Kazman. Information Retrieval Data Structures Algorithm. Prentice Hall, Englewood Cliffs, New Jersey 0732. 1992.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Turning yahoo into an automatic web-page classifier",
+                "authors": [
+                    {
+                        "first": "Marko",
+                        "middle": [],
+                        "last": "Frobelink",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Dunja",
+                        "middle": [],
+                        "last": "Mladenic",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the 13 th European Conference on Aritficial Intelligence",
+                "volume": "",
+                "issue": "",
+                "pages": "473--474",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Marko Frobelink and Dunja Mladenic. \"Turning yahoo into an automatic web-page classifier.\" In Proceedings of the 13 th European Conference on Aritficial Intelligence, pages 473-474. 1998.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Automatic classification for news written in Chinese",
+                "authors": [
+                    {
+                        "first": "Yi-Ling",
+                        "middle": [],
+                        "last": "Huang Sen-Yuan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ja-Chen",
+                        "middle": [],
+                        "last": "Chou",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Lin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Computer Processing of Oriental Languages",
+                "volume": "12",
+                "issue": "2",
+                "pages": "143--159",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Huang Sen-Yuan, Yi-Ling Chou, and Ja-Chen Lin. \"Automatic classification for news written in Chinese.\" Computer Processing of Oriental Languages, 12(2):143-159. 1998.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Finding Groups in Data Analysis : An Introduction to Cluster Analysis",
+                "authors": [
+                    {
+                        "first": "Leonard",
+                        "middle": [],
+                        "last": "Kaufman",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Peter",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Rousseeuw",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Leonard Kaufman and Peter J. Rousseeuw. \"Finding Groups in Data Analysis : An Introduction to Cluster Analysis.\" John Wiley and Sons,Inc., New York. 1990.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Training algorithms for linear text classifiers",
+                "authors": [
+                    {
+                        "first": "David",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Lewis",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Robert",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Schapire",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "James",
+                        "middle": [
+                            "P"
+                        ],
+                        "last": "Callan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ron",
+                        "middle": [],
+                        "last": "Papka",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Proceedings of the 19 th Ann Int ACM SIFIR Conference on Research and Development in Information Retrieval (SIGIR'96)",
+                "volume": "",
+                "issue": "",
+                "pages": "298--306",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "David D. Lewis, Robert E. Schapire, James P. Callan, and Ron Papka. \"Training algorithms for linear text classifiers.\" In Proceedings of the 19 th Ann Int ACM SIFIR Conference on Research and Development in Information Retrieval (SIGIR'96), pages 298-306. 1996.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "A way to extract unknown words without dictionary from Chinese corpus and its applications",
+                "authors": [
+                    {
+                        "first": "Lin",
+                        "middle": [],
+                        "last": "Yih-Jeng",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ming-Shing",
+                        "middle": [],
+                        "last": "Yu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Shyh-Yang",
+                        "middle": [],
+                        "last": "Hwang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ming-Jer",
+                        "middle": [],
+                        "last": "Wu",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Research on Computational Linguistics Conference (ROCLING XI)",
+                "volume": "",
+                "issue": "",
+                "pages": "217--226",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lin Yih-Jeng, Ming-Shing Yu, Shyh-Yang Hwang, and Ming-Jer Wu. \"A way to extract unknown words without dictionary from Chinese corpus and its applications.\" In Research on Computational Linguistics Conference (ROCLING XI), pages 217-226. 1998.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Machine Learning. The McGraw",
+                "authors": [
+                    {
+                        "first": "Tom",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Mitchell",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Tom M. Mitchell. Machine Learning. The McGraw-Hill Companies, Inc. 1997 .",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "Term Weighting Revisited",
+                "authors": [
+                    {
+                        "first": "Amitabh Kumar",
+                        "middle": [],
+                        "last": "Singhal",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "PHD theses",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Amitabh Kumar Singhal. \"Term Weighting Revisited. \" PHD theses, Cornell University. 1997.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "Term selection with distributional clustering for Chinese text categorization using n-grams",
+                "authors": [
+                    {
+                        "first": "Tsay",
+                        "middle": [],
+                        "last": "Jyh",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "-Jong",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Jing-Doo",
+                        "middle": [],
+                        "last": "Wang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Research on Computational Linguistics Conference XII",
+                "volume": "",
+                "issue": "",
+                "pages": "151--170",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Tsay Jyh-Jong and Jing-Doo Wang. \"Term selection with distributional clustering for Chinese text categorization using n-grams.\" In Research on Computational Linguistics Conference XII, pages 151-170. 1999.",
+                "links": null
+            },
+            "BIBREF16": {
+                "ref_id": "b16",
+                "title": "Improving automatic Chinese text categorization by error correction",
+                "authors": [
+                    {
+                        "first": "Tsay",
+                        "middle": [],
+                        "last": "Jyh",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "-Jong",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Jing-Doo",
+                        "middle": [],
+                        "last": "Wang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "The Fifth International Workshop on Information Retrieval with Asian Languages(IRAL2000)",
+                "volume": "",
+                "issue": "",
+                "pages": "1--8",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Tsay Jyh-Jong and Jing-Doo Wang. \"Improving automatic Chinese text categorization by error correction.\" In The Fifth International Workshop on Information Retrieval with Asian Languages(IRAL2000), pages 1-8. 2000.",
+                "links": null
+            },
+            "BIBREF17": {
+                "ref_id": "b17",
+                "title": "A scalable approach for Chinese term extraction",
+                "authors": [
+                    {
+                        "first": "Jyh-Jong",
+                        "middle": [],
+                        "last": "Tsay",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Jing-Doo",
+                        "middle": [],
+                        "last": "Wang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "2000 International Computer Sympoyium(ICS2000)",
+                "volume": "",
+                "issue": "",
+                "pages": "246--253",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Jyh-Jong Tsay and Jing-Doo Wang. \"A scalable approach for Chinese term extraction.\" In 2000 International Computer Sympoyium(ICS2000), Taiwan, R.O.C, pages 246-253. 2000.",
+                "links": null
+            },
+            "BIBREF18": {
+                "ref_id": "b18",
+                "title": "The Thesaurus of Daily Wordings",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Yang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R.C.Yang. The Thesaurus of Daily Wordings. Book-Spring Publishing Company, Taiwan. 1995.",
+                "links": null
+            },
+            "BIBREF19": {
+                "ref_id": "b19",
+                "title": "A re-examination of text categorization methods",
+                "authors": [
+                    {
+                        "first": "Yang",
+                        "middle": [],
+                        "last": "Yiming",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Xin",
+                        "middle": [],
+                        "last": "Liu",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of the 22th Ann Int ACM SIFIR Conference on Research and Development in Information Retrieval(SIGIR'99)",
+                "volume": "",
+                "issue": "",
+                "pages": "42--49",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Yang Yiming and Xin Liu. \"A re-examination of text categorization methods.\" In Proceedings of the 22th Ann Int ACM SIFIR Conference on Research and Development in Information Retrieval(SIGIR'99),pages 42-49. 1999.",
+                "links": null
+            },
+            "BIBREF20": {
+                "ref_id": "b20",
+                "title": "A comparative study on feature selection in text categorization",
+                "authors": [
+                    {
+                        "first": "Yang",
+                        "middle": [],
+                        "last": "Yiming",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Jan",
+                        "middle": [
+                            "O"
+                        ],
+                        "last": "Pedersen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the Fourteenth International Conference on Machine Learning(ICML'97)",
+                "volume": "",
+                "issue": "",
+                "pages": "412--420",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Yang Yiming and Jan O.Pedersen. \"A comparative study on feature selection in text categorization.\" In Proceedings of the Fourteenth International Conference on Machine Learning(ICML'97), pages 412-420. 1997.",
+                "links": null
+            },
+            "BIBREF21": {
+                "ref_id": "b21",
+                "title": "A study of document auto-classification in mandarin Chinese",
+                "authors": [
+                    {
+                        "first": "Yang",
+                        "middle": [],
+                        "last": "Yun-Yan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Research on Computational Linguistics Conference(ROCLING VI)",
+                "volume": "",
+                "issue": "",
+                "pages": "217--233",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Yang Yun-Yan. \"A study of document auto-classification in mandarin Chinese.\" In Research on Computational Linguistics Conference(ROCLING VI), pages 217-233. 1993.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "for linear classifiers and was initially developed for information retrieval in the vector space model. The basic idea is to construct one prototype vector per class, using a training set of documents. Given a class, the training document collection consists of positive and negative examples. Positive examples are those documents belonging to that class, while negative examples are those documents not belonging to that class. The prototype vector of a class is the centroid of positive examples, tuned using negative examples. Let i D be a document in the training collection D, represented as a vector ,1",
+                "num": null
+            },
+            "FIGREF1": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "where \u03b7 is the parameter that adjusts the relative impact of positive and negative examples. We have experimented with different values for \u03b7, including 0.25, 0.5, 0.75 and 1. The best choice of \u03b7 in our experiment was found to be 0To classify a request document X, we compute the cosine similarity between X and each prototype vector i G , and assign to X the class whose prototype vector has the highest degree of cosine similarity with X. Cosine similarity is defined as",
+                "num": null
+            },
+            "FIGREF2": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "of occurrence of term j t in documents of class k C and |T| is the total number of distinct terms used in the domain of document representation. The formula used to predict the probability of class value k C for a given document X is )",
+                "num": null
+            },
+            "FIGREF3": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "In a kNN algorithm, each training document i D as well as the request document X are represented by means of vectors as )",
+                "num": null
+            },
+            "FIGREF4": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "Three performance measures were used to evaluate the performance of each classifier: MicroAccuracy, MacroAccuracy and Design and Evaluation of Approaches to Automatic Chinese Text Categorization 51 AccuracyVariance. Let |C| be the number of predefined classes, and let | | i C be the number of testing news articles that are preclassified into the ith class, and let the average of the classification accuracy within classes.",
+                "num": null
+            },
+            "FIGREF5": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "Classification time.",
+                "num": null
+            },
+            "FIGREF6": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "MicroAccuacy comparison(term selection). Design and Evaluation of Approaches to Automatic Chinese Text Categorization 53 MicroAccuacy comparison(term clustering).",
+                "num": null
+            },
+            "FIGREF7": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "Term clustering examples.",
+                "num": null
+            },
+            "FIGREF8": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "Term frequencies in each class.",
+                "num": null
+            },
+            "FIGREF9": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "averaged statistics of terms were more robust estimates when the percentage of none-zero items increased from 0.12%(=106/90000) to 6.58%(=79/1200) due to term clustering.",
+                "num": null
+            },
+            "FIGREF10": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "Recall(%)/precision(%) comparison(n=90000).",
+                "num": null
+            },
+            "FIGREF11": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "Recall(%)/precision(%) comparison(n=1200).",
+                "num": null
+            },
+            "TABREF1": {
+                "type_str": "table",
+                "content": "<table><tr><td>d d</td><td>d</td><td colspan=\"3\">in</td><td colspan=\"6\">, where</td><td/><td colspan=\"12\">j d i j i tf</td><td>be the term frequency of</td></tr><tr><td colspan=\"20\">the jth term in document i D , and let defined as</td><td/><td>d</td><td>i</td><td>,</td><td>j</td><td>=</td><td>log</td><td>2</td><td>(</td><td>tf</td><td>i</td><td>,</td><td>j</td><td>+</td><td>) 1</td><td>*</td><td>log</td><td>2</td><td>(</td><td>df N</td><td>j</td><td>)</td><td>, where N</td></tr><tr><td colspan=\"17\">is the total number of documents in the training collection.</td><td/><td/><td/><td/><td/><td/></tr><tr><td>The prototype vector</td><td colspan=\"2\">G</td><td>i</td><td>=</td><td>(</td><td>g</td><td>i</td><td>1 ,</td><td>,</td><td>g</td><td>i</td><td>,</td><td>2</td><td>, \u22c5</td><td>\u22c5</td><td>, \u22c5</td><td>g</td><td>i</td><td>,</td><td>n</td><td>)</td><td/><td>of class</td></tr></table>",
+                "html": null,
+                "num": null,
+                "text": ", is the weight assigned to the jth term and n is the dimension of the document space. To determine , i j d , we use the TF-IDF weighting method[15], which has been shown to be effective when used in the vector space model. Let ,"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O00/O00-3004.json

@@ -0,0 +1,761 @@
+{
+    "paper_id": "O00-3004",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T07:59:14.342952Z"
+    },
+    "title": "Compiling Taiwanese Learner Corpus of English",
+    "authors": [
+        {
+            "first": "Rebecca",
+            "middle": [],
+            "last": "Hsue",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Sun Yat-sen University",
+                "location": {
+                    "settlement": "Kaohsiung",
+                    "country": "Taiwan"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Hueh",
+            "middle": [],
+            "last": "Shih",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Sun Yat-sen University",
+                "location": {
+                    "settlement": "Kaohsiung",
+                    "country": "Taiwan"
+                }
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "This paper presents the mechanisms of and criteria for compiling a new learner corpus of English, the quantitative characteristics of the corpus and a practical example of its pedagogical application. The Taiwanese Learner Corpus of English (TLCE), probably the largest annotated learner corpus of English in Taiwan so far, contains 2105 pieces of English writing (around 730,000 words) from Taiwanese college students majoring in English. It is a useful resource for scholars in Second Language Acquisition (SLA) and English Language Teaching (ELT) areas who wish to find out how people in Taiwan learn English and how to help them learn better. The quantitative information shown in the work reflects the characteristics of learner English in terms of part-of-speech distribution, lexical density, and trigram distribution. The usefulness of the corpus is demonstrated by a means of corpus-based investigation of learners' lack of adverbial collocation knowledge.",
+    "pdf_parse": {
+        "paper_id": "O00-3004",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "This paper presents the mechanisms of and criteria for compiling a new learner corpus of English, the quantitative characteristics of the corpus and a practical example of its pedagogical application. The Taiwanese Learner Corpus of English (TLCE), probably the largest annotated learner corpus of English in Taiwan so far, contains 2105 pieces of English writing (around 730,000 words) from Taiwanese college students majoring in English. It is a useful resource for scholars in Second Language Acquisition (SLA) and English Language Teaching (ELT) areas who wish to find out how people in Taiwan learn English and how to help them learn better. The quantitative information shown in the work reflects the characteristics of learner English in terms of part-of-speech distribution, lexical density, and trigram distribution. The usefulness of the corpus is demonstrated by a means of corpus-based investigation of learners' lack of adverbial collocation knowledge.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "A computer corpus is a body of computerized written text or transcribed speech. Computer corpora are useful for a wide variety of research purposes, in fields such as lexicography, natural language processing, and all varieties of linguistics. The first computer corpus made its appearance in the early 1960s when two scholars at Brown University compiled a one-million-word corpus, known as the Brown Corpus [Francis & Kucera, 1964] . It contains a wide range of American English texts with grammatical annotation. For decades, this pioneering work was an important source for linguistic scholars who wished to perform quantitative as well as qualitative that is crucial for a broad coverage system. Third, a static WSD model is unlikely to be robust and portable, since it is very difficult to build a single model relevant to a wide analysis of language structure and use [Francis & Kucera, 1982] . In the early 1970s, an equivalent British collection, the Lancaster-Oslo-Bergn (LOB) Corpus, was designed and compiled to facilitate comparative studies. Quantitative information on the distribution of various linguistic features in these two corpora became available [Johansson & Norheim, 1988; Nakamura, 1993] . The two corpora and other subsequently compiled corpora are similar in structure and size, and are considered to be first generation corpora.",
+                "cite_spans": [
+                    {
+                        "start": 409,
+                        "end": 433,
+                        "text": "[Francis & Kucera, 1964]",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 875,
+                        "end": 899,
+                        "text": "[Francis & Kucera, 1982]",
+                        "ref_id": "BIBREF4"
+                    },
+                    {
+                        "start": 1170,
+                        "end": 1197,
+                        "text": "[Johansson & Norheim, 1988;",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 1198,
+                        "end": 1213,
+                        "text": "Nakamura, 1993]",
+                        "ref_id": "BIBREF13"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "With the fast development in technology needed for text capture, storage and analysis, the scale of computer corpora has increased considerably, and a corpus of one million words seems to be inadequate for large scale studies on lexis. In the early 1980s, the publisher Collins and Birmingham University compiled the first mega-size corpus, the Cobuild Corpus, for the production of a new English dictionary. The scale of the corpus reached 13-million words by the time the dictionary was published in 1987 [Collins, 1987] . In preparation for a new generation of language reference publications, the corpus was transformed into the Bank of English in 1991 and has been growing larger in size ever since. Another well-known mega-corpus, the British National Corpus, was compiled between 1991 and 1994 by a consortium of academics and publishing houses. This corpus consists of 100 million words of part-of-speech tagged contemporary written and spoken British English. Access to the corpus was originally restricted within Europe, and it was not until very recently that the corpus was made accessible worldwide. Due to the need for comparative studies of different English varieties as in the first generation, the International Corpus of English compilation project was launched in the 1990s [Greenbaum, 1996] to gather written and spoken forms of national varieties of English throughout the world. The project aims to collect up to 20 subcorpora, each containing one million words of English used in countries where English is the first language, and in countries such as India and Singapore where English is an additional office language. The corpus will enable researchers to use each national subcorpus independently for descriptive research and also to undertake comparative studies.",
+                "cite_spans": [
+                    {
+                        "start": 507,
+                        "end": 522,
+                        "text": "[Collins, 1987]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 1294,
+                        "end": 1311,
+                        "text": "[Greenbaum, 1996]",
+                        "ref_id": "BIBREF9"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "For nearly fifty years, machine-readable language corpora have greatly benefited people in both linguistics and publishing houses. Linguistic scholars have been able to better understand language structure and use with the aid of quantitative data. Publishers have produced new pedagogical tools that reflect the real use of language. However, it was not until the 1990s that scholars in the EFL and SLA sectors began to recognize the theoretical as well as practical potential of corpora and to believe that with the aid of quantitative information, computer learner corpora can form an authoritative basis for obtaining further insights into the interlanguage systems of language learners. Publishing houses also realize the vital role that learner corpora play on designing EFL tools, which can be improved \"with the NS (native speaker) data giving information about what is typical in English, and the NNS (non-native speaker) data highlighting what is difficult for learners in general and for specific groups of learners\" [Granger, 1998a] However, it is difficult to create learner corpora on the huge scale of native corpora mainly because each collection is usually confined to classroom language.",
+                "cite_spans": [
+                    {
+                        "start": 1028,
+                        "end": 1044,
+                        "text": "[Granger, 1998a]",
+                        "ref_id": "BIBREF7"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In 1993 the International Corpus of Learner English (ICLE) was launched [Granger, 1993] through academic collaboration worldwide. At present, the corpus contains 14 different national varieties, some of which are subdivided regionally, and each subcorpus contains 200,000 words. A great deal of comparative research has been done based on the ICLE, providing statistics-based interpretation of the learners' lexicon, grammar, and discourse [Granger, 1998b] . Another learner corpus, and probably the largest corpus of single group learners so far, is the Hong Kong University of Science and Technology Learner Corpus [Milton & Tong, 1991] , which consists of five million words of written English from Cantonese learners. This corpus is intended to be used for the development of English teaching materials in Hong Kong. SLA scholars in Japan soon followed the trend, and several learner corpus projects were launched, such as the JEFLL corpus of around 200,000 words from Japanese EFL learners' written data, the SST Corpus of 1 million spoken words of learners, and the CEJL Corpus of junior high school to university students. In China, Chinese Middle School Students' Written English and Chinese Middle School Students' Spoken English are two learner corpora forming the Corpus of Middle School English Education that was compiled at South China Normal University beginning in 1998. Apart from academic circles, publishing houses such as Longman and Cambridge University Press have also compiled their own learner corpora for the development of their own language related publications. While many countries around the world have been creating their own learner corpora, little work has been done in Taiwan. The Soochow Colber Student Corpus [Bernath, 1998] , which was compiled between 1984 and 1995 at Soochow University, can be viewed as a pioneering Taiwanese corpus of learner English. It contains around 227,000 words of written text from junior and senior students of Soochow University and National Taiwan University. No other corpus of comparable size was compiled until 1999 when a one-million-word learner corpus project, the Taiwanese Learner Corpus of English, was launched at Sun Yat-sen University. This corpus is a collection of written data from college students majoring in English at the university. The data has been annotated for various linguistic features using the TOSCA-ICLE tagger/lemmatizer [Aarts, Barkema, & Oostdijk, 1997] , assigning to each word its lemma and a tag of its morphological, syntactic and semantic information. With the permission of the compiler of the Soochow Colber Student Corpus to incorporate 85% of its contents, consisting of written data from students majoring in English, the scale of the TLCE has increased from its original 530,000 to 730,000 words. The corpus continues to grow in size. Currently, the TLCE is probably by far the largest annotated learner corpus of English in Taiwan. In the following sections, a complete description of the TLCE will be given, including its purpose, design criteria, method of data capture and documentation, corpus structure and grammatical annotations. The quantitative characteristics of the TLCE as well as its pedagogical application will be depicted and illustrated at the end of the paper.",
+                "cite_spans": [
+                    {
+                        "start": 72,
+                        "end": 87,
+                        "text": "[Granger, 1993]",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 440,
+                        "end": 456,
+                        "text": "[Granger, 1998b]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 579,
+                        "end": 638,
+                        "text": "Science and Technology Learner Corpus [Milton & Tong, 1991]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 1745,
+                        "end": 1760,
+                        "text": "[Bernath, 1998]",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 2421,
+                        "end": 2455,
+                        "text": "[Aarts, Barkema, & Oostdijk, 1997]",
+                        "ref_id": "BIBREF0"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Compiling Taiwanese Learner Corpus of English 89",
+                "sec_num": null
+            },
+            {
+                "text": "The history of the computer learner corpus is less than a decade old, but it has been widely considered as \"a useful resource for anyone wanting to find out how people learn languages and how they can be helped to learn them better\" [Leech, 1998] . Learner output is indeed hard data that SLA scholars can utilize to depict learners' interlanguage systems. The TLCE has been compiled in the hope that it will become a useful resource for SLA scholars who want to understand the internal learning process of Taiwanese learners of English, and in the hope that with corpus-based research findings, EFL teachers will be able to tailor their teaching to students' needs.",
+                "cite_spans": [
+                    {
+                        "start": 233,
+                        "end": 246,
+                        "text": "[Leech, 1998]",
+                        "ref_id": "BIBREF11"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Purpose",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "It is important to have clear design criteria when compiling a learner corpus because of the heterogeneous nature of learners and learning situations. Clear criteria help make it possible to interpret research results correctly and help justify the results of comparative studies on different corpora. Table 1 shows the design criteria of the TLCE. The subjects who have contributed data to the corpus are students majoring in English at the three universities, ranging from freshmen to seniors (aged 19 to 22). Their English proficiency varies from intermediate to advanced levels. The TLCE includes written production of two different genres, namely, informal writings and essay writings. Informal writings consist of daily or weekly journals, which the learners are encouraged to keep during their writing courses, and essay writings are the compositions they are asked to submit regularly for their courses. The types of compositions are mainly descriptive, narrative, expository and argumentative. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 302,
+                        "end": 309,
+                        "text": "Table 1",
+                        "ref_id": "TABREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Corpus Design Criteria",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "The data of the TLCE are in three forms: electronic files, printouts and handwritten texts. More than half of the collection has been submitted through e-mail, which is the easiest way of gathering data for the corpus. E-mail or Microsoft Word files are converted into text files. Another source of data, learners' printouts, have been scanned and transformed into a machine readable format. Post-editing of the scanned data is",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Data capture and documentation",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "Compiling Taiwanese Learner Corpus of English 91 necessary to remove scanning errors. The most time-consuming task is the collection of handwritten texts; all the data have to be keyboarded. As the issue of spelling errors is not a concern in the project and errors would hinder part-of-speech tagging in the subsequent annotation work, all the data in the corpus are spellchecked.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Data capture and documentation",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "The documentation of each piece of writing is needed for researchers to create their own subcorpora according to selection based on pre-defined attributes, and to carry out different comparison studies. For this reason, details about attributes are recorded as an SGML file header for each text. The information includes the university where the learner is studying, the academic year in which the text is collected, the school year (proficiency level) of the learner, and the genre of the text. For instance, the header <#nsysu-891-f-DES> indicates that the text is a descriptive composition written by a freshman at Sun Yat-sen University in the first semester of the 1989 academic year.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Data capture and documentation",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "As stated in Section 2.2, journals and compositions are the two genres of writing collected in the corpus. Journals are informal writings from students, recording what concerns them the most during a day or a week. The journals are sent to their teachers through e-mail systems. Compositions are the essay writings based mainly on different writing strategies: description, narration, exposition and argumentation. The first two are often taught in the first year at universities, whereas the expository and argumentative types are practiced in the second and the third years. Table 2 illustrates the structure of the corpus, including the total numbers of texts and words, and the percentage of the corpus each genre represents. As indicated in the table, the ratio of journals to compositions in the corpus stands at around 3 to 7. Expository and argumentative types of writings are most numerous, making up more than 46% of the whole corpus. Data classified as others came originally from the Soochow Colber Student Corpus with type labels that did not fit into the TLCE categories. For instance, they are labeled as autobiographical writings, letters, imaginative writings or creative writings.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 577,
+                        "end": 584,
+                        "text": "Table 2",
+                        "ref_id": "TABREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Corpus structure",
+                "sec_num": "2.4"
+            },
+            {
+                "text": "Computer corpora are either raw corpora or annotated corpora. Raw corpora simply contain plain text, whereas annotated corpora have extra encoded features obtained through part-of-speech tagging or syntactic parsing. Part-of-speech tagging is a process of attaching a category and probably other attributes to each word, whereas syntactic parsing provides the structural analysis of each sentence. The former is usually done automatically by rule-based, probabilistic or mixed taggers, and the average tagging accuracy is about 95%; the latter can be done by automatic full/partial parsers outputting one or more syntactic structures for a sentence.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Grammatical Annotation",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "The text in the TLCE is currently part-of-speech tagged using the TOSCA-ICLE tagger [Aarts et al., 1997] . TOSCA-ICLE is a stochastic tagger, supplemented with a rule-based component, which tries to correct observed systematic errors of the statistical components. Each word is given its lemma, and a part-of-speech tag, which consists of a major wordclass label, followed by attributes for subclasses and for its morphological information. There are 17 major word classes in the tag set (see Appendix A) and a total of 270 different attribute combinations.",
+                "cite_spans": [
+                    {
+                        "start": 84,
+                        "end": 104,
+                        "text": "[Aarts et al., 1997]",
+                        "ref_id": "BIBREF0"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Grammatical Annotation",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "A major advantage of the corpus approach lies in the usefulness for conducting quantitative analysis. The quantitative features of a corpus provide a basic but global view of the characteristics of the learners' writings. The following findings depict the characteristics of the TLCE as a learner corpus. Figure 1 shows the part-of-speech distribution of the corpus. The graph only indicates those parts of speech individually making up at least 5% of the total corpus. As can be seen, Nouns (N) and verbs (VB) exist in similar proportions in the corpus. Pronouns (PRON) are third, followed by prepositions (PREP), adverbs (ADV), adjectives (ADJ), articles (ART) and conjunctions (CONJUNC). Note that the words in nominal form (N or PRON) make up nearly one third of the whole corpus. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 305,
+                        "end": 313,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Quantitative Analysis",
+                "sec_num": "3."
+            },
+            {
+                "text": "For open classes, N, VB, ADV, and ADJ, it is desirable to know their type/token ratios. The type-token ratio, also called the lexical density, is often used as a measure of the lexical complexity of a text. Here, it is used as the measure of the word versatility of an open class. It is the ratio of different words to the total number of words in the class and is calculated by the formula Although N and VB have similar distributions as shown in Figure 1 ,their lexical densities show great discrepancy. As can be seen in Figure 2 , the lexical density of N is four times higher than that of VB. This phenomenon is also found in the pair consisting of ADJ and ADV, where ADJ has a much higher density value than ADV. In other words, although the frequency counts of VB and ADV in the learner corpus are similar to those of N and ADJ, respectively, the variety of actual words used in the categories of VB and ADV is much more limited than in the N and ADJ categories. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 448,
+                        "end": 456,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 524,
+                        "end": 532,
+                        "text": "Figure 2",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Type/Token Ratio (Lexical Density)",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "A POS trigram is a pattern of three adjacent POSs. It reveals to a certain extent the habitual use of syntactic structures by language learners. The corpus has a total of 777,096 trigrams from 2202 different patterns. Hence, the type-token ratio of POS trigrams is as low as 2.8. Table 3 shows the distribution of the front rank trigram patterns according to frequency of use. As can be seen, the first 50 patterns make up a large proportion of use in the distribution diagram. In fact, it is calculated that the top 220 ranking patterns make up 82% of the trigrams. In other words, learners use only 10% of the POS trigram patterns in 80% of their writings. These figures demonstrate the serious lack of structural variations in learners' writings. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 280,
+                        "end": 287,
+                        "text": "Table 3",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Part-of-speech trigrams",
+                "sec_num": "3.3"
+            },
+            {
+                "text": "The main purpose of compiling the TLCE is to provide Taiwanese researchers in the SLA and EFL areas with a large quantity of authentic learner data, which can be used to conduct qualitative analysis based on quantitative information. With the availability of this useful resource, they can utilize advanced corpus analysis tools to systematically uncover the features of non-nativeness existing in learner English. The findings will enable EFL teachers to focus on areas where remedial work is needed. In this section, a pedagogical application of the TLCE is demonstrated through an investigation of learners' lack of adverbial collocation knowledge from both overuse and underuse perspectives. A series of experiments were carried out based on a contrastive approach, comparing learner English (from the TLCE) with native English (from a one-million-word subset of the BNC).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pedagogical Application",
+                "sec_num": "4."
+            },
+            {
+                "text": "A frequency list of adverbs with the \"-ly\" suffix was obtained from each corpus, and their top 10 adverbs were taken into consideration. The left column of Table 3 shows the top 10 adverbs used by the learners, and the right column shows those used by the native speakers. The bracketed number following an adverb indicates the adverb's rank in the other corpus. Table 3 . Top 10 adverbs in the two corpora.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 156,
+                        "end": 163,
+                        "text": "Table 3",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 363,
+                        "end": 370,
+                        "text": "Table 3",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Top 10 adverbs in the BNC and the TLCE",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "Top N TLCE(learner) BNC (native) 5110 quickly 14especially 4As can be seen in the table, four of the top 10 adverbs in the TLCE appear in the BNC list, namely, really, usually, especially and actually. The rest fall into BNC's top 20 group except for deeply, whose counterpart is ranked 60. This implies that deeply is very overused by Taiwanese learners. By contrast, particularly, with the third highest rank in the BNC list, is the one least used by the learners. Sections 4.2 and 4.3 provide a closer examination of these two phenomena, respectively, based on the contexts in which they appear.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Top 10 adverbs in the BNC and the TLCE",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "----------------------------------------------------------------------",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Top 10 adverbs in the BNC and the TLCE",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "This experiment examined the context in which deeply appears in the TLCE. The adverb can be used to intensify adjectives or verbs. According to the estimation of Mutual Information, the top 10 adjectives or verbs that highly collocate with deeply those listed in the left column of Table 4 .",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 282,
+                        "end": 289,
+                        "text": "Table 4",
+                        "ref_id": "TABREF4"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Overuse phenomenon",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "The middle and right columns show the adverbs (including deeply) which are used by the learners and native speakers, respectively, to intensify words. They are listed in the descending order of their joint frequencies with the corresponding words. As can be seen, deeply seems to be chosen most often when learners wish to use an adverb to modify these words, whereas in the BNC, the native speakers use other synonyms (words in bold type) more frequently than deeply to intensify the same set of words. Extremely distressed, strongly/greatly influenced, greatly impressed, strongly/greatly attracted, firmly convinced and extremely confused are collocations that do not exist in the TLCE. This finding suggests that instead of the Compiling Taiwanese Learner Corpus of English 97 monotonous use of deeply, Taiwanese learners should be made aware of native speakers' strong preference for the above collocations. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Overuse phenomenon",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "To understand the learners' use of particularly, its concordancing lists from the corpora were investigated. There are only 4 instances of the adverb in the TLCE, whereas in the BNC, there are 217 examples. Following is the complete TLCE list and a selected sample of the BNC list:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Underuse phenomenon",
+                "sec_num": "4.3"
+            },
+            {
+                "text": "TLCE concordancing list First of all, the government, <particularly> the Ministry of Administration, self-defense, the teachers, <particularly> the elementary school teachers, is still applied universally , <particularly> in cram schools for high schools take your words seriously, <particularly> in foreign countries. They might BNC concordancing list (selected) ncy food aid in 1990. 'We're <particularly> concerned about the situation in may have been linked with a <particularly> violent six-week strike by rail n French international thinking <particularly> over France's role as the motor ront and other radical groups, <particularly> among the rapidly expanding he past six months, and many, <particularly> the US, are expected to argue st and provide grants for artists, <particularly> students, in the region. Thr strial and social development, <particularly> after Renault was nationalised in hey still have a very useful role, <particularly> when it is the function of t the landscape study shown here. 1 <particularly> liked the rounds for their v hot poker. These colours work <particularly> well in late summer and early",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Underuse phenomenon",
+                "sec_num": "4.3"
+            },
+            {
+                "text": "As can be seen in the TLCE concordancing list, there are only two different functions of particularly in the learners' writings: it is used to modify either a noun phrase or a preposition phrase. However, there are more functions of the adverb in the native speakers' writings. Apart from noun and prepositional phrases, the native speakers also use it to intensify clauses, verb phrases, adjectives and even adverbs. Table 5 shows the percentage of each of the grammatical functions used in each corpus. The findings are two fold. First, the learners seem to possess limited knowledge of particularly's grammatical behaviours. Only two out of the six functions are actually found in the TLCE. Second, the learners are not clear about the possible uses of particularly. Its collocation with adjectives makes up 42% of the BNC examples, the highest among all, but yet it is not used in this way by the learners at all. The above findings suggest that learners should be informed of the grammatical function of the adverb during the learning process. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 418,
+                        "end": 425,
+                        "text": "Table 5",
+                        "ref_id": "TABREF5"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Underuse phenomenon",
+                "sec_num": "4.3"
+            },
+            {
+                "text": "This is the first large-scale tagged Taiwanese learner corpus of English. Preliminary results show several interesting characteristics of the learner corpus in terms of its part-of-speech distribution, the lexical density of its main categories, and the distribution of its trigram structures. An example of pedagogical application has been used to illustrate the usefulness of the corpus. These efforts have been made in the hope that scholars in language education and research will benefit from this pioneer learner corpus, which will be made available soon on website with software tailored to facilitate corpus analysis.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary and Outlook",
+                "sec_num": "5."
+            }
+        ],
+        "back_matter": [
+            {
+                "text": "Financial support from the National Science Council of the Republic of China of this work under contract No. NSC 89-2411-H-110-024 is gratefully acknowledged. I would also like to express my gratitude to Colman Bernath, the compiler of the Soochow Colber Student Corpus, for allowing his corpus to be incorporated into the TLCE. I am also greatly indebted to the following colleagues for helping me collect data: Dr. Shu-ing Shyu, Dr.Ching-yuan Tsai, Dr. Shu-li Chang, Dr. Shu-Fang Lai, Dr.Yuan-jung Cheng, Hue-jen Wen, Alex K.T. Chung, Chu-jen Loh, Dr. Ting-yao Luo at Sun Yat-sen University and Dr. Zhao-ming Gao at National Taiwan University.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Acknowledgements",
+                "sec_num": null
+            },
+            {
+                "text": "Appendix A: part of speech set of TOSCA Tagger ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "annex",
+                "sec_num": null
+            }
+        ],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "The TOSCA-ICLE Tagset",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Aarts",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Barkema",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "N",
+                        "middle": [],
+                        "last": "Oostdijk",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Aarts, J., Barkema, H., & Oostdijk, N. 1997. The TOSCA-ICLE Tagset . Nijmegen: University of Nijmegen, The Netherlands.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Soochow Colber Student Corpus",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Bernath",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Bernath, C. 1998. Soochow Colber Student Corpus. Available: ftp://ftp.scu.edu/tw/scu/english/colber.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Frequency Analysis of English Usage:Lexicon and Grammar",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [],
+                        "last": "Francis",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Kucera",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1982,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Francis, W., & Kucera, H. 1982. Frequency Analysis of English Usage:Lexicon and Grammar. Boston: Houghton Mifflin.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Manual of Information to accompany ' a Standard Sample of Resent-Day Edited American English, for Use with Digital Computers",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [
+                            "N"
+                        ],
+                        "last": "Francis",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Kucera",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1964,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Francis, W. N., & Kucera, H. 1964. Manual of Information to accompany ' a Standard Sample of Resent-Day Edited American English, for Use with Digital Computers' Department of Linguistics, Brown University.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "English language Corpora: Design, Analysis and Exploitation",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Granger",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "57--69",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Granger, S. 1993. The International Corpus of Learner English. In J. Aarts, P. d. Haan, & N. Oostdijk (Eds.), English language Corpora: Design, Analysis and Exploitation. pp. 57-69 Armsterdam: Rodopi.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "The computer learner corpus: a versatile new source of data for SLA research",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Granger",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Learner English on Computer",
+                "volume": "",
+                "issue": "",
+                "pages": "3--18",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Granger, S. 1998a. The computer learner corpus: a versatile new source of data for SLA research. In S. Granger (Ed.), Learner English on Computer. pp. 3-18 London and New York: Longman.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Comparing English Worldwise: The Interational Corpus of English",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Greenbaum",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Greenbaum, S. (Ed.). 1996. Comparing English Worldwise: The Interational Corpus of English. Oxford: Clarendon Press.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "The subjunctive in British and American English",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Johansson",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "E",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Norheim",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1988,
+                "venue": "ICAME Journal",
+                "volume": "",
+                "issue": "12",
+                "pages": "27--36",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Johansson, S., & Norheim, E. H. 1988. The subjunctive in British and American English. ICAME Journal (12), 27-36.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Preface",
+                "authors": [
+                    {
+                        "first": "G",
+                        "middle": [],
+                        "last": "Leech",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Learner English on Computer",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Leech, G. 1998. Preface. In S. Granger (Ed.), Learner English on Computer . New York: Longman.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "Text Analysis in Computer Assisted Language Learning",
+                "authors": [],
+                "year": 1991,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Milton, J., & Tong, K. (Eds.). 1991. Text Analysis in Computer Assisted Language Learning. Hong Kong: Hong Kong University of Science and Technology.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Quantitative comparison of modals in the Brown and LOB corpora",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Nakamura",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "ICAME",
+                "volume": "",
+                "issue": "17",
+                "pages": "29--48",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Nakamura, J. 1993. Quantitative comparison of modals in the Brown and LOB corpora. ICAME (17), 29-48.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF1": {
+                "type_str": "figure",
+                "text": "Figure 1. POS Distribution",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF2": {
+                "type_str": "figure",
+                "text": "Figure 2. Lexical Density",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF3": {
+                "type_str": "figure",
+                "text": "Figure 3. Trigram Distribution",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF4": {
+                "type_str": "figure",
+                "text": "",
+                "num": null,
+                "uris": null
+            },
+            "TABREF0": {
+                "html": null,
+                "content": "<table><tr><td/><td>attributes</td></tr><tr><td>age</td><td>19-22</td></tr><tr><td>level</td><td>Intermediate to advanced</td></tr><tr><td>Mother tongue</td><td>Chinese</td></tr><tr><td>Learning context</td><td>EFL</td></tr><tr><td>medium</td><td>Written text</td></tr><tr><td>genre</td><td>journals and compositions</td></tr></table>",
+                "num": null,
+                "type_str": "table",
+                "text": "TLCE Design Criteria"
+            },
+            "TABREF1": {
+                "html": null,
+                "content": "<table><tr><td>823</td><td>213091</td><td>29.4</td></tr><tr><td>composition</td><td/><td/></tr><tr><td>Description/narration</td><td/><td/></tr><tr><td>(first year)</td><td/><td/></tr><tr><td>Exposition/argumentation (second/third years) others</td><td/><td/></tr></table>",
+                "num": null,
+                "type_str": "table",
+                "text": ""
+            },
+            "TABREF4": {
+                "html": null,
+                "content": "<table><tr><td>Intensified words</td><td>Adverbs in TLCE</td><td>Adverbs (Synonyms) in BNC</td></tr><tr><td>Distressed</td><td>Deeply</td><td>extremely, deeply, \u2026</td></tr><tr><td>Influenced</td><td>deeply, directly, rapidly</td><td>strongly, greatly, deeply,\u2026</td></tr><tr><td>Moved</td><td>deeply, really, suddenly, \u2026</td><td>deeply,\u2026</td></tr><tr><td>impressed</td><td>deeply, especially, really</td><td>greatly, deeply,\u2026</td></tr><tr><td>attracted</td><td>deeply, really, fully</td><td>strongly, greatly, deeply,\u2026</td></tr><tr><td>convinced</td><td>deeply, obsessively</td><td>firmly, deeply,\u2026</td></tr><tr><td>touched</td><td>really, deeply</td><td>deeply,\u2026</td></tr><tr><td>concerned</td><td>deeply, obsessively</td><td>deeply,\u2026</td></tr><tr><td>confused</td><td>deeply, really</td><td>extremely, deeply,\u2026</td></tr><tr><td>interested</td><td>really, deeply, keenly</td><td>deeply,\u2026</td></tr></table>",
+                "num": null,
+                "type_str": "table",
+                "text": "Adverb Alternatives"
+            },
+            "TABREF5": {
+                "html": null,
+                "content": "<table><tr><td>Grammatical collocation</td><td>BNC(%)</td><td>TLCE(%)</td></tr><tr><td>ADJECTIVE</td><td>42</td><td>-</td></tr><tr><td>PREPOSITION PHRASE</td><td>28</td><td>50</td></tr><tr><td>NOUN PHRASE</td><td>15</td><td>50</td></tr><tr><td>CLAUSE</td><td>7</td><td>-</td></tr><tr><td>VERB</td><td>6</td><td>-</td></tr><tr><td>ADVERB</td><td>2</td><td>-</td></tr></table>",
+                "num": null,
+                "type_str": "table",
+                "text": "Distribution of Grammatical Collocations of \"particularly\""
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1001.json

@@ -0,0 +1,577 @@
+{
+    "paper_id": "O01-1001",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:11.876645Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O01-1001",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "/0h!B?(/0$l4!(\"#&'( 1 \u00e2\u00aa\u00ab2$\u00cf(\"#%N1U(\"#%N!_g \u00d3= 0!2 Q &' M\u00f1!\u00cbg\u00a9!N\u00d4R$l4X\u00cf!> bc\"#'!\u00cf\u00d089G\u00bdZ\u00d5()1 XN\u00d4R!\u00e0A_`23\u00d3l|}\u00b8\u00b9\u00a49! $7 \u00af!(/0\u00a4(/0Y\u00d6AbNr\u00d9!/0$fN\u00d4RU\u00a4Y 7(/0!_g\u00da\u00d7e K$ \u00a4|1$bc * X Y!/0 23N5 X Y\u00a4G\u00bd!E(F X Y! \u00d8/0 !() g Y/0 !o\u00d5/\u00d91U\u00cb23\u00e2E(FZ\u00d5!/0 \u00d2 !(/01 \u00d7 = + = $ \u00b6 YK/0 !R\u00d9 \u00b6/0 X\u00aa\u00ab2Y! \u00b6\u00aa\u00ab2Y\u00a4\u00bd\u00db1 U\u00cbB?N\u00d4R$(\"#!%N123$B?N\u00d4R!()r \u00cf'\u00da!/0\u00d5U\u00f5bc(!!&3ZP!N\u00d4R1 23\u00dbN!bc \u00dan/0 \u00baG !N\u00d4RI\u00e2D\u00bdE(F 1pC@\"C\u00cfC \u00a4D\u00bd!E(F CCE(F!_g\u00da\u00b1\u00d7e 7 \u00a4|1 \u2211 \u2229 \u2208 7 \u00b6 X Y!/01 I23\u00a6Y\u00b0bcC!E(FpwC!E(F\u00d5\u00e4\u00d7\u00a6sY (\"#!\u00dc123\u00aa\u00ab2\u00cf!\u00a4\u00bd\u00dd!\u00d7\u00de\u00b0b Y(C!\"#\u00df1op23l\u00b0c\"#X\u00cf\u00a4D\u00bd!()\u00dc 1U\u00a9Kmn@A(\"#\u00a4X!CH1qr\u00dd23\u00e2K _g!L}TG\u00e0 1 \u00e0 \u00b6(\"#_g!\u00a6Cg1 +,*- -.89Q.G\u00bdZ\u00d5!HN}~23FY!C%3ZP-./ 0IKCQ.&G\u00bdZ\u00d5!HN}~JK@A\u00ed!\u00cf\u00d01\u00b0: -.\u00cf\u00d0JLGu\u00dfs23g\u00a4Y\u00a3s\u00e1\u00e2!|}1U\u00f5XbY \u00ed\u00bb&\u00fa\u00b1!/0$B\u00b1b\\\u00e3'\u00edB\u00be!\u00e4\u00ed\u00e5ae\u00aaST=/0! P\u00be) \u00b6s23X(/0rL}YrtH\u00b1/0!\u00e7w\u00f4\u00e3\u00e41 GB\u00be! A23F\u00ed\u00aa\u00ab\u00bbP\u00be)!/0$B\u00b1b.\u00fa LBdH\u00e8\u00e9\"!/0X!G\u00bd\u00d5~HN51JK23\u00cc\u00cd7 \u00ed\u00bd\u00bb!HN/01 =7Q-.\u00cf\u00d0GHN!/023 !\u00ae\u00ea \u00a9N)\u00eb-7bc/0!\u00b15/Q5/1DE23FX\u00ff\u00ec/0/5! N5Y\u00a4t!/0!&QK/0G\u00bdbN!HN}~1JK23\u00a9 ./\"#0/12342 \u00aa\u00abv x\u00edb \u00b6\u00a9\u00d7e Q $7\u00aa\u00ab2\u00cf!(/01 x\u00ed+ \u00b6\u00e2!(/0Q-.Rmr!N\u00d4R1 x\u00ed. \u00b6\u00a9\u00d7e 7C\u00cfC!E(FhwE(F$N \u00cf\"#!()\u00dc1 x\u00ed9 \u00b6x\u00ed.!|}lQ4NXK\u00aa\u00ab2Y (\"#!\u00a4X1 \u00acv x\u00edb \u00b6 @A(\"#\u00a4X!C$H1 $+,-./0L Q-./0ZHN!\u00b15/K5/QN5\u00cf !/0h\u00e2B?/0\u00d2-./0$\u00ff\u00cf\u00d0P\u00be)/0\u00a4\u00ac$!\u00ee\u00ef1 X \u00cf\u00bd?b/P5!/0&D\u00bdPc/51JKX+,/0\u00d5 23\u00f0\u00f1\u00de/0!\u00fev/5\u00b3\u00d7:\u00c923\u00a5\u00ee\u00ff/5\u00f2e!\u00b5.1 =c_g$-./0/5bc: !N5Q/0 !(\u00f3)$/5\u00f2ebcQ\u00a928/5 \u00e7w/5\u00f2ebc1 5 6789:; Bc_g:$X-.Ybc/0%G\u00bdR/5K/5\u00a4\u00f4$!HN/0 XY !\u00e7w!&\u00d5LG'`/5!\u00f4/01G:23-.\u00cf!J /K/QH\u00d0/ $7\u00b0c/0X\u00fa\u00b1/5\u00a4D\u00bd!\u00b15/Q 5/*/5XN5\u00cf\u00a4!/0\u00e2. bc/0 r1U\u00cb \u00e2 r\u00cf!/0Q\u00cf\u00d0r23\u00f5 (\u00f6)o\u00d5! r\u00a4\"B!/5 NK/0!/51\u00e0 $ B|K_gG/0\u00a4H! r\u00e0Y!/0\u00bd / 5UB /5\u00bd!\u00b15/K5/QN5\u00cf!/0JK&\u00bd c/0 r1 I23\u00a9 -./0\u00ed/5bc\u00b2bc/0!\u00a4\u00bd/0 r\u00bd \u00bbX\u00cf \u00bcb\u00besg\u00cb P\u00f7$%&\u00fev/51K\u00b2 23\u00a8\u00f8(*+/5bc_g\u00a928/5 \u00e7w/5\u00f2ebc1 <= >?@A6789:; *+c/5bc_gI:\u00db\u00f9bc/0X28Yo9 !/51\u00a4! 2:-4/5!6728-+ 142X 7$ \u00c8YK28\u00bd\u00c4p !\u00c91K28\u00cf!/0= $/5-423B?/0 l/0X\u00fa\u00b1/5 !\u00e7w1JKXbc/5!L}Y23TU\u00fa\u00e0 \u00d9wo\u00d5!/5$\u00edBc/0 X-.\u00b2\u00a4G\u00bd!/51B\u00be! _e23\u00a8W bc/5!\u00fa\u00e1Ua\u00d5\u00fbbcw1 \u00a9\u00dd_g/5bcp\u00d7-./0!/5$+,'HN! /01/0+,!mn\u00d7:\u00e2-./0!5/K\u00b15/QN5\u00cf!H\u00b1/)/0 @A$HK-./0!HNR123\u00e2-./0Q+,p!/0\u00d2(/0 U\u00cbC\u00b0cC\u00a4D\u00bd!(/0\u00edC!\u00abp\u00faZ\u00d5! C$\u00ed1+,-./0_g!\u00a6Cg23G\u00e0 71 \u00e0 7 \u00b6+,-./0_g!\u00a6Cg1 +,*-0/12342 x\u00edb \u00b6 !-./0 7'/5!/0 r1 x\u00ed+ \u00b6\u00e2/0 !\u00a4\u00bd/0 rQ \u00cf!/0rC r! 1 x\u00ed. \u00b6C/0!o\u00d5\u00d9/5Q\u00d5\u00d9/5r\u00d9F % \u00f4GR\u00fc\u00eeN !\u00fd\u00feF IN/0 !/5 o\u00d5! r\u00a4\"B!/5q\u00d1 ITU\u00db\u00f928Y \u00d9wo\u00d5!/51 x\u00ed9 \u00b6\u00a9bc!/5$+,-./0\u00ff/5!\u00b15RK5RQN 5\u00cf\u00b1/)!/01 x\u00ed: \u00b6mr-./0Q+,p!/0\u00ed!(/01 x\u00ed) \u00b6C \u00cfC \u00a4D\u00bd!(/0\u00fa\u00e0D\u00bdZP(/0! C\u00ed\u00cf\u00d01 BC12 !]^:bHI\u00f5\u00ca!\u00ec\u00ed1b\u00caUy]^ !",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u00bdO\u00d8!uvw1J!Bcfg23\u00a5bcdkl$VW23!] \u00ec\u00ed1 < * \u00b3 1 $%%%\u00a4a!d_gY\u00fa\u00e0='Ybc_g$ \u00c8ijdkl1\u00ee23\u00e22Y!/0\u00c8v\u00c8\u00e4Rhg R\u00f5B?$!/0\u00f9#H!!3op\u00a4s!/0\u00d7:S R21U\u00cb23\u00aa\u00ab2Y\u00b0cd \u00a4D\u00bd!0\u00db'SR $l\u00db\u00f9`3!\u00bd!E(F UCE(F!_e\u00da\u00d7e 9 \u00a4|1 \u00d7 = 9 /3 = \u00b6SR Xd !\u00d9w \u00b6d!\u00db \u00b6SR L!d\u00db1 \u00db\u00f9b\u00ac2! 23XN\u00a4\u00ba!d\u00b2\u00f0\u00f1\u00cf\u00a43! dSRE(F!T\u00dbstX\u00fa\u00b1d!\u00dbE(F 1U \u00cbld!\u00dbE(F\u00e2^\u00baG\u00dbE(Fo\u00d5!d C_e \u00da\u00d7e \u00a4|1 \u2211 \u2208 \u00d7 = !\" 3 = \u00b6SR X !\u00d9w \u00b6d !SR r1 \u00dd!_g23ij=bcdkl123\u00e2>\u00bfL! \u00ac2dklstuvw !d n23Kuvw$I\u00ed xyuvw$f rZ!xy1 \\\u00a4a!_g n23!2m 7[ \\|2=bk![\\$\u00ca_g! 1\u00ee23\u00ee/0|2!\u00d5 #1U\u00cb\u00e2/0[\\_g3[\\Q[\\mnMop [\\mn6=1 D#EF X\u00d523G[\\|2!\u00d5#\u00bd\u00b5cx\u00ed \u00b6\u00ee\u00f0\u00f1 gR4+\u00ae\u00faG(5!/01U\u00cb:#H!Bx\u00ed Pae/0G\u00bd\u00fa\u00b1/)!1\u00da /K/K/'B.c/0\u00aa G\u00bdHa!5\u00b0Xl&\u00d2\u00fa\u00b1!/0qU#H\u00bb\u00e2Bd!/0 \u00a7 Z !R\u00da /JK0Kd/0XY\u00a4;\u00bd!()1 !\"#$%&'()* 23\u00d3X 9 \u00c8Ya!_g$4!(\"#\u00a4X1\u00a4\u00bd\u00aa\u00ab2Y !\u00fa:c(Cw\"#!()Q \u00d9w$U 4(\"#\u00a4X1DEG2!",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "+,*-",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "\u00ecV*! ;(;;4E",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "\u00ecV*! ;(;;4E (323 ;",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "23=X 9$ Q 9$$ \u00c8\u00a4 a!c_g$/5bc1hbc!/5$-./0+",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "U\u00eb-.\u00cf!j/K/Qh\u00f0",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "U\u00cb-.\u00cf!J/K/QH\u00d0/23=X 9$ Q 9$$ \u00c8\u00a4 a!c_g$/5bc1hbc!/5$-./0+,_ ,Q-./0HN! \u00b15/K5/QN5\u00cf\u00d01",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "GHIJ \u00ee23\u00a9 9$ \u00c8\u00a4a!*bc_g$-./0bc/5123\u00e2B?",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "GHIJ \u00ee23\u00a9 9$ \u00c8\u00a4a!*bc_g$-./0bc/5123\u00e2B?-",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "0Q rp\u00db(\u00bd 9%% P c/0 G Y1sG?\u00fa3X \u00cf!/023\u00bfL\u00ca",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/QH\u00d0/!/0123\u00a9B?/0Q rp\u00db(\u00bd 9%% P c/0 G Y1sG?\u00fa3X \u00cf!/023\u00bfL\u00ca",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "IJ MMqJB?/0>",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "'Y\u00bdP:\u00baG\u00ef\u00bdJ/\u00d9\u00da G-)K1H)?IJ MMqJB?/0>",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "G Y23g`3/0+",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "G Y23g`3/0+,1U\u00cb23\u00a4 9%% Pc/0",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "C\u00b0c rY!/0rt!",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C\u00b0c rY!/0rt!1bc-./023\u00e0o\u00d5",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "B!/5\u00edK/0!/51\u00abUIro\u00d5! r\u00a1Lbc!o H23\u00a8g%&K/0^\u00baG>bc/51B\u00b223\u00a8\u00a9*+/5bcg",
+                "authors": [
+                    {
+                        "first": "!",
+                        "middle": [],
+                        "last": "R\u00a4",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "! r\u00a4\"B!/5\u00edK/0!/51\u00abUIro\u00d5! r\u00a1Lbc!o H23\u00a8g%&K/0^\u00baG>bc/51B\u00b223\u00a8\u00a9*+/5bcg",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "/0!/5p23\u00a8w<B?%&!/5$ /0/0+,!\u00ec\u00ed1/0+,!_e:\u00a4 Y\u00e2/5!",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "I\u00fb\u00a2%&-./0!/5p23\u00a8w<B?%&!/5$ /0/0+,!\u00ec\u00ed1/0+,!_e:\u00a4 Y\u00e2/5!",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "/0!\u00ee\u00efX1 `230c\u00d9Cfc\u00df@!\u00c91B :b\u00ffp!_H1($ !R:\u00cfY !-.RU23 !RI:+,p!R123\u00ca** bKQ-./0rp\u00c9\u00bdbc/0 8E XKqUX+",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "\u00b15/K5/QN5\u00cfH\u00b1/)!/0@A$B?+,/0$/#\u00bd! -./0ST-./0!\u00ee\u00efX1 `230c\u00d9Cfc\u00df@!\u00c91B :b\u00ffp!_H1($ !R:\u00cfY !-.RU23 !RI:+,p!R123\u00ca** bKQ-./0rp\u00c9\u00bdbc/0 8E XKqUX+,/0p",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "HN/0s**b \u00a7sQ-.ZHN1",
+                "authors": [
+                    {
+                        "first": "B!#",
+                        "middle": [],
+                        "last": "C\u00ebb",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "B!#C\u00cbB?HN/0s**b \u00a7sQ-.ZHN1",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "\u00d5p!\u00c4\u00d91 \"# \u00cf\u00d0 -. K;>0I;K:L;-I*0;I:;I>0;&G>;HM0",
+                "authors": [
+                    {
+                        "first": "B \u00b6/",
+                        "middle": [],
+                        "last": "0+",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "B \u00b6/0+,\u00d5p!\u00c4\u00d91 \"# \u00cf\u00d0 -. K;>0I;K:L;-I*0;I:;I>0;&G>;HM0(: ? ;",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "23 G/ ) 2L \"#2' 8 '+ 48/ !# 2L #!4 #3",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "3",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "-E / 23 G/ ) 2L \"#2' 8 '+ 48/ !# 2L #!4 #3 #/ 8 2/!3//",
+                "links": null
+            },
+            "BIBREF16": {
+                "ref_id": "b16",
+                "title": "!\u00b15/B#3!#C: !5/B2 #C: 2 !5",
+                "authors": [
+                    {
+                        "first": "\u00a3",
+                        "middle": [],
+                        "last": "\u00b6b#c",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "\u00a3 \u00b6B#C: / !\u00b15/B#3!#C: !5/B2 #C: 2 !5/1",
+                "links": null
+            },
+            "BIBREF18": {
+                "ref_id": "b18",
+                "title": "0Q+,p!/0h4\u00bd X\u00cf\u00afs23XCCE(F\u00b2\u00cf \u00a4\u00bd!CE(F %JUg@A56C$H1B\u00c9=+,-. /0_gXR?g7'\u00fb 1B +,-./0_g![\\ mn%\u00a4\u00bd 77% $Cmnst !duvw1 B \u00b6+",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "0Q+,p!/0h4\u00bd X\u00cf\u00afs23XCCE(F\u00b2\u00cf \u00a4\u00bd!CE(F %JUg@A56C$H1B\u00c9=+,-. /0_gXR?g7'\u00fb 1B +,-./0_g![\\ mn%\u00a4\u00bd 77% $Cmnst !duvw1 B \u00b6+,-./0![\\mn",
+                "links": null
+            },
+            "BIBREF20": {
+                "ref_id": "b20",
+                "title": "3st %7!duvw1JK\u00a4\u00dd![\\mn23\\=+,-./ 0_g\u00a4!'\u00cf\u00d0!v3=\u00cf!(C'\u00bbstZ8",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "3st %7!duvw1JK\u00a4\u00dd![\\mn23\\=+,-./ 0_g\u00a4!'\u00cf\u00d0!v3=\u00cf!(C'\u00bbstZ8",
+                "links": null
+            },
+            "BIBREF21": {
+                "ref_id": "b21",
+                "title": "!d 1OFsba!:X/5bcv23!_g\u00fa\u00a59:!CQ\u00f3\" \u00d7-./07'o\u00bd\u00bb/5$VW+",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "!d 1OFsba!:X/5bcv23!_g\u00fa\u00a59:!CQ\u00f3\" \u00d7-./07'o\u00bd\u00bb/5$VW+,-./0_g!1",
+                "links": null
+            },
+            "BIBREF22": {
+                "ref_id": "b22",
+                "title": "/0_g$@A\u00cf!CH1\u00b0X/ 0rL}Y23 h\u00fa:\u00a4\u00bd!\u00cf\u00d0\u00bc& -./0B\u00be!oHs \u00bd?J\u00fa\u00bbX\u00cf\u00d0Yrt-./0Ug@A56C$JKB !\u00cf\u00d0\u00e2&:!1=",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "X 7 \u00c8Y23\u00a9+,-./0_g$@A\u00cf!CH1\u00b0X/ 0rL}Y23 h\u00fa:\u00a4\u00bd!\u00cf\u00d0\u00bc& -./0B\u00be!oHs \u00bd?J\u00fa\u00bbX\u00cf\u00d0Yrt-./0Ug@A56C$JKB !\u00cf\u00d0\u00e2&:!1=;",
+                "links": null
+            },
+            "BIBREF23": {
+                "ref_id": "b23",
+                "title": "#_g![\\mnN\u00a9K_g$TW1 GHIJ =b",
+                "authors": [
+                    {
+                        "first": "$",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "$ \u00c8(\"#_g![\\mnN\u00a9K_g$TW1 GHIJ =b[\\23\u00ee\u00a9+,-./0_g$\u00a4\u00bd",
+                "links": null
+            },
+            "BIBREF24": {
+                "ref_id": "b24",
+                "title": "#1DE23 [\\mnY\\=\u00b4~%\u00bbt!.b\u00e2a\u00de\u00e1\u00e2!(|}\u00db \u00f91 X+",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "*:![\\mn23 (\"#[\\stU!xyuv w![\\mn\u00c9=\u00cf!vX\u00cbQ.HN!SN(\"#1DE23 [\\mnY\\=\u00b4~%\u00bbt!.b\u00e2a\u00de\u00e1\u00e2!(|}\u00db \u00f91 X+,-./0[\\Y23\u00a9= !S)\u00cb`!\u00ae \u00ea-./0+,ST=-.!\u00ee\u00efX7Q-.ZHN!\u00ed \u00cf\u00d01\u00faLX[\\L}Y23 Bc_g\u00bd'\u00fa\u00e1:G! \u00cf\u00d0\u00fabN\u00bd\u00cb-./0''HN/0!X1JK23a=bqr+,-.",
+                "links": null
+            },
+            "BIBREF25": {
+                "ref_id": "b25",
+                "title": "#!_gmrc_g$VW\u00a4\u00bd!\u00bc",
+                "authors": [
+                    {
+                        "first": "/",
+                        "middle": [],
+                        "last": "0q",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/0Q(\"#!_gmrc_g$VW\u00a4\u00bd!\u00bc\u00bb(!",
+                "links": null
+            },
+            "BIBREF27": {
+                "ref_id": "b27",
+                "title": "M#2! ;2!!.5 G ?#3 7 I2",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "M#2! ;2!!.5 G ?#3 7 I2/ )3",
+                "links": null
+            },
+            "BIBREF30": {
+                "ref_id": "b30",
+                "title": "2#5G?#39G!'/I",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "323",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "2#5G?#39G!'/I(323 44",
+                "links": null
+            },
+            "BIBREF31": {
+                "ref_id": "b31",
+                "title": "!2&?P+)I2",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "!2&?P+)I2",
+                "links": null
+            },
+            "BIBREF32": {
+                "ref_id": "b32",
+                "title": "#/M8>)G>?",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "//82!1<I0/##(\"#/M8>)G>?",
+                "links": null
+            },
+            "BIBREF33": {
+                "ref_id": "b33",
+                "title": "/50//&+( 44 $ $ & ? N ; ? P12# I",
+                "authors": [
+                    {
+                        "first": "G!'",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "G!'/50//&+( 44 $ $ & ? N ; ? P12# I ;",
+                "links": null
+            },
+            "BIBREF34": {
+                "ref_id": "b34",
+                "title": "PI2! >\" ;2!!. ;2!!5 I'# I2!#>",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "&' 0 1 : ( PI2! >\" ;2!!. ;2!!5 I'# I2!#>\";2!!.>)G>?",
+                "links": null
+            },
+            "BIBREF35": {
+                "ref_id": "b35",
+                "title": "PG3 >4# 8 ?5 G ?#3",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "1",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "&' 0 1 : ( PG3 >4# 8 ?5 G ?#3",
+                "links": null
+            },
+            "BIBREF36": {
+                "ref_id": "b36",
+                "title": "#I44/2/(323?#3I",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "3--4",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "#I44/2/(323?#3I(?3M1",
+                "links": null
+            },
+            "BIBREF37": {
+                "ref_id": "b37",
+                "title": "*;NN1PI>\"13.-;2!!.>#A25G ?#3(?GE4I1($%%%$%%%44 7",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "*;NN1PI>\"13.-;2!!.>#A25G ?#3(?GE4I1($%%%$%%%44 7",
+                "links": null
+            },
+            "BIBREF41": {
+                "ref_id": "b41",
+                "title": "I P1#42/ G\"3< I -H3. *+/35",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "2",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "I P1#42/ G\"3< I -H3. *+/35 ;2 (8",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF0": {
+                "num": null,
+                "content": "<table><tr><td colspan=\"2\">cv\u00c4p!|2GB 1</td><td/><td/></tr><tr><td colspan=\"3\">B  \u00b6G-4/5!6728\u00c4p|}1</td><td/></tr><tr><td>J\u00c1</td><td>\u00c2\u00c3c</td><td>/5-4\u00c4\u00c5</td><td>-4/5!/0</td></tr><tr><td>8+</td><td>%7</td><td>J/K/KH\u00d0/QAE/</td><td>%$</td></tr><tr><td>8+$</td><td>7</td><td>J/K/KH\u00d0/QAE/</td><td>9 9</td></tr><tr><td>8+'</td><td/><td>/</td><td>9 $</td></tr><tr><td>&gt;/</td><td>7$</td><td/><td>$799</td></tr><tr><td colspan=\"4\">\u00a4B \u00cf\u00d0\u00cbBc6728(\u00bd 7$ c\u00c2\u00c31J \u00c93J/K</td></tr><tr><td colspan=\"4\">/KH\u00d0/QAE/B9/)JK\u00c9\u00bbB9/)!/0/5-4\u00ec</td></tr><tr><td colspan=\"4\">\u00ed1Xc28Y\u00db(\u00bd $799 c/0-4 /5123\u00e2\u00a9B</td></tr><tr><td colspan=\"4\">c6728\u00a4a\u00de!/0Q/5|}$VW1</td></tr><tr><td colspan=\"4\">X23\u00e2_g\u00c7Q]^v6=1X 9 s 9$ \u00c8Y</td></tr><tr><td colspan=\"4\">23\u00c9X\u00b2\u00a4!_g1DEabcd_</td></tr><tr><td colspan=\"4\">e$]^ !fgh23\u00a4!dklX 97 \u00c8Y\u00c4p!</td></tr><tr><td>\u00c91</td><td/><td/><td/></tr><tr><td colspan=\"3\">!\"#$%&amp;'()*</td><td/></tr><tr><td colspan=\"4\">Xb\u00ca!\u00cf\u00d0!.89&amp;;&lt;\u00cbR?SN!\"#1JK</td></tr><tr><td colspan=\"4\">\u00cc\u00cd\u00bbLPae\"#!()$7!.\u00a4X10!2 F3-.R</td></tr><tr><td colspan=\"4\">!&amp;Q.ZHN0!2 1DEo(!89&amp;</td></tr><tr><td colspan=\"4\">XZ\u00d5`':Zp`!\"#1KE&amp;' M\u00f1\u00b3F.9;&lt;SN\"#</td></tr><tr><td colspan=\"4\">Y3\u00ce\u00cf\u00d0HN!a\u00d1\u00e2-.RQ\u00df/\u00cf!R\u00d2!(</td></tr></table>",
+                "html": null,
+                "text": "4<+++#3#4#2=+ \u00c0\u00a3s 8+ 8+$ 8+' .",
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1002.json

@@ -0,0 +1,574 @@
+{
+    "paper_id": "O01-1002",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:08:51.003920Z"
+    },
+    "title": "2.2 n-gram \u6a21\u578b\u4e4b\u5efa\u7acb",
+    "authors": [
+        {
+            "first": "T",
+            "middle": [],
+            "last": "\u4e26\u4e14\u662f\u7531\u4e00\u6bb5\u8a5e\u5e8f\u5217",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "\u2026w",
+            "middle": [
+                "T"
+            ],
+            "last": "\u6240\u7d44\u6210\uff0c\u5247",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "S",
+            "middle": [],
+            "last": "\u51fa\u73fe\u7684\u6a5f\u7387\u53ef\u4ee5\u5beb\u6210",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O01-1002",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "\u53ef\u80fd\u7684\u5c0d\u61c9\u6587\u53e5(sentence) S * \uff0c\u4f7f\u7528\u8c9d\u5f0f\u5206\u985e\u67b6\u69cb\u662f\u627e\u51fa\u6700\u4f73\u4e8b\u5f8c\u6a5f\u7387\u7684\u6587\u53e5 * argmax ( | ) argmax ( | ) ( ) S S S PS X PX S PS = = \u22c5 (1) \u5176\u4e2d n-gram \u6a21\u578b P(S)\u626e\u6f14\u8457\u4e8b\u524d\u6a5f\u7387\u7684\u89d2\u8272\uff0c\u900f\u904e\u8072\u5b78\u6a21\u578b\u8a08\u7b97\u53ef\u7372\u5f97\u4e00\u6bb5\u6587 \u53e5\u7684\u8072\u5b78\u6a21\u578b\u5206\u6578 P(X|S)\uff0c\u518d\u900f\u904e\u8a9e\u8a00\u6a21\u578b\u8a08\u7b97\u53ef\u7372\u5f97\u6b64\u6587\u53e5\u7684\u8a9e\u8a00\u6a21\u578b\u5206\u6578 P(S)\uff0c\u5c07\u5169\u6a5f\u7387\u76f8\u4e58\u6c42\u5f97\u6700\u4f73\u5316\u4e4b\u6587\u53e5\uff0c\u5373\u70ba\u6b64\u8072\u5b78\u8a0a\u865f\u6700\u6709\u53ef\u80fd\u4e4b\u5c0d\u61c9\u6587\u53e5\u3002 \u5c31\u6587\u4ef6\u5206\u985e\u7684\u9818\u57df\u800c\u8a00\uff0c\u7d66\u5b9a\u4e00\u7bc7\u6587\u4ef6 d\uff0c\u76ee\u6a19\u662f\u53bb\u627e\u5c0b\u6b64\u7bc7\u6587\u4ef6\u6240\u5c6c\u7684\u985e \u5225 c (category)\uff0c\u5047\u8a2d\u6211\u5011\u7e3d\u5171\u5b9a\u7fa9\u4e86 k \u500b\u985e\u5225\uff0c\u4e26\u4e14\u4f7f\u7528\u9019\u4e9b\u985e\u5225\u6240\u5c6c\u7684\u6587\u4ef6\u8a13 \u7df4\u597d\u4e0d\u540c\u985e\u5225\u7684 n-gram \u6a21\u578b L 1 \u3001L 2 \u3001\u2026\u2026\u3001L k \uff0c \u4f7f\u7528\u8c9d\u5f0f\u5206\u985e\u5668\u6c42\u5f97\u6b64\u7bc7\u6587\u4ef6 \u6240\u5c6c\u7684\u985e\u5225 c * \u53ef\u5beb\u6210 (3) \u4f46\u662f\u6b64\u7a2e\u65b9\u6cd5\u5728\u8a08\u6bcf\u4e00\u500b\u8a5e\u7684\u689d\u4ef6\u6a5f\u7387\u6642\u90fd\u8981\u727d\u6d89\u5230\u524d\u9762\u6240\u6709\u7684\u8a5e\u5e8f\u5217\uff0c\u4f7f\u5f97\u8a08 \u7b97\u91cf\u592a\u5927\u800c\u7121\u6cd5\u5be6\u73fe\uff0c\u70ba\u89e3\u6c7a\u9019\u500b\u554f\u984c\u6240\u4ee5\u6709 n-gram \u6a21\u578b\u7684\u7522\u751f\uff0c\u5728 n-gram \u6a21 \u578b\u4e2d\uff0c\u5b83\u662f\u5047\u8a2d\u4e00\u500b\u8a5e\u51fa\u73fe\u7684\u6a5f\u7387\u53ea\u8ddf\u524d\u9762 n-1 \u500b\u8a5e\u6709\u95dc\uff0c\u56e0\u6b64(3)\u5f0f\u53ef\u4ee5\u8fd1\u4f3c\u70ba (4) \u5176\u4e2d 1 1 i i n W \u2212 \u2212 + \u4ee3\u8868 1 2 1 ...... i n i n i W W W \u2212 + \u2212 + \u2212 \u8a5e\u5e8f\u5217\uff0c\u5982\u6b64\u4e00\u4f86\u4f7f\u7528 n-gram \u53ef\u4ee5\u5927\u91cf\u7bc0\u7701\u8a08 \u7b97\u6642\u9593\u8207\u8a18\u61b6\u9ad4\uff0c\u8b93\u5be6\u7528\u6027\u5927\u70ba\u63d0\u9ad8\u3002\u800c\u4e00\u822c\u5728\u5efa\u7acb n-gram \u6a5f\u7387\u6a21\u578b 1 1 ( | ) i i i n P W W \u2212 \u2212 + \u6700\u76f4\u89ba\u7684\u65b9\u6cd5\u5c31\u662f\u7d71\u8a08\u5728\u8a5e\u5e8f\u5217 1 2 1 ...... i n i n i W W W \u2212 + \u2212 + \u2212 \u5f8c\u51fa\u73fe i W \u7684\u6b21\u6578\u518d\u9664\u4ee5\u8a5e\u5e8f \u5217 1 2 1 ...... i n i n i W W W \u2212 + \u2212 + \u2212 \u5728\u8a13\u7df4\u6587\u96c6\u4e2d\u51fa\u73fe\u7684\u6b21\u6578\uff0c\u4e5f\u5c31\u662f (5) \u5176\u4e2d ( ) j i C W \u4ee3\u8868 j i W \u5728\u8a13\u7df4\u6587\u96c6\u4e2d\u51fa\u73fe\u7684\u6b21\u6578\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "= = = \u220f 1 1 1 1 1 1 1 ( ) ( ) ( | ) ( ) ( ) i i i i i n i n i i n i i i n i n W C W C W P W W C W C W \u2212 \u2212 + \u2212 + \u2212 + \u2212 \u2212 + \u2212 + = = \u2211 1 2",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "\u80a1\"\u9019\u6bb5\u8a5e\u5e8f\u5217\u6703\u4e0d\u65b7\u51fa\u73fe\uff0c\u900f\u904e\u6211\u5011\u7684\u8a5e\u5178\uff0c\u6703\u5c07\u6b64\u8a5e\u5e8f\u5217\u65b7\u8a5e\u70ba\"\u91d1\u878d\"\u8207\"\u80a1\" \u5169\u500b\u8a5e\uff0c\u6b64\u6642\u82e5\u6211\u5011\u5728\u7b2c\u4e00\u6b21\u6e2c\u8a66\u5230\u6b64\u8a5e\u5e8f\u5217\u6642\uff0c\u5c07 \"\u91d1\u878d\" \u5f8c\u9762\u63a5 \"\u80a1\" \u7684\u6a5f \u7387\u63d0\u9ad8\uff0c\u81ea\u7136\u53ef\u4ee5\u589e\u5f37\u6211\u5011\u6a21\u578b\u7684\u6e96\u78ba\u6027\uff0c\u5728\u5feb\u53d6\u6a21\u578b\u4e2d\u6703\u4fdd\u7559\u4e00\u584a\u5feb\u53d6\u8a18\u61b6 \u9ad4\uff0c\u800c\u505a\u6587\u4ef6\u6e2c\u8a66\u6642\uff0c\u6703\u5c07\u6700\u8fd1\u6e2c\u8a66\u904e\u7684\u6587\u53e5\u62ff\u4f86\u8a13\u7df4\u51fa\u5feb\u53d6 n-gram \u6a21\u578b P c \u5c07 \u5176\u8207\u539f\u59cb\u7684\u7d71\u8a08\u6a21\u578b P S \u505a\u7d50\u5408\uff0c\u6211\u5011\u5c07\u6a21\u578b\u6a5f\u7387\u7528(9)\u5f0f\u8868\u793a (9) \u5176\u4e2d \u00b5 \u4ee3\u8868\u7d50\u5408\u6bd4\u91cd\u3002 \u800c\u5728\u672c\u8ad6\u6587\u4e2d\uff0c\u6211\u5011\u4f7f\u7528\u7684\u662f\u6587\u53e5\u968e\u5c64\u6df7\u5408\u5f0f n-gram \u6a21\u578b(sentence-level mixture n-gram model)\uff0c\u5728\u6bcf\u7d93\u904e\u4e00\u6587\u53e5\u5f8c\uff0c\u5c31\u5229\u7528\u6b64\u6587\u53e5\u6240\u63d0\u4f9b\u7684\u8cc7\u8a0a\u8abf\u6574\u6df7 \u5408\u6a21\u578b\u7684\u6bd4\u91cd\u53c3\u6578\u3002\u6211\u5011\u662f\u5229\u7528\u5947\u6469\u7db2\u7ad9\u5df2\u5206\u985e\u597d\u7684\u65b0\u805e\uff0c\u505a\u70ba\u6211\u5011\u7684\u5206\u985e\u7fa4 \u7d44\u3002\u800c\u6211\u5011\u6703\u4f9d\u64da\u5206\u7fa4\u904e\u5f8c\u4e4b\u6587\u96c6\u8a13\u7df4\u51fa\u5c0d\u61c9\u65bc\u5404\u7fa4\u7d44\u4e4b n-gram \u6a21\u578b\uff0c\u5728\u9019\u908a \u4ee5 P k \u8868\u793a\u7b2c k \u500b\u7fa4\u7d44\u7684 n-gram \u6a21\u578b\uff0c\u800c\u5728\u505a\u6e2c\u8a66\u6642\uff0c\u4f7f\u7528\u6b0a\u91cd k \u03bb \u5c07\u5404\u7fa4\u4e4b\u6a21 \u578b \u505a \u7d44 \u5408 \u6210 \u70ba \u6e2c \u8a66 \u7528 \u7684 n-gram \u6a21 \u578b \uff0c \u4e5f \u5c31 \u662f \u8aaa \u5047 \u8a2d \u6709 \u4e00 \u6587 \u53e5 S \u70ba W 1 W 2 W 3 \u2026\u2026W T \uff0c\u5247\u6b64\u6587\u53e5\u51fa\u73fe\u7684\u6a5f\u7387\u70ba (10) \u5176\u4e2d m \u4ee3\u8868\u6df7\u5408\u6578\u500b\u6578\uff0c\u4f46\u70ba\u6b64\u6a21\u578b\u9084\u9808\u505a\u5169\u9ede\u6539\u9032\uff0c\u7b2c\u4e00\u3001\u70ba\u4e86\u907f\u514d\u6bcf\u500b\u7fa4 \u7d44\u4e2d\u7684\u8a13\u7df4\u6587\u96c6\u592a\u5c11\uff0c\u9020\u6210\u8cc7\u6599\u7a00\u758f(data sparseness)\uff0c\u6bcf\u500b\u55ae\u4e00\u7fa4\u7d44\u6a21\u578b\u9700\u8981\u518d \u7d50\u5408\u4e00\u500b\u4e00\u822c\u5316\u7684\u6a21\u578b(general model)\uff0c\u7528\u4ee5\u589e\u52a0\u6a21\u578b\u7684\u53ef\u9760\u5ea6\uff0c\u7b2c\u4e8c\u3001\u5728\u6e2c\u8a66 \u6642\u53ef\u80fd\u6703\u6709\u7121\u9818\u57df(nontopic)\u7684\u6587\u96c6\u5b58\u5728\uff0c\u6240\u4ee5\u6211\u5011\u53c8\u5fc5\u9808\u5c07\u4e00\u822c\u5316\u6a21\u578b\u52a0\u5165\uff0c \u8996\u70ba\u4e00\u500b\u7121\u9818\u57df\u7684\u7fa4\u7d44\uff0c\u5728\u6b64\u6211\u5011\u5c07\u4e00\u822c\u5316\u6a21\u578b\u4ee5 P G \u8868\u793a\uff0c\u6545\u4e0a\u5f0f\u53ef\u6539\u5beb\u70ba , 1 1 1 1 1 1 1 ( ) [ ( | ) (1 ) ( | )] m G T i i k k k i in k G i in k i P S P W W P W W \u03bb \u03b1 \u03b1 + \u2212 \u2212 \u2212 + \u2212 + = = = + \u2212 \u2211 \u220f (11) \u5176\u4e2d k \u03b1 \u70ba\u7b2c k \u500b\u7fa4\u7d44\u6a21\u578b\u8207\u4e00\u822c\u5316\u6a21\u578b\u7684\u7d44\u5408\u6b0a\u91cd\u3002\u5728\u6df7\u5408\u5f0f n-gram \u6a21\u578b\u4e2d\uff0c \u6709\u5169\u500b\u6b0a\u91cd k \u03b1 \u53ca k \u03bb \u5b58\u5728\uff0c\u57fa\u672c\u4e0a\u6df7\u5408\u5f0f n-gram \u6a21\u578b\u662f\u4f9d\u64da\u524d\u6587\u4f86\u52d5\u614b\u7684\u8abf\u6574\u6b64 \u4e8c\u6b0a\u91cd\uff0c\u5728\u521d\u59cb\u6642\u6703\u4f7f\u7528\u5c11\u6578\u4fdd\u7559\u6587\u96c6\u4f30\u6e2c\u51fa\u5176\u521d\u59cb\u503c\uff0c\u6e2c\u8a66\u6642\u6703\u5728\u6bcf\u4e00\u6587\u53e5\u7d50 \u675f\u6642\u518d\u53bb\u505a\u4e00\u6b21\u6b0a\u91cd\u7684\u8abf\u6574\uff0c\u800c\u8abf\u6574\u7684\u52d5\u4f5c\u53ef\u4ee5\u5206\u5225\u5beb\u6210(12)(13)\u5f0f",
+                        "eq_num": "(12)"
+                    }
+                ],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "1 1 1 1 1 1 1 1 1 1 ( | ) ( | ) (1 ) ( | ) k l k N T old i new k k i i n k N old i old i l i k k i i n k G i i n l l P W W P W W P W W T \u03b1 \u03b1 \u03b1 \u03b1 \u2212 \u2212 + \u2212 \u2212 = = \u2212 + \u2212 + = = + \u2212 \u2211\u2211 \u2211 1 1 1 1 2 1 1 1 ( ...... ) [(1 ) ( | ) ( | )] T s i c i T i i n i i n i P WW W P W W P W W \u00b5 \u00b5 + \u2212 \u2212 \u2212 + \u2212 + = = \u2212 + \u220f 1 1 1 1 1 1 1 11 ( ) ( | ) ( | ) T T m i i i i n k k i i n i",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u6a21\u578b\u4e2d\u6c92\u6709\u8a13\u7df4\u5230\u7684\u8a5e\u5e8f\u5217\u6a5f\u7387\u6a21\u578b\u4f7f\u7528(n-1)-gram \u6a21\u578b\u505a\u88dc\u511f\uff0c\u4e5f\u5c31\u662f 1 1 1 1 1 1 1 interp 1 1 interp 2 ( | ) ( | ) (1 ) ( | ) i i i N i N i i i i i N i i N i i N W W P W W P W W p W W \u03bb \u03bb \u2212 \u2212 \u2212 + \u2212 + \u2212 \u2212 \u2212 \u2212 + \u2212 + \u2212 + = + \u2212 (14) \u9019\u662f\u4e00\u500b\u905e\u8ff4\u5f0f\u7684\u5b9a\u7fa9\uff0c\u6240\u6709\u7684 n-gram \u6a21\u578b\u90fd\u5fc5\u9808\u5229\u7528(n-1)-gram \u6a21\u578b\u505a\u88dc\u511f\uff0c \u5176\u4e2d 1 1 i i N W \u03bb \u2212 \u2212 + \u4ee3\u8868\u7684\u662f\u5408\u4f75 n-gram \u8207 (n-1)-gram \u4e4b\u6b0a\u91cd\uff0c\u800c Witten-Bell \u5e73\u6ed1\u5316\u6280 \u8853\u5c0d\u6b64\u4e00\u6b0a\u91cd\u6709\u4e00\u500b\u7279\u6b8a\u7684\u4f30\u6e2c\u65b9\u5f0f\uff0c\u5728\u9019\u908a\u5148\u5c0d\u7b26\u865f\u505a\u4ee5\u4e0b\u7684\u5b9a\u7fa9 1 1 1 1 1 ( ) |{ : ( ) 0}| i i i n i i n i N W W C W W \u2212 \u2212 + \u2212 + \u2212 + \u22c5 = > (15) 1 1 1 ( ) i i n N W \u2212 + \u2212 + \u22c5 \u4ee3\u8868\u5728 1 1 i i n W \u2212 \u2212 + \u5f8c\u53ef\u63a5\u7684\u8a5e\u6578\uff0c\u5176\u4e2d\u4e0b\u6a19\u300c1+\u300d\u4ee3\u8868\u662f\u9023\u63a5\u4e00\u500b\u8a5e \u4ee5\u4e0a\u4e4b\u610f\u3002\u6b0a\u91cd\u56e0\u6578\u5b9a\u7fa9\u70ba (16) \u5373\u70ba Witten-Bell \u7684 n-gram \u6a21\u578b\u5efa\u7acb\u65b9\u5f0f\uff0c\u5176\u7269\u7406\u610f\u7fa9\u8868\u793a\u5728\u7d71\u8a08 1 1 i i n W \u2212 \u2212 + \u51fa\u73fe\u6b21\u6578 \u6642\uff0c\u5982\u679c 1 1 i i n W \u2212 \u2212 + \u5f8c\u9762\u53ef\u63a5\u7684\u8a5e\u6578\u8d8a\u5c11\uff0c\u6211\u5011\u7d66 1 1 ( | ) i i i n P W W \u2212 \u2212 + \u8f03\u5927\u7684\u6b0a\u91cd\uff0c\u53cd\u4e4b\u5247\u4f7f \u7528\u8f03\u591a\u7684(n-1)-gram \u505a\u88dc\u511f\uff0c\u5047\u8a2d\u5728\u505a bigram \u7d71\u8a08\u6642\uff0c\u8a5e\u5178\u4e2d\u6709\u4e00\u8a5e\u70ba \u300c\u985e\u795e\u7d93\u300d \uff0c \u6211\u5011\u767c\u73fe\u5728\u8a13\u7df4\u6587\u96c6\u4e2d\u300c\u985e\u795e\u7d93\u300d\u5f8c\u90fd\u63a5\u300c\u7db2\u8def\u300d\u4e00\u8a5e\uff0c\u6b64\u6642\u5c31\u4e0d\u9700\u8981\u592a\u591a\u7684 unigram \u505a\u88dc\u511f\uff0c\u9019\u662f\u56e0\u70ba\u6b64\u540d\u8a5e\u6709\u7368\u7279\u6027\uff0c\u5f8c\u9762\u5e7e\u4e4e\u90fd\u63a5\u5c11\u91cf\u7279\u5b9a\u7684\u8a5e\uff0c\u800c\u82e5 \u6b32\u7d71\u8a08\u4e00\u8a5e\u300c\u5e7e\u4e4e\u300d\u5f8c\u53ef\u63a5\u8a5e\u7684 bigram \u6a5f\u7387\uff0c\u53ef\u80fd\u6703\u767c\u73fe\u8a13\u7df4\u6587\u96c6\u4e2d\u5176\u5f8c\u53ef\u63a5 1 , 1 1 1 ( ,..., ) 1 ( ,..., ) i i old N k k T new k m G old i j j T j P W W N P W W \u03bb \u03bb \u03bb = = = \u2211 \u2211 1 1 1 1 1 1 1 1 1 ( ) 1 ( ) ( ) i i n i i i n i i W i n i n W N W N",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(2)\u5927\u9678 \u5c0f\u4e09\u901a \u21d2 \u5169\u5cb8 confidence = 100% support = 2.25%",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u4ee5\u4e0a\u4f8b\u800c\u8a00\uff0c\u51fa\u73fe\"\u5c0f\u4e09\u901a\"\u5f8c\uff0c\u65b0\u805e\u8a9e\u6599\u6709 90 % \u7684\u6a5f\u7387\u4e5f\u6703\u51fa\u73fe \"\u5927\u9678\" \u9019\u500b \u8a5e\uff0c\u9019\u500b\u95dc\u806f\u6cd5\u5247\u4f54\u4e86\u7e3d\u6587\u53e5 6.25%\uff0c\u5982\u679c\"\u5927\u9678\"\u8207\"\u5c0f\u4e09\u901a\"\u540c\u6642\u51fa\u73fe\u5f8c\uff0c\u65b0\u805e \u8a9e\u6599\u6703\u6709 100 %\u7684\u6a5f\u7387\u4e5f\u6703\u51fa\u73fe\"\u5169\u5cb8\"\uff0c\u800c\u6b64\u95dc\u806f\u6cd5\u5247\u4f54\u4e86\u7e3d\u6587\u53e5 2.25%\u3002 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Fast Algorithms for Mining Association Rules",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Agrawal",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Srikant",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "Proceedings of the 20 th VLDB Conference",
+                "volume": "",
+                "issue": "",
+                "pages": "487--499",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R. Agrawal and R. Srikant, \"Fast Algorithms for Mining Association Rules\", Proceedings of the 20 th VLDB Conference, Santiago-Chile, pp.487-499, 1994\u3002",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "An Empirical Study of Smoothing Techniques for Language Modeling",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Goodman",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Computer Speech and Language",
+                "volume": "13",
+                "issue": "",
+                "pages": "359--394",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "S. F. Chen and J. Goodman, \"An Empirical Study of Smoothing Techniques for Language Modeling\", Computer Speech and Language , vol.13, 359-394 , 1999\u3002",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Language Model Adaptation Using Mixtures and an Exponentially Decaying Cache",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Clarkson",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Robinson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proc. of ICASSP",
+                "volume": "",
+                "issue": "",
+                "pages": "799--802",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "P. R. Clarkson and A. J. Robinson, \"Language Model Adaptation Using Mixtures and an Exponentially Decaying Cache\", Proc. of ICASSP, pp.799-802 , 1997\u3002",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Relevance weighting for combining multi-domain data for n-gram language modeling",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Iyer",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Ostendorf",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Computer Speech and Language",
+                "volume": "13",
+                "issue": "",
+                "pages": "267--282",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R. Iyer and M. Ostendorf, \"Relevance weighting for combining multi-domain data for n-gram language modeling\", Computer Speech and Language, vol.13, pp.267-282, 1999\u3002",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Modeling long distance dependence in language : Topic Mixtures Versus dynamic cache models",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Iyer",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Ostendorf",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "IEEE Transaction on speech and audio processing",
+                "volume": "7",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R. M. Iyer and M. Ostendorf, \"Modeling long distance dependence in language : Topic Mixtures Versus dynamic cache models\", IEEE Transaction on speech and audio processing , vol.7 , January 1999\u3002",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Interpolation estimation of Markov source parameters from sparse data",
+                "authors": [
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Jelinek",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Mercer",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1980,
+                "venue": "Proceedings of the workshop on pattern recognition in Practice",
+                "volume": "",
+                "issue": "",
+                "pages": "381--397",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "F. Jelinek and R. L. Mercer, \"Interpolation estimation of Markov source parameters from sparse data\", Proceedings of the workshop on pattern recognition in Practice, North-Holland, Amsterdam, The Netherlands, pp.381-397, May 1980\u3002",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Selecting Articles from the Language Model Training Corpus",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Klakow",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Proc of ICASSP",
+                "volume": "",
+                "issue": "",
+                "pages": "1695--1698",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "D. Klakow, \"Selecting Articles from the Language Model Training Corpus\", Proc of ICASSP, pp.1695 -1698, 2000\u3002",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "A maximum entropy approach to adaptive statistical language model",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Rosenfeld",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Computer Speech and Language",
+                "volume": "10",
+                "issue": "",
+                "pages": "187--228",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R. Rosenfeld, \"A maximum entropy approach to adaptive statistical language model\", Computer Speech and Language, vol 10 , pp.187-228 , 1996.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Trigger-based language models: A maximum entropy approach",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Lau",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Rosenfeld",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Roukos",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Proc. Int. Conf. Acoustics, Speech, Signal Processing",
+                "volume": "II",
+                "issue": "",
+                "pages": "45--48",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R. Lau, R. Rosenfeld, and S. Roukos, \"Trigger-based language models: A maximum entropy approach\" , in Proc. Int. Conf. Acoustics, Speech, Signal Processing, vol. II, pp. 45-48. , 1993",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Statistical language modeling combining N -gram and context free grammars",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Meteer",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Rohlicek",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Proc. Int. Conf. Acoustics, Speech, Signal Processing",
+                "volume": "II",
+                "issue": "",
+                "pages": "37--40",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "M. Meteer and J. R. Rohlicek, \"Statistical language modeling combining N -gram and context free grammars\" , in Proc. Int. Conf. Acoustics, Speech, Signal Processing, vol. II, pp. 37-40 , 1993.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Foundations of statistical natural language processing",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Manning",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Schutze",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Massachusetts Institute of Technology",
+                "volume": "",
+                "issue": "",
+                "pages": "315--407",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C. D. Manning, H. Schutze, \"Foundations of statistical natural language processing\", Massachusetts Institute of Technology pp.315-407, 1999\u3002",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Funadamental of Speech Recognition",
+                "authors": [
+                    {
+                        "first": "L",
+                        "middle": [],
+                        "last": "Rabiner",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "B",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Juang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "321--387",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "L. Rabiner and B.H. Juang, \"Funadamental of Speech Recognition\", Prentice Hall, pp.321-387, 1993\u3002",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "The zero-frequency problem : Estimating the probabilities of novel events in adaptive text compression",
+                "authors": [
+                    {
+                        "first": "I",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Witten",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Bell",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "IEEE Transactions on Information Theory",
+                "volume": "37",
+                "issue": "",
+                "pages": "1085--1094",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "I. H. Witten and T. C. Bell, \"The zero-frequency problem : Estimating the probabilities of novel events in adaptive text compression.\", IEEE Transactions on Information Theory , vol.37, pp.1085-1094, 1991\u3002",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Interpolation of n-gram and mutual-information based trigger pair language models for Mandarin speech recognition",
+                "authors": [
+                    {
+                        "first": "G",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Zhou",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "T"
+                        ],
+                        "last": "Lua",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Computer Speech and Language",
+                "volume": "13",
+                "issue": "",
+                "pages": "125--141",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "G. D. Zhou and K. T. Lua, \"Interpolation of n-gram and mutual-information based trigger pair language models for Mandarin speech recognition\", Computer Speech and Language, vol. 13, pp.125-141, 1999\u3002",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "type_str": "figure",
+                "num": null,
+                "uris": null,
+                "text": ",b)\uff0c(a,c)\uff0c(b,c)\uff0c\u4e26\u4e14\u5c0d\u8cc7\u6599\u5eab\u641c\u5c0b\u6bcf\u4e00\u5e8f\u5c0d\uff0c\u662f \u5426\u540c\u6642\u51fa\u73fe\u5728\u65bc\u540c\u4e00\u6587\u53e5\u4e2d\uff0c\u5047\u8a2d\u53ea\u6709(a,b)\uff0c(b,c)\u5e8f\u5c0d\u7b26\u5408\u9019\u9805\u689d\u4ef6\uff0c\u5247\u5c07(a,c) \u522a\u9664\uff0c\u6b64\u6642\u6211\u5011\u5efa\u7acb(a,b)\uff0c(b,c)\u7684\u95dc\u806f\u6cd5\u5247\uff0c\u6b64\u95dc\u806f\u6cd5\u5247\u7684\u5c64\u7d1a(step) \u70ba\u4e8c \uff0c\u4e0d \u904e\u6211\u5011\u5fc5\u9808\u8a08\u7b97\u5176\u4fe1\u8cf4\u5ea6\u8207\u652f\u63f4\u5ea6\uff0c\u4f8b\u5982\u6211\u5011\u53ef\u4ee5\u8a08\u7b97\u540c\u4e00\u7bc7\u6587\u7ae0\u51fa\u73fe\u8a5e a \u4e14\u51fa \u73fe\u8a5e b \u7684\u6a5f\u7387\uff0c\u5373\u70ba\u5176\u4fe1\u8cf4\u5ea6\u3002\u800c\u6211\u5011\u6703\u518d\u5c07\u5269\u9918\u4e0b\u4e4b\u5e8f\u5c0d(a,b)\uff0c(b,c)\u505a\u7d50\u5408\u6210 \u70ba(a,b,c) \u4e26\u641c\u5c0b\u8a13\u7df4\u6587\u96c6\u4e2d(a,b,c\u7684\u6587\u7ae0\u4e2d\u6709 c%\u7684\u6a5f\u7387\u6703\u51fa\u73fe Y \uff0c\u800c\u6709 s% \u7684\u6587\u7ae0\u540c\u6642\u5305\u542b\u4e86 WordSeq \u8207 Y\u3002\u4ee5\u4e0b\u70ba\u5229\u7528\u897f\u5143\u4e8c\u5343\u96f6\u4e00\u5e74\u5341\u4e8c\u6708\u4e8c\u5341\u516b\u865f\u5230\u897f\u5143\u4e8c\u5343\u96f6\u4e00\u5e74\u5341\u4e8c\u6708\u4e09\u5341 \u4e00\u865f\u671f\u9593\u7684\u653f\u6cbb\u65b0\u805e\u6240\u64f7\u53d6\u51fa\u7684\u5169\u689d\u95dc\u806f\u6cd5\u5247\u7bc4\u4f8b\uff0c\u7b2c\u4e00\u689d\u95dc\u806f\u6cd5\u5247\u7684\u5c64\u7d1a\u70ba \u4e8c\uff0c\u7b2c\u4e8c\u689d\u7684\u5c64\u7d1a\u5247\u70ba\u4e09 \u5c0f\u4e09\u901a \u21d2 \u5927\u9678 confidence = 90% support = 6.25%"
+            },
+            "FIGREF1": {
+                "type_str": "figure",
+                "num": null,
+                "uris": null,
+                "text": "ig ra m + A s s o c ia tio n ru le B ig ra m + T rig g e r p a ir perplexity a s s o c ia tio n s te p"
+            },
+            "TABREF0": {
+                "text": "",
+                "content": "<table><tr><td>\u7684\u9ed1\u624b\u5c31\u662f\u5efa\u7acb\u5728 \"\u4f7f\u4eba\u4fbf\u5229\" \u7684\u57fa\u790e\u4e4b\u4e0a\uff0c\u4f46\u662f\u96fb\u8166\u81ea\u5f9e\u5728\u767c\u660e\u4e4b\u521d\u5c31\u5b58\u5728\u4e00 \u65b0\u805e\u6587\u4ef6\u5206\u985e\u7684\u7cfb\u7d71\u4e0a\u6709\u4e00\u5b9a\u5e45\u5ea6\u7684\u5e6b\u52a9\u3002</td></tr><tr><td>\u500b\u8207\u4eba\u6027\u80cc\u9053\u800c\u99b3\u7684\u7f3a\u9ede\uff0c\u8207\u5b83\u5011\u7684\u6e9d\u901a\u9700\u8981\u900f\u904e\u4e00\u500b\u7279\u5b9a\u7684\u6309\u9375\u88dd\u7f6e\uff0c\u6bd4\u65b9\u8aaa</td></tr><tr><td>\u8981\u8207\u500b\u4eba\u96fb\u8166\u6e9d\u901a\u5c31\u5fc5\u9808\u900f\u904e\u9375\u76e4\u6216\u6ed1\u9f20\u7b49\u88dd\u7f6e\uff0c\u4e8b\u5be6\u4e0a\u9019\u662f\u4f7f\u8a31\u591a\u4eba\u5c0d\u96fb\u8166\u671b 2. n-gram \u8a9e\u8a00\u6a21\u578b\u7c21\u4ecb</td></tr><tr><td>\u4e4b\u537b\u6b65\u7684\u539f\u56e0\uff0c\u8981\u5b78\u7fd2\u5982\u4f55\u4f7f\u7528\u9375\u76e4\u8207\u96fb\u8166\u505a\u6e9d\u901a\u5c31\u7b49\u65bc\u662f\u5f37\u8feb\u4eba\u53bb\u5b78\u7fd2\u3127\u7a2e \u76ee\u524d n-gram[11]\u6a21\u578b\u7684\u63a2\u8a0e\u65bc\u5404\u76f8\u95dc\u5b78\u8853\u6703\u8b70\u53ca\u671f\u520a\u8ad6\u6587\u4e0a\u5df2\u6709\u76f8\u7576\u591a\u7684</td></tr><tr><td>\"\u96fb\u8166\u8a9e\u8a00\"\uff0c\u9019\u8207 \"\u4f7f\u4eba\u4fbf\u5229\" \u7684\u539f\u5247\u7576\u7136\u662f\u4e92\u76f8\u9055\u80cc\u7684\uff0c\u4f46\u662f\u53cd\u904e\u4f86\u8aaa\u5982\u80fd\u8b93 \u6587\u737b\u767c\u8868\uff0c\u986f\u793a\u5404\u7a2e\u7814\u7a76\u6a5f\u69cb\u5c0d\u6b64\u4e00\u9818\u57df\u7684\u767c\u5c55\u6709\u76f8\u7576\u5927\u7684\u671f\u8a31\uff0c\u6545\u6295\u8eab\u65bc\u5176</td></tr><tr><td>\u96fb\u8166\u5b78\u7fd2\u4eba\u985e\u7684\u8a9e\u8a00\uff0c\u4f7f\u96fb\u8166\u80fd\u66f4\u63a5\u8fd1\u4eba\u985e\uff0c\u4e5f\u5c31\u80fd\u4f7f\u5176\u8207\u4eba\u985e\u751f\u6d3b\u7684\u7d50\u5408\u66f4\u52a0 \u4e2d\uff0c\u800c\u5728\u5404\u65b9\u90fd\u81f4\u529b\u65bc\u6539\u9032 n-gram \u6a21\u578b\u4e4b\u4e0b\uff0cn-gram \u6a21\u578b\u5728\u6548\u80fd\u4e0a\u5df2\u7372\u5f97\u76f8\u7576</td></tr><tr><td>\u7c21\u4ec1\u5b97 \u9673\u9d3b\u5100 \u7dca\u5bc6\uff0c\u9032\u4e00\u6b65\u5982\u679c\u96fb\u8166\u80fd\u900f\u904e\u8a9e\u8a00\u7684\u5b78\u7fd2\u800c\u5177\u5099\u4e86\u95b1\u8b80\u7684\u80fd\u529b\uff0c\u6211\u5011\u5c31\u53ef\u4ee5\u8b93\u96fb \u4e0d\u932f\u4e4b\u6210\u679c\uff0c\u5728\u672c\u7ae0\u4e2d\u6211\u5011\u5c07\u6703\u5c0d n-gram \u6a21\u578b\u7684\u57fa\u672c\u6982\u5ff5\u505a\u4e00\u7c21\u55ae\u4e4b\u4ecb\u7d39\u3002</td></tr><tr><td>\u570b\u7acb\u6210\u529f\u5927\u5b78\u8cc7\u8a0a\u5de5\u7a0b\u5b78\u7cfb \u8166\u70ba\u6211\u5011\u904e\u6ffe\u4ea6\u6216\u5206\u985e\u6bcf\u5929\u6240\u9700\u95b1\u8b80\u7684\u6587\u4ef6\uff0c\u6bd4\u65b9\u8aaa\u53ef\u4ee5\u61c9\u7528\u5728\u65bc e-mail \u5ee3\u544a\u904e</td></tr><tr><td>Email\uff1ajtchien@mail.ncku.edu.tw \u6ffe\u6216\u662f\u65b0\u805e\u6587\u4ef6\u5206\u985e\u7b49\u7b49\uff0c\u5c31\u53ef\u4ee5\u8b93\u96fb\u8166\u70ba\u6211\u5011\u7701\u4e0b\u66f4\u591a\u7684\u6642\u9593\u3002 2.1 n-gram \u6a21\u578b\u4e4b\u61c9\u7528</td></tr><tr><td>\u8981\u514b\u670d\u96fb\u8166\u8207\u4eba\u5728\u8a9e\u8a00\u4e0a\u7684\u9d3b\u6e9d\uff0c\u5728\u8a9e\u8a00\u6280\u8853\u7684\u9818\u57df\u6709\u4e86\u8072\u5b78\u6a21\u578b(acoustic \u4e00\u822c\u800c\u8a00 n-gram \u8a9e\u8a00\u6a21\u578b\u901a\u5e38\u61c9\u7528\u65bc\u8c9d\u5f0f\u5206\u985e\u5668(Bayes classifier)\uff0c\u626e\u6f14\u8457</td></tr><tr><td>\u6458\u8981 model)\u8207\u81ea\u7136\u8a9e\u8a00\u6a21\u578b(natural language model)\u7684\u7522\u751f\uff0c\u800c\u9019\u5169\u9805\u6280\u8853\u7684\u767c\u5c55\u5728\u570b \u4e8b\u524d\u6a5f\u7387(priori probability)\u6216\u662f\u53ef\u80fd\u6027( likelihood )\u7684\u89d2\u8272\uff0c\u4ee5\u8a9e\u97f3\u8fa8\u8b58\u70ba\u4f8b\u5b50\u800c</td></tr><tr><td>\u672c\u8ad6\u6587\u63d0\u51fa\u4e00\u7a2e\u80fd\u64f7\u53d6\u9577\u8ddd\u96e2\u8cc7\u8a0a\u7684\u8a9e\u8a00\u6a21\u578b\uff0c\u5b83\u53ef\u4ee5\u64f7\u53d6\u591a\u8a5e\u5f59\u4e4b\u9593\u7684\u95dc \u5916\u5df2\u7d93\u884c\u4e4b\u6709\u5e74\uff0c\u53f0\u7063\u81ea\u897f\u5143\u4e00\u4e5d\u516b\u4e8c\u5e74\u8d77\u4fbf\u958b\u59cb\u6709\u4e86\u4e2d\u6587\u8072\u5b78\u6a21\u578b\u65b9\u9762\u7684\u7814 \u8a00\uff0c\u5047\u8a2d\u6709\u4e00\u6bb5\u8072\u5b78\u8a0a\u865f(acoustic signal)X\uff0c\u6211\u5011\u7684\u76ee\u6a19\u662f\u53bb\u627e\u5c0b\u51fa\u6b64\u8a0a\u865f\u6700\u6709</td></tr><tr><td>\u806f\u6027\uff0c\u64f7\u53d6\u7684\u65b9\u5f0f\u662f\u4f7f\u7528\u8cc7\u6599\u63a2\u52d8\u4e2d\u5341\u5206\u6d41\u884c\u7684 Apriori \u6f14\u7b97\u6cd5\uff0c\u50b3\u7d71\u4e0a n-gram \u7a76\uff0c\u8a31\u591a\u7814\u7a76\u55ae\u4f4d\u5305\u62ec\u53f0\u6e05\u4ea4\u6210\u7b49\u5927\u5c08\u9662\u6821\uff0c\u4ee5\u53ca\u5de5\u7814\u9662\u3001\u4ea4\u901a\u90e8\u3001\u4e2d\u7814\u9662\u3001\u4e2d</td></tr><tr><td>\u8a9e\u8a00\u6a21\u578b\u53ea\u80fd\u5728 n-gram \u8996\u7a97\u5167\u64f7\u53d6\u5230\u6709\u9650\u8ddd\u96e2\u7684\u8cc7\u8a0a\uff0c\u8f03\u9577\u8ddd\u96e2\u7684\u8cc7\u8a0a\u4e5f\u5c31\u56e0 \u83ef\u96fb\u4fe1\u7b49\u90fd\u7a4d\u6975\u7684\u6295\u5165\u7814\u7a76\u7684\u5de5\u4f5c\u4e26\u4e14\u64c1\u6709\u4e86\u5341\u5206\u8c50\u78a9\u7684\u7814\u7a76\u6210\u679c\uff0c\u800c\u5728\u8072\u5b78\u6a21</td></tr><tr><td>\u6b64\u800c\u6d41\u5931\uff0c\u7136\u800c\u9019\u4e9b\u5931\u53bb\u7684\u9577\u8ddd\u96e2\u8cc7\u8a0a\u5c0d\u65bc\u8a9e\u8a00\u6a21\u578b\u662f\u5341\u5206\u91cd\u8981\u7684\uff0c\u6240\u4ee5\u5982\u4f55\u514b \u578b\u5df2\u65e5\u76ca\u6210\u719f\u7684\u57fa\u790e\u4e0b\uff0c\u81ea\u7136\u8a9e\u8a00\u6a21\u578b\u7684\u767c\u5c55\u4e5f\u5099\u53d7\u77da\u76ee\uff0c\u8aa0\u5982\u524d\u6587\u6240\u8ff0\uff0c\u8a9e\u97f3</td></tr><tr><td>\u670d n-gram \u6a21\u578b\u7f3a\u4e4f\u9577\u8ddd\u96e2\u8cc7\u8a0a\u4e00\u76f4\u662f\u975e\u5e38\u71b1\u9580\u7684\u7814\u7a76\u8ab2\u984c\uff0c\u89f8\u767c\u5e8f\u5c0d\u5c31\u662f\u5176\u4e2d \u6280\u8853\u767c\u5c55\u7684\u6700\u7d42\u76ee\u7684\u5c31\u662f\u8981\u5c07\u96fb\u8166\u8207\u4eba\u985e\u7684\u6e9d\u901a\u4fbf\u5229\u5316\uff0c\u800c\u8981\u9054\u5230\u9019\u500b\u76ee\u7684\uff0c\u5c07</td></tr><tr><td>\u4e00\u7a2e\u6709\u6548\u7684\u65b9\u6cd5\uff0c\u5176\u4e3b\u8981\u529f\u80fd\u662f\u5728\u64f7\u53d6\u9577\u8ddd\u96e2\u4e4b\u8a5e\u5e8f\u5c0d\u8cc7\u8a0a\uff0c\u4e5f\u5c31\u662f\u5efa\u7acb\u8d77\u8a5e\u8207 \u8a9e\u97f3\u6a21\u578b\u8207\u81ea\u7136\u8a9e\u8a00\u6a21\u578b\u505a\u7d50\u5408\u662f\u5fc5\u9808\u7684\uff0c\u6211\u5011\u7684\u8ad6\u6587\u4e3b\u8981\u5c31\u662f\u8457\u58a8\u65bc\u81ea\u7136\u8a9e\u8a00</td></tr><tr><td>\u8a5e\u4e4b\u9593\u7684\u95dc\u806f\u6027\uff0c\u7136\u800c\u6211\u5011\u6240\u63d0\u51fa\u7684\u95dc\u806f\u6cd5\u5247\u6280\u8853\u80fd\u64f7\u53d6\u591a\u5143\u8a5e\u7d44\u9593\u7684\u95dc\u806f\u6027\uff0c \u6a21\u578b\u7684\u63a2\u8a0e\uff0c\u6211\u5011\u5c07\u6703\u5c0d\u81ea\u7136\u8a9e\u8a00\u6a21\u578b\u4e2d\u7684\u4e00\u9805\u5341\u5206\u6210\u529f\u4e14\u5ee3\u6cdb\u904b\u7528\u7684\u6280\u8853</td></tr><tr><td>\u53ef\u4ee5\u8aaa\u662f\u9032\u4e00\u6b65\u6539\u826f\u8a5e\u7d44\u6578\u4e26\u5efa\u7acb\u66f4\u9577\u8ddd\u96e2\u8cc7\u8a0a\uff0c\u800c\u5be6\u9a57\u7d50\u679c\u4e5f\u986f\u793a\u672c\u8ad6\u6587\u65b9\u6cd5 n-gram \u8a9e\u8a00\u6a21\u578b\u505a\u4ecb\u7d39\uff0c\u4e26\u4e14\u5206\u6790\u5176\u5728\u50b3\u7d71\u4e0a\u7684\u7f3a\u9ede\u8207\u6539\u9032\u6280\u8853\uff0c\u800c\u672c\u8ad6\u6587\u4e5f</td></tr><tr><td>\u6bd4\u8d77\u50b3\u7d71\u89f8\u767c\u5e8f\u5c0d\u7372\u5f97\u8f03\u4f4e\u7684 perplexity\uff0c\u6b64\u95dc\u806f\u6cd5\u5247\u6280\u8853\u4e5f\u53ef\u4ee5\u6709\u6548\u7684\u8207\u5176\u4ed6 \u5c07\u6703\u91dd\u5c0d n-gram \u6a21\u578b\u5176\u4e2d\u4e00\u9805\u7f3a\u9ede-\u9577\u8ddd\u96e2\u8cc7\u8a0a\u7684\u7f3a\u4e4f\uff0c\u63d0\u51fa\u4e00\u5957\u65b0\u7684\u6539\u9032\u65b9</td></tr><tr><td>\u6a21\u578b\u8abf\u6574\u53ca\u6a21\u578b\u5e73\u6ed1\u5316\u7684\u6280\u8853\u7d50\u5408\uff0c\u5728\u8a9e\u8a00\u6a21\u578b\u7684\u6548\u7387\u6539\u5584\u65b9\u9762\u80fd\u6709\u66f4\u826f\u597d\u7684\u6548 \u6cd5\uff0c\u4e26\u4e14\u7d50\u5408\u5176\u4ed6\u6539\u9032\u65b9\u6cd5\uff0c\u9032\u800c\u767c\u5c55\u51fa\u4e00\u5957\u8f03\u6709\u6548\u7387\u7684 n-gram \u6a21\u578b\uff0c\u6211\u5011\u5c07</td></tr><tr><td>\u679c\uff0c\u6700\u5f8c\u672c\u8ad6\u6587\u4e5f\u5c07\u63d0\u51fa\u7684\u8a9e\u8a00\u6a21\u578b\u6210\u529f\u7684\u61c9\u7528\u5728\u8a9e\u97f3\u8fa8\u8b58\u8207\u6587\u4ef6\u5206\u985e\u4e0a\uff0c\u4e26\u5efa \u6703\u5c07\u5176\u61c9\u7528\u5728\u7d50\u5408\u8072\u5b78\u6a21\u578b\u505a\u8a9e\u97f3\u8fa8\u8b58\u548c\u6587\u4ef6\u5206\u985e\u7684\u9818\u57df\u4e4b\u4e0a\uff0c\u671f\u671b\u5c0d\u5176\u6b63\u78ba\u7387</td></tr><tr><td>\u7acb\u4e00\u5957\u500b\u4eba\u5316\u4e4b\u65b0\u805e\u700f\u89bd\u5668\u4e4b\u5c55\u793a\u7cfb\u7d71\u3002 \u6709\u4e00\u5b9a\u5e45\u5ea6\u7684\u6539\u5584\u3002</td></tr><tr><td>\u800c\u81ea\u7136\u8a9e\u8a00\u6a21\u578b\u65b9\u9762\u5728\u73fe\u4eca\u6709\u8a31\u591a\u4e0d\u540c\u7684\u767c\u5c55\uff0c\u4f9d\u5176\u5167\u5bb9\u4e3b\u8981\u5206\u70ba\u4e09\u500b\u65b9</td></tr><tr><td>1. \u7c21\u4ecb \u5411\uff0c\u4e00\u3001\u6839\u64da\u8a9e\u8a00\u5b78\u6240\u767c\u5c55\u51fa\u7684\u6587\u6cd5(grammar)\u5206\u6790\uff0c\u4e8c\u3001\u4ee5\u77e5\u8b58\u70ba\u57fa\u790e\u800c\u767c\u5c55</td></tr><tr><td>\u62dc\u786c\u9ad4\u6280\u8853\u4e0d\u65b7\u9032\u6b65\u7684\u8ca2\u737b\u4e4b\u4e0b\uff0c\uff0c\u4e00\u822c\u4eba\u6703\u5f88\u7406\u6240\u7576\u7136\u7684\u4f7f\u7528\u81ea\u52d5\u6ac3\u54e1\u6a5f \u7684\u8a9e\u8a00\u8cc7\u6599\u5eab\uff0c\u4e09\u3001\u8457\u91cd\u65bc\u7d71\u8a08\u800c\u767c\u5c55\u51fa\u7684 n-gram \u6a21\u578b\u3002\u800c\u6211\u5011\u4e3b\u8981\u662f\u8457\u58a8\u65bc</td></tr><tr><td>\u63d0\u6b3e\u6216\u662f\u5229\u7528\u81ea\u52d5\u7a7a\u8abf\u8a2d\u5099\u4f86\u63a7\u5236\u5ba4\u5167\u7684\u6eab\u5ea6\uff0c\u800c\u9019\u90fd\u662f\u7531\u65bc\u96fb\u8166\u7684\u81ea\u52d5\u5316\u7ba1\u7406 \u7d71\u8a08\u5f0f\u7684 n-gram \u6a21\u578b\uff0c\u5728\u7b2c\u4e8c\u7ae0\u4e2d\uff0c\u6211\u5011\u5c07\u5c0d n-gram \u6a21\u578b\u505a\u8a73\u7d30\u7684\u4ecb\u7d39\uff0c\u4e26\u5c0d</td></tr><tr><td>\u8b93\u751f\u6d3b\u8b8a\u7684\u5982\u6b64\u4fbf\u5229\uff0c\u6b63\u6240\u8b02 \"\u79d1\u6280\u59cb\u7d42\u4f86\u81ea\u4eba\u6027\" \uff0c\u63a8\u52d5\u79d1\u6280\u9032\u6b65\u7684\u90a3\u96bb\u5e55\u5f8c \u5176\u7f3a\u9ede\u52a0\u4ee5\u63a2\u8a0e\uff0c\u7b2c\u4e09\u7ae0\u4e2d\u5c07\u6703\u4ecb\u7d39\u50b3\u7d71\u4e0a\u91dd\u5c0d n-gram \u6a21\u578b\u7684\u7f3a\u9ede\u6240\u884d\u751f\u51fa\u7684</td></tr><tr><td>\u6539\u9032\u65b9\u6cd5\uff0c\u4e26\u4e14\u63d0\u51fa\u4e00\u7a2e\u80fd\u64f7\u53d6\u9577\u8ddd\u96e2\u8cc7\u8a0a\u7684\u8a9e\u8a00\u6a21\u578b\uff0c\u5c07\u5b83\u61c9\u7528\u5728\u8a9e\u97f3\u8fa8\u8b58\u6216</td></tr></table>",
+                "html": null,
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF1": {
+                "text": "",
+                "content": "<table><tr><td colspan=\"8\">1 ( , ,....., ) 2 T P S P W W ( ) W</td></tr><tr><td/><td/><td colspan=\"6\">1 ( ) ( | )..... ( | 2 1 T P W P W W P W W W 1 , 2</td><td>,.....,</td><td>T W</td><td>1 \u2212</td><td>)</td></tr><tr><td/><td/><td colspan=\"2\">T</td><td/><td/><td/></tr><tr><td/><td/><td>i</td><td>1 =</td><td colspan=\"4\">1 2 P W WW ( | i</td><td>,......,</td><td>i W</td><td>1 \u2212</td><td>)</td></tr><tr><td colspan=\"8\">\u57fa\u672c\u4e0a\u5728\u8a55\u4f30\u4e00\u500b n-gram \u6a21\u578b\u7684\u6548\u679c\u6642\u5e38\u4f7f\u7528 perplexity[12]\u9019\u500b\u8a55\u4f30\u6a19</td></tr><tr><td colspan=\"8\">\u6e96\uff0c\u800c\u4e8b\u5be6\u4e0a\u5b83\u662f\u5728\u8a08\u7b97\u6a5f\u7387\u6a21\u578b\u7684 entropy\uff0centropy \u5728\u8a0a\u606f\u7406\u8ad6\u4e0a\u6307\u7684\u662f\u5c07\u6a5f</td></tr><tr><td colspan=\"8\">\u7387 P \u4e58\u4ee5\u8cc7\u8a0a-logP\uff0c\u61c9\u7528\u5728 n-gram \u6a21\u578b\u7684\u8a55\u4f30\u5247\u8868\u793a\u70ba:</td></tr><tr><td>H</td><td>p</td><td colspan=\"3\">= \u2212</td><td>P</td><td>log</td><td>P</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>(6)</td></tr><tr><td colspan=\"8\">\u5176\u7269\u7406\u610f\u7fa9\u8868\u793a\u5728\u8a08\u7b97\u4e00\u500b n-gram \u6a21\u578b\u7684 entropy \u6642\uff0c\u5fc5\u9808\u5148\u5c07\u8a5e\u5178\u4e2d\u7684\u8a5e\u505a</td></tr><tr><td colspan=\"8\">\u7d44\u5408\uff0c\u5f62\u6210\u70ba\u7121\u9650\u9577\u7684\u8a5e\u5e8f\u5217 1 2 ...... Q WW W \uff0c\u4e26\u4e14\u5c07\u6240\u6709\u7684\u53ef\u80fd\u8a5e\u5e8f\u5217\u8a08\u7b97\u5176\u6a5f\u7387</td></tr><tr><td colspan=\"8\">\u8207\u8cc7\u8a0a\u7684\u4e58\u7a4d\u5f8c\u52a0\u7e3d\uff0c\u5373\u53ef\u5f97\u5230\u6b64 n-gram \u6a21\u578b\u7684 entropy\u3002\u4f46\u4e8b\u5be6\u4e0a\u4e0d\u5bb9\u6613\u5be6\u73fe</td></tr><tr><td colspan=\"8\">\u5982\u6b64\u8907\u96dc\u7684\u8a08\u7b97\uff0c\u5fc5\u9808\u5047\u8a2d\u53ef\u4ee5\u63d0\u4f9b\u4e00\u6bb5\u8db3\u5920\u9577\u7684\u8a5e\u5e8f\u5217\u4f86\u4ee3\u8868\u6240\u6709\u7684\u8a5e\u5e8f\u5217\u7d44</td></tr></table>",
+                "html": null,
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF2": {
+                "text": "\u9577\u8ddd\u96e2\u8cc7\u8a0a(long distance)\u7f3a\u4e4f\u4e4b\u554f\u984c : n-gram \u6a21\u578b\u5728\u8a08\u7b97\u4e0a\u7684\u512a\u52e2\u662f\u5728\u65bc\u5b83\u4f7f\u7528\u4e86 n-gram \u8996\u7a97(n-gram window)\u505a",
+                "content": "<table><tr><td>\u5408\uff0c\u9019\u7a2e\u5047\u8a2d\u5728\u7d71\u8a08\u5b78\u4e0a\u7a31\u70ba\u6b64\u8a5e\u5e8f\u5217\u70ba ergodic\uff0c\u6545(8)\u5f0f\u53ef\u6539\u5beb\u70ba 1 2 1 ( ) log ( ...... ) p Q H PW W W Q = \u2212 3.\u70ba\u57fa\u790e\uff0c\u7bc0\u7701\u4e86\u5927\u91cf\u7684\u8a18\u61b6\u9ad4\u8207\u904b\u7b97\u6642\u9593\uff0c\u4f46\u4e5f\u56e0\u70ba\u4f7f\u7528\u4e86\u9019\u500b\u6982\u5ff5\u4f7f\u5f97 n-gram (7)</td></tr><tr><td>\u800c perplexity \u7684\u5b9a\u7fa9\u70ba \u6a21\u578b\u53ea\u80fd\u64f7\u53d6\u5230\u8996\u7a97\u4e4b\u5167\u7684\u8cc7\u8a0a\uff0c\u9577\u8ddd\u96e2\u7684\u8cc7\u8a0a\u5c31\u56e0\u6b64\u800c\u6d41\u5931\uff0c\u800c\u9019\u4e9b\u6d41\u5931\u7684\u8cc7</td></tr><tr><td>2 p H \u8a0a\u5f88\u53ef\u80fd\u6703\u9020\u6210 n-gram \u6a21\u578b\u6e2c\u8a66\u6642\u76f8\u7576\u7a0b\u5ea6\u7684\u8aa4\u5dee\uff0c\u6545\u5982\u4f55\u64f7\u53d6\u9577\u8ddd\u96e2\u7684\u8cc7\u8a0a perplexity = (8)</td></tr><tr><td>perplexity \u4ee3\u8868\u4e86 n-gram \u6a21\u578b\u4e2d\u7684\u5e73\u5747\u5206\u652f\u56e0\u6578(average branching factor)\uff0c \u4e00\u76f4\u90fd\u662f n-gram \u6a21\u578b\u4e2d\u76f8\u7576\u53d7\u5230\u77da\u76ee\u7684\u7814\u7a76\u7684\u8ab2\u984c\u3002\u4e00\u822c\u800c\u8a00\u76ee\u524d n-gram \u6a21</td></tr><tr><td>perplexity \u8d8a\u4f4e\u4ee3\u8868 n-gram \u6a21\u578b\u5728\u505a\u6a5f\u7387\u8a55\u4f30\u6642\uff0c\u6240\u9047\u5230\u7684\u5206\u652f\u8d8a\u5c11\uff0c\u4e5f\u5c31\u662f\u6b64 \u578b\u7684\u7814\u7a76\u5747\u4ee5\u89e3\u6c7a\u6b64\u4e09\u9805\u554f\u984c\u70ba\u4e3b\uff0c\u672c\u8ad6\u6587\u91dd\u5c0d\u4e0a\u8ff0\u7b2c\u4e09\u9805\u9577\u8ddd\u96e2\u8cc7\u8a0a\u7684\u64f7\u53d6\u63d0</td></tr><tr><td>1 i n i PW W \u2212 ( | i \u2212 + 1 ) \u2245 \u220f 1 T i = \u51fa\u6539\u9032\u65b9\u6cd5\uff0c\u671f\u671b\u80fd\u63d0\u6607 n-gram \u6a21\u578b\u7684\u6548\u679c\u3002 1 2 3 ( ) ( , , ,....., ) T P S PW W W W = \u6a21\u578b\u7684\u6548\u7387\u8d8a\u597d\u3002</td></tr><tr><td>2.4 n-gram \u6a21\u578b\u7684\u7f3a\u9ede 3. n-gram \u6a21\u578b\u6539\u9032\u65b9\u5411</td></tr><tr><td>n-gram \u6a21\u578b\u9577\u4e45\u5df2\u4f86\u5c31\u5b58\u5728\u8457\u4e09\u500b\u91cd\u8981\u7684\u554f\u984c\uff0c\u4e5f\u662f\u7814\u7a76 n-gram \u6a21\u578b\u7684\u4eba</td></tr><tr><td>\u4e00\u76f4\u52aa\u529b\u7684\u76ee\u6a19\uff0c\u6211\u5011\u5206\u8ff0\u5982\u4e0b:</td></tr><tr><td>1.\u8a13\u7df4\u6587\u96c6\u8207\u6e2c\u8a66\u6587\u96c6\u9818\u57df\u4e0a\u4e4b\u5dee\u8ddd(domain mismatch) :</td></tr><tr><td>n-gram \u6a21\u578b\u5728\u5efa\u7acb\u6642\uff0c\u5fc5\u9808\u8981\u6709\u4e00\u8a13\u7df4\u6587\u96c6\u4f86\u7d71\u8a08\u51fa\u6b64\u6a21\u578b\u7684\u6a5f\u7387\uff0c\u56e0\u6b64</td></tr><tr><td>n-gram \u6a21\u578b\u53d7\u5236\u65bc\u5b83\u7684\u8a13\u7df4\u6587\u96c6\uff0c\u7576\u8a13\u7df4\u6587\u96c6\u4e0d\u5e73\u5747\u6642\u53ef\u80fd\u6703\u4f7f n-gram \u6a21\u578b\u8f03</td></tr><tr><td>\u504f\u5411\u67d0\u7a2e\u9818\u57df(domain)\uff0c\u5047\u8a2d\u6211\u5011\u7684\u8a13\u7df4\u6587\u96c6\u662f\u8ca1\u7d93\u985e\u7684\u65b0\u805e\uff0c\u4f46\u662f\u6b64 n-gram \u6a21 (Trigger pair)[8][9]\u7684\u7c21\u4ecb\uff0c\u89f8\u767c\u5e8f\u5c0d\u662f\u5728\u64f7\u53d6\u9577\u8ddd\u96e2\u8cc7\u8a0a\u7684\u4e00\u7a2e\u6709\u6548\u7684\u65b9\u6cd5\uff0c\u53ef</td></tr><tr><td>\u578b\u7684\u76ee\u7684\u662f\u7528\u4f86\u6e2c\u8a66\u653f\u6cbb\u65b0\u805e\uff0c\u90a3\u9ebc\u5c31\u6703\u9020\u6210\u8f03\u5927\u7684\u8aa4\u5dee\uff0c\u5728\u9019\u65b9\u9762\u901a\u5e38\u6703\u4f7f\u7528 \u4ee5\u7528\u4f86\u88dc\u511f n-gram \u6a21\u578b\u9577\u8ddd\u96e2\u8cc7\u8a0a\u7684\u4e0d\u8db3\uff0c\u800c\u672c\u8ad6\u6587\u4e5f\u5c07\u63d0\u51fa\u4e00\u7a2e\u6539\u9032\u89f8\u767c\u5e8f</td></tr><tr><td>\u8f03\u4e00\u822c\u5316\u7684\u7684\u5e73\u8861\u6587\u96c6\u4f5c\u70ba\u8a13\u7df4\u6587\u96c6\u4f86\u89e3\u6c7a\u9019\u500b\u554f\u984c\u3002\u4f46\u77db\u76fe\u7684\u662f\u5982\u679c\u6211\u5011\u4f7f\u7528 \u5c0d\u7684\u65b9\u6cd5\u70ba\u5c0d\u7167\u7d44\uff0c\u4e26\u5728\u5be6\u9a57\u4e2d\u505a\u6bd4\u8f03\u7814\u7a76\u3002</td></tr><tr><td>\u8f03\u70ba\u5e73\u8861\u7684\u6587\u96c6\u8a13\u7df4\u51fa\u6211\u5011\u7684 n-gram \u6a21\u578b\uff0c\u6b64 n-gram \u6a21\u578b\u7528\u4f86\u6e2c\u8a66\u67d0\u4e9b\u7279\u5b9a\u9818</td></tr><tr><td>\u57df\u7684\u65b0\u805e\u662f\u5426\u6070\u7576?\u4e8b\u5be6\u4e0a\u6211\u5011\u5e0c\u671b\u5728\u6e2c\u8a66\u653f\u6cbb\u65b0\u805e\u6642\u6211\u5011\u7684 n-gram \u6a21\u578b\u662f\u504f\u5411 3.1 \u5feb\u53d6(cache)n-gram \u6a21\u578b\u8207\u6df7\u5408\u5f0f(mixture)n-gram \u6a21\u578b</td></tr><tr><td>\u653f\u6cbb\u985e\u7684\uff0c\u6e2c\u8a66\u8ca1\u7d93\u65b0\u805e\u6642 n-gram \u6a21\u578b\u662f\u504f\u5411\u8ca1\u7d93\u985e\u7684\uff0c\u70ba\u4e86\u8981\u5b8c\u6210\u9019\u9805\u9700\u6c42\uff0c \u70ba\u4e86\u8981\u4f7f n-gram \u6a21\u578b\u80fd\u5920\u66f4\u7b26\u5408\u6e2c\u8a66\u6642\u7684\u9818\u57df\uff0c\u6240\u4ee5\u7522\u751f\u4e86\u6a21\u578b\u8abf\u6574\u7684\u6982</td></tr><tr><td>\u5c31\u5fc5\u9808\u5c0d n-gram \u6a21\u578b\u518d\u505a\u6539\u9032\uff0c\u4f7f\u5176\u5177\u6709\u8abf\u6574\u4e4b\u6548\u679c[3][4][5][7]\u3002 \u5ff5\uff0c\u5b83\u7684\u6982\u5ff5\u662f\u57fa\u65bc\u3127\u7bc7\u6587\u7ae0\u6216\u662f\u4e00\u6bb5\u6587\u53e5\u6703\u6709\u4e00\u500b\u8fd1\u4f3c\u7684\u4e3b\u984c\uff0c\u6bd4\u65b9\u8aaa\u68d2\u7403\u985e</td></tr><tr><td>\u7684\u65b0\u805e\u5c31\u6bd4\u8f03\u504f\u5411\u904b\u52d5\u985e\u7684\u9818\u57df\uff0c\u8207\u5176\u4ed6\u985e\u5225\u7684\u65b0\u805e(\u5982\u8ca1\u7d93\u65b0\u805e)\u5c31\u6709\u4e00\u6bb5\u76f8\u7576</td></tr><tr><td>1 2 P WW W ( ... )log ( Q P WW W 1 2 ... ) Q \u5927\u7684\u5dee\u8ddd\uff0c\u800c\u5e0c\u671b\u80fd\u5728\u505a\u6e2c\u8a66\u6642\uff0c\u5229\u7528\u6587\u7ae0\u524d\u9762\u51fa\u73fe\u7684\u8cc7\u8a0a\uff0c\u52d5\u614b\u7684\u8abf\u6574\u6211\u5011\u7684 ... 1 lim Q Q WW W Q \u2192\u221e = \u2212 2.\u8a13\u7df4\u6587\u96c6\u4e0d\u8db3(data spareness) : \u2211 n-gram \u6a21\u578b\u5728\u8a13\u7df4\u6642\uff0c\u4e26\u4e0d\u80fd\u4fdd\u8b49\u8a13\u7df4\u6587\u96c6\u80fd\u5920\u5305\u542b\u6240\u6709\u8a5e\u7684\u7d44\u5408\uff0c\u4ee5\u81f3 \u65bc\u6240\u8a13\u7df4\u51fa\u4f86\u7684\u6a5f\u7387\u6a21\u578b\u67d0\u4e9b\u8a5e\u7d44\u76f8\u9023\u7684\u6a5f\u7387\u70ba\u96f6\uff0c\u6216\u662f\u56e0\u70ba\u8a13\u7df4\u6587\u96c6\u7684\u4e0d\u5e73 n-gram \u6a21\u578b\uff0c\u4f7f\u5f97\u6211\u5011\u7684\u6a21\u578b\u66f4\u80fd\u7b26\u5408\u6211\u5011\u6e2c\u8a66\u6587\u96c6\u7684\u9818\u57df\uff0c\u57fa\u65bc\u6b64\u7a2e\u6982\u5ff5\uff0c</td></tr><tr><td>\u5c31\u6709\u5feb\u53d6\u6a21\u578b\u8207\u6df7\u5408\u5f0f\u6a21\u578b\u7684\u6280\u8853\u7522\u751f\u3002</td></tr><tr><td>\u8861\uff0c\u9020\u6210\u7d71\u8a08\u51fa\u4f86\u7684\u6a5f\u7387\u6a21\u578b\u4e26\u4e0d\u5920\u4e00\u822c\u5316\uff0c\u800c\u70ba\u4e86\u89e3\u6c7a\u9019\u500b\u554f\u984c\uff0c\u5c31\u6709\u5e73\u6ed1\u5316</td></tr><tr><td>\u6280\u8853\u7684\u7522\u751f\uff0c\u5728\u53c3\u8003\u6587\u737b[2]\u4e2d\u5c0d\u50b3\u7d71\u4e0a\u53d7\u6b61\u8fce\u7684\u5e73\u6ed1\u5316\u6280\u8853\u6709\u8a73\u76e1\u7684\u8aaa\u660e\u3002 \u5feb\u53d6 n-gram \u6a21\u578b\u9867\u540d\u601d\u7fa9\u5c31\u662f\u76f8\u540c\u7684\u8a5e\u5e8f\u5217\u6703\u5728\u9130\u8fd1\u7684\u6642\u9593\u9ede\u4e0a\u4e0d\u65b7\u51fa</td></tr><tr><td>\u73fe\uff0c\u6bd4\u65b9\u8aaa\u6211\u5011\u7684\u6e2c\u8a66\u6587\u96c6\u662f\u4e00\u7bc7\u6709\u95dc\u91d1\u878d\u80a1\u7684\u65b0\u805e\uff0c\u4e5f\u5c31\u662f\u8aaa\u6b64\u7bc7\u6587\u4ef6\"\u91d1\u878d</td></tr></table>",
+                "html": null,
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF3": {
+                "text": "\u4ee3\u8868\u5728\u6587\u53e5 l \u7684\u8a5e\u6578\uff0cN k \u8868\u793a\u5728\u7fa4\u7d44 k \u7684\u7e3d\u6587\u53e5\u6578\u3002",
+                "content": "<table><tr><td colspan=\"10\">\u5176\u4e2d T l (13)</td></tr><tr><td colspan=\"10\">\u5176\u4e2d N \u4ee3\u8868\u8abf\u6574\u7684\u7e3d\u6587\u53e5\u6578\u3002\u6b0a\u91cd\u7684\u8abf\u6574\u7684\u4e3b\u8981\u6839\u64da\u6e2c\u8a66\u6642\u6587\u4ef6\u6240\u51fa\u73fe\u7684\u8cc7\u8a0a\uff0c</td></tr><tr><td colspan=\"10\">\u6df7\u5408\u5f0f n-gram \u6a21\u578b\u6703\u4f9d\u524d\u6587\u5728\u6bcf\u500b\u7fa4\u7d44\u6a21\u578b\u51fa\u73fe\u7684\u6a5f\u7387\u70ba\u6b0a\u91cd\uff0c\u52d5\u614b\u7684\u8abf\u6574\u6e2c</td></tr><tr><td colspan=\"10\">\u8a66\u6a21\u578b\u7684\u7d44\u5408\u6b0a\u91cd\uff0c\u6bd4\u65b9\u8aaa\u5728\u6e2c\u8a66\u6587\u4ef6\u4e2d\u4e0d\u65b7\u63d0\u5230\u91d1\u878d\u6d88\u606f\uff0c\u6df7\u5408\u5f0f n-gram \u6a21</td></tr><tr><td colspan=\"10\">\u578b\u5c31\u6703\u5c07\u6a21\u578b\u9010\u6b65\u7684\u8abf\u6574\u5230\u8ca1\u7d93\u9818\u57df\uff0c\u518d\u5229\u7528\u9019\u8abf\u6574\u904e\u5f8c\u4e4b n-gram \u6a21\u578b\u7e7c\u7e8c\u6e2c</td></tr><tr><td colspan=\"10\">\u8a66\u5f8c\u9762\u7684\u6587\u53e5\uff0c\u7136\u5f8c\u518d\u5c07\u6e2c\u8a66\u800c\u5f97\u7684\u65b0\u8cc7\u8a0a\u7e7c\u7e8c\u505a\u8abf\u6574\uff0c\u9019\u7a2e\u905e\u8ff4\u5f0f\u7684\u505a\u6cd5\u662f\u4e00</td></tr><tr><td colspan=\"7\">\u7a2e\u7a31\u70ba\u8cc7\u8a0a\u7d50\u69cb(Information structure)\u7684\u6982\u5ff5\u3002</td><td/><td/><td/></tr><tr><td>P S</td><td>=</td><td>+ = \u220f</td><td>P W W</td><td>\u2212 \u2212 +</td><td>=</td><td>ik + == \u220f\u2211</td><td>\u03bb</td><td>P W W</td><td>\u2212 \u2212 +</td></tr></table>",
+                "html": null,
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF4": {
+                "text": "\u76f8\u7368\u7acb\uff0c\u5982\u679c\u5728\u6240\u6709\u5305\u542b X \u7684\u6587\u7ae0\u4e2d\u6709 c% \u540c\u6642\u4e5f\u5305\u542b\u4e86 Y\uff0c\u5247\u6211\u5011\u53ef\u4ee5\u7a31\u95dc\u806f \u6cd5\u5247 X Y \u21d2 \u5b58\u5728\u65bc\u8cc7\u6599\u5eab D \u4e2d\u7684\u4fe1\u8cf4\u5ea6(confidence)\u70ba c\uff0c\u6b64\u5916\u82e5\u6709 s% \u7684\u6587\u7ae0\u540c \u6642\u5305\u542b X \u8207 Y\uff0c\u5247\u6211\u5011\u53ef\u7a31\u95dc\u806f\u6cd5\u5247 X Y \u21d2 \u4ee5\u652f\u6301\u5ea6(support) s \u5b58\u5728\u65bc\u8cc7\u6599\u5eab D",
+                "content": "<table><tr><td colspan=\"10\">\u7684\u8a5e\u975e\u5e38\u591a\uff0c\u6b64\u6642 unigram \u7684\u6b0a\u91cd\u53ef\u4ee5\u9069\u5ea6\u52a0\u5927\uff0c\u4ee5\u5f4c\u88dc\u53ef\u80fd\u8f03\u591a\u7684\u8cc7\u8a0a\u640d\u5931\uff0c \u6bd4\u8d77 n-gram \u6a21\u578b\u591a\u4e86\u9577\u8ddd\u96e2\u7684\u8cc7\u8a0a\uff0c\u70ba\u4e86\u65b9\u4fbf\u8d77\u898b\u4f7f\u7528\u5c0d\u6578\u8868\u793a\u70ba</td></tr><tr><td>\u4f7f\u8a9e\u8a00\u6a21\u578b\u7684\u6e96\u78ba\u6027\u63d0\u9ad8\u3002 1 log ( ) log ( ) T i i P S PW = = \u2211</td><td>+</td><td>max(1, i ws 1 j i \u2212 = \u2212 \u2211 \u2211 2 i T =</td><td>)</td><td>( M I T r i g g e rW \u2212</td><td>j</td><td>\u2192</td><td>i W</td><td>)</td><td>(19)</td></tr><tr><td colspan=\"10\">3.3 \u89f8\u767c\u5e8f\u5c0d (Trigger Pair) \u6f14\u7b97\u6cd5 \u5176\u4e2d logP(W i ) \u5373\u70ba unigram \u6a21\u578b\u6a5f\u7387\uff0cws \u4ee3\u8868 window size\uff0c\u73fe\u5728\u5728\u6211\u5011\u7684\u8ad6\u6587 \u4e2d\uff0c\u63db\u53e5\u8a71\u8aaa\uff0c\u4fe1\u8cf4\u5ea6\u662f\u4e00\u7a2e\u91cf\u6e2c\u95dc\u806f\u6cd5\u5247\u5f37\u5f31\u7684\u6a19\u6e96\uff0c\u800c\u652f\u6301\u5ea6\u5247\u662f\u8868\u793a\u7d71\u8a08</td></tr><tr><td colspan=\"10\">\u5728\u81ea\u7136\u8a9e\u8a00\u4e2d\uff0c\u5b58\u5728\u8457\u8a31\u591a\u9ad8\u5ea6\u95dc\u806f\u6027\u7684\u8a5e\u7d44\uff0c\u6bd4\u65b9\u8aaa\"\u91ab\u751f\u3001\"\u8b77\u58eb\"\u6216\u662f\"\u967d \u4e2d\u5c07 window size \u5b9a\u70ba\u6587\u53e5\u9577\u5ea6\uff0c\u4e5f\u5c31\u662f\u8aaa\u5728\u6211\u5011\u8ad6\u6587\u4e2d\u7684\u89f8\u767c\u5e8f\u5c0d\u662f\u6587\u53e5\u968e\u5c64 \u4e0a\u51fa\u73fe\u7684\u983b\u7387\uff0c\u4e8b\u5be6\u4e0a\u6211\u5011\u5be6\u4f5c\u6642\u6703\u8a02\u5b9a\u4fe1\u8cf4\u5ea6\u8207\u652f\u6301\u5ea6\u7684\u9580\u6abb\uff0c\u6211\u5011\u64f7\u53d6\u51fa\u4f86</td></tr><tr><td colspan=\"10\">\u5149\"\u3001\"\u71b1\"\u7b49\u5c31\u7d93\u5e38\u51fa\u73fe\u65bc\u65bc\u540c\u4e00\u53e5\u5b50\u4e4b\u4e2d\uff0c\u4f46\u7531\u65bc\u5b83\u5011\u901a\u5e38\u5728\u53e5\u5b50\u4e2d\u4e26\u4e0d\u76f8\u9023\uff0c \u7684\u89f8\u767c\u5e8f\u5c0d(sentence-level trigger pair) \uff0c\u4ee3\u8868\u6211\u5011\u53ea\u80fd\u64f7\u53d6\u540c\u4e00\u6587\u53e5\u4e2d\u7684\u89f8\u767c\u5e8f\u5c0d \u4e4b\u95dc\u806f\u6cd5\u5247\u7684\u4fe1\u8cf4\u5ea6\u8207\u652f\u6301\u5ea6\u5747\u5fc5\u9808\u5927\u65bc\u6b64\u9580\u6abb\u3002</td></tr><tr><td colspan=\"10\">\u6240\u4ee5 n-gram \u6a21\u578b\u4e26\u6c92\u6709\u8fa6\u6cd5\u64f7\u53d6\u5230\u9019\u4e9b\u8a5e\u4e4b\u9593\u7684\u95dc\u806f\u8cc7\u8a0a\uff0c\u56e0\u6b64\u5c31\u6709\u4e86\u89f8\u767c\u5e8f \u8cc7\u8a0a\u3002\u73fe\u5728\u6211\u5011\u5fc5\u9808\u5c07\u89f8\u767c\u5e8f\u5c0d\u52a0\u5165 n-gram \u6a21\u578b\u4e4b\u4e2d\u505a\u70ba\u9577\u8ddd\u96e2\u8cc7\u8a0a\u64f7\u53d6\u4e4b\u8f14 \u4ee5\u4e0b\u5373\u70ba\u64f7\u53d6\u95dc\u806f\u6cd5\u5247\u7684\u6f14\u7b97\u6cd5\u6d41\u7a0b\uff0c\u662f\u4ee5\u8cc7\u6599\u63a2\u52d8\u4e2d\u7684 Apriori \u6f14\u7b97\u6cd5\u505a</td></tr><tr><td colspan=\"10\">\u5c0d\u7684\u7522\u751f\uff0c\u89f8\u767c\u5e8f\u5c0d\u7684\u8a2d\u8a08\u4e3b\u8981\u5728\u65bc\u89e3\u6c7a\u9577\u8ddd\u96e2\u8cc7\u8a0a\u5f4c\u88dc n-gram \u6a21\u578b\u7684\u4e0d\u8db3\u7684 \u52a9\uff0c\u900f\u904e\u7dda\u6027\u63d2\u88dc(linear interpolation)\u7684\u65b9\u5f0f\uff0c\u6211\u5011\u53ef\u4ee5\u4e00\u6b0a\u91cd a i \u5c07\u5176\u505a\u5408\u4f75\uff0c \u70ba\u57fa\u790e\u6240\u6539\u5beb\u800c\u6210\u82e5\u6211\u5011\u4ee5\u7c21\u55ae\u7684\u4f8b\u5b50\u8aaa\u660e\u4e4b\uff0c\u5047\u8a2d\u6211\u5011\u5171\u6709\u4e09\u8a5e\uff0c\u5206\u5225\u4ee5 a\u3001</td></tr><tr><td colspan=\"10\">\u554f\u984c\uff0c\u89f8\u767c\u5e8f\u5c0d\u7531\u65bc\u5176\u6c92\u6709\u6f14\u7b97\u6cd5\u8207\u8cc7\u6599\u7d50\u69cb\u53ef\u4ee5\u5feb\u901f\u7684\u5c0d\u8cc7\u6599\u5eab\u505a\u6c42\u53d6\uff0c\u6545\u89f8 \u4e5f\u5c31\u662f b\u3001c \u4ee3\u8868\uff0cApriori \u6f14\u7b97\u6cd5\u5c31\u662f\u5728\u627e\u5c0b\u6b64\u4e09\u8a5e\u7684\u95dc\u806f\u6027\uff0c\u5b83\u7684\u6982\u5ff5\u5c31\u662f\u5148\u5c07\u9019\u4e9b</td></tr><tr><td colspan=\"10\">\u767c\u5e8f\u5c0d\u6703\u9650\u5236\u672c\u8eab\u70ba \"\u5e8f\u5c0d\" \u3001\u5373\u82e5\u6709\u4e00\u8fad\u5178 V\uff0c\u89f8\u767c\u5e8f\u5c0d\u6703\u5c0d\u5176\u4e2d\u6240\u6709\u53ef\u80fd\u7684 \u8a5e\u5e8f\u5c0d\u505a\u8003\u616e\uff0c\u5982\u6b64\u4e00\u4f86\u53ef\u5c07\u4fc3\u767c\u5e8f\u5c0d\u7684\u7e3d\u500b\u6578\u63a7\u5236\u65bc|V| 2 \u5167\u3002 1 log ( ) log ( ) k MERGED i i i P S a P S \u5176\u4e2d 0 1 i a \u2264 \u2264 \u4e14 k 1 i a = \uff0c\u5728\u9019\u908a\u6211\u5011\u6709\u5169\u500b\u6a21\u578b\u6a5f\u7387\u5b58\u5728 \u2211 = = \u22c5 \u2211 (20) \u8a5e\u5169\u5169\u70ba\u4e00\u7d44\u5efa\u7acb\u5e8f\u5c0d\u96c6\u5408(a</td></tr><tr><td>i=1</td><td/><td/><td/><td/><td/><td/><td/><td/><td/></tr><tr><td colspan=\"7\">1. P 1 (S) = P n-gram (S) \u70ba n-gram \u6a21\u578b\u5c0d\u6587\u53e5 S \u6240\u4f30\u6e2c\u51fa\u4e4b\u6a5f\u7387\u3002</td><td/><td/><td/></tr><tr><td colspan=\"9\">2. P 2 (S) = P MI-Trigger-pair (S) \u70ba\u89f8\u767c\u5e8f\u5c0d\u6a21\u578b\u5c0d\u6587\u53e5 S \u6240\u4f30\u6e2c\u51fa\u4e4b\u6a5f\u7387\u3002</td><td/></tr><tr><td colspan=\"10\">\u900f\u904e(20)\u5f0f\u7684\u8a08\u7b97\uff0c\u6211\u5011\u53ef\u4ee5\u4f7f\u7528\u89f8\u767c\u5e8f\u5c0d\u8a08\u7b97\u51fa\u4e00\u6bb5\u6587\u53e5\u7684\u6a5f\u7387\uff0c\u4e14\u6b64\u6a5f\u7387\u6709</td></tr><tr><td colspan=\"9\">( , ) ( ) ( ) i j i j P W W P W P W \u9577\u8ddd\u96e2\u8cc7\u8a0a\u5b58\u5728\uff0c\u6bd4\u8d77\u50b3\u7d71\u7684 n-gram \u6a21\u578b\u5728\u8cc7\u8a0a\u64f7\u53d6\u4e0a\u70ba\u4f73\u3002 ( , ) ( ; ) ( , )log ( , )log ( ) ( ) i j i j i j i j i j AMI W W P W W P W W P W P W = + P W W</td><td/></tr><tr><td colspan=\"10\">( , ) ( ) ( ) i j i j P W W P W P W 4. \u95dc\u806f\u6cd5\u5247\u8207\u5176\u61c9\u7528 ( , ) i j P W W + \u5728\u9019\u908a\u6211\u5011\u5f15\u5165\u4e86\u4e00\u500b\u5728\u8cc7\u6599\u63a2\u52d8(Data Ming)\u9818\u57df\u53d7\u5230\u5341\u5206\u5ee3\u6cdb\u904b\u7528\u7684 ( , ) ( , ) ( ) ( ) i j i j i j P W W P W W + P W P W</td></tr><tr><td colspan=\"10\">Apriori \u6f14\u7b97\u6cd5[1]\uff0c\u6b64\u6f14\u7b97\u6cd5\u53ef\u4ee5\u7528\u4f86\u5efa\u7acb\u95dc\u9375\u8a5e\u7684\u95dc\u806f\u6cd5\u5247\uff0c\u8209\u4f8b\u800c\u8a00\uff0c\u5047\u8a2d</td></tr><tr><td colspan=\"10\">i P W W \u4ee3\u8868\u5728\u540c\u4e00\u500b\u8996 j \u6709\u4e00\u7d44\u4ea4\u6613\u7d00\u9304\u8cc7\u6599\u5eab\uff0c\u6b64\u8cc7\u6599\u5eab\u8a18\u9304\u8457\u6bcf\u7b46\u4ea4\u6613\u6240\u5305\u542b\u7684\u5546\u54c1\uff0c\u95dc\u806f\u6cd5\u5247\u6240\u8981</td></tr><tr><td colspan=\"10\">W \u7a97\u4e2d\u53ea\u51fa\u73fe W i ,\u800c\u6c92\u51fa\u73fe W j \u7684\u6a5f\u7387\u3002\u900f\u904e AMI \u8a55\u4f30\u6a19\u6e96\uff0c\u6211\u5011\u5c07\u5176\u9078\u70ba\u89f8\u767c\u5e8f C W \u03bb \u2212 \u2212 + \u2212 + \u2212 + \u2212 + \u2212 + \u2212 + \u22c5 \u2212 \u64f7\u53d6\u7684\u5c31\u662f\u6bcf\u500b\u5546\u54c1\u9593\u7684\u76f8\u4e92\u95dc\u4fc2\uff0c\u4e5f\u5c31\u662f\u8aaa\u6211\u5011\u60f3\u77e5\u9053\u4e00\u7b46\u4ea4\u6613\u51fa\u73fe\u4e86\u67d0\u7a2e\u5546 = \u22c5 + \u2211 \u5c0d\uff0c\u4ee5\u7b26\u865f( i j W W \u2192 ) \u8868\u793a\u3002\u7576\u5e8f\u5c0d\u9078\u53d6\u5b8c\u7562\u5f8c\uff0c\u5fc5\u9808\u8981\u5c0d\u6bcf\u500b\u89f8\u767c\u5e8f\u5c0d\u8a08\u7b97\u5176 \u54c1\u5f8c\uff0c\u9084\u6709\u54ea\u4e9b\u5546\u54c1\u662f\u53ef\u80fd\u51fa\u73fe\u5728\u540c\u4e00\u7b46\u4ea4\u6613\u7d00\u9304\u4e4b\u4e2d\uff0c\u5982\u679c\u8aaa\u5546\u5bb6\u5f9e\u95dc\u806f\u6cd5\u5247</td></tr><tr><td colspan=\"10\">\u76f8\u4e92\u8cc7\u8a0a MI (mutual information)\uff0c\u7528\u5c0d\u6578\u8868\u793a\u4e4b\u5982\u4e0b \u4e2d\u77e5\u9053\u9867\u5ba2\u8cb7\u4e86\u5546\u54c1\u7532\u5f8c\uff0c\u9084\u6709\u5f88\u5927\u7684\u6a5f\u7387\u6703\u53bb\u8cb7\u5546\u54c1\u4e59\uff0c\u5247\u53ef\u5c07\u5546\u54c1\u7532\u8207\u5546\u54c1</td></tr><tr><td colspan=\"5\">( , ) ( ) ( ) i j i j P W W P W P W \u4e59\u653e\u5728\u9644\u8fd1\u589e\u52a0\u9867\u5ba2\u7684\u65b9\u4fbf\u6027\u8207\u5546\u5bb6\u7684\u696d\u7e3e\u3002 ( ; ) log i j MI W W =</td><td/><td/><td/><td/><td>(18)</td></tr><tr><td colspan=\"10\">\u5982\u679c W i \u548c W j \u662f\u76f8\u4e92\u7368\u7acb\u7684\u8a71\uff0c\u5247 MI(W i , W j ) = 0\uff0c\u76f8\u4e92\u8cc7\u8a0a\u53cd\u6620\u4e86\u89f8\u767c\u5e8f\u5c0d\u4e2d\u5169 4.1 Apriori \u6f14\u7b97\u6cd5</td></tr><tr><td colspan=\"10\">\u500b\u8a5e\u76f8\u4e92\u9593\u7684\u8cc7\u8a0a\u8b8a\u5316\u3002\u800c\u89f8\u767c\u5e8f\u5c0d\u4e26\u7121\u6cd5\u55ae\u7368\u4f7f\u7528[14]\uff0c\u56e0\u70ba\u5b83\u53ea\u80fd\u53cd\u6620\u51fa\u8a5e \u5047\u8a2d\u6211\u5011\u6709\u4e00\u7d44\u65b0\u805e\u6587\u4ef6\u8cc7\u6599\u5eab D\uff0c\u88e1\u9762\u5305\u542b\u4e86|D|\u7bc7\u6587\u7ae0\uff0c\u6bcf\u7bc7\u6587\u7ae0\u5747\u662f</td></tr><tr><td colspan=\"10\">\u8207\u8a5e\u7684\u8cc7\u8a0a\u8b8a\u5316\uff0c\u6240\u4ee5\u6211\u5011\u5fc5\u9808\u5c07\u5176\u8207 unigram \u505a\u7d50\u5408\uff0c\u5982\u6b64\u4e00\u4f86\u6240\u7372\u5f97\u7684\u8cc7\u8a0a \u8fad\u5178 1 2 { , ,......, } n L w w w = \u7684\u5b50\u96c6\u5408\uff0c\u7528\u4e0a\u9762\u7684\u4f8b\u5b50\u89e3\u91cb\u5c31\u662f\u5404\u7a2e\u5546\u54c1\u7684\u96c6\u5408\u4e4b</td></tr><tr><td colspan=\"10\">\u610f\uff0c\u800c\u95dc\u806f\u6cd5\u5247\u4ee5 X Y \u21d2 \u7684\u578b\u5f0f\u8868\u793a\uff0c\u5176\u4e2d X\u3001Y \u5747\u662f L \u7684\u5b50\u96c6\u5408(subset)\u4e14\u4e92</td></tr></table>",
+                "html": null,
+                "num": null,
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1004.json

@@ -0,0 +1,417 @@
+{
+    "paper_id": "O01-1004",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:45.371484Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O01-1004",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "\u548c k-NN \u65b9\u6cd5 [Aas, 1999] ",
+                "cite_spans": [
+                    {
+                        "start": 10,
+                        "end": 21,
+                        "text": "[Aas, 1999]",
+                        "ref_id": "BIBREF0"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Text categorization: A Survey",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Aas",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "L",
+                        "middle": [],
+                        "last": "Eikvil",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Aas, K., and Eikvil, L., \"Text categorization: A Survey\", Report No. 941, Norwegian Computing Center, June, 1999. ISBN 82-539-0425-8",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "\u2033A Corpus-based Approach to Language Learning\u2033, phD Dissertation",
+                "authors": [
+                    {
+                        "first": "Eric",
+                        "middle": [],
+                        "last": "Brill",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Eric Brill, \u2033A Corpus-based Approach to Language Learning\u2033, phD Dissertation, University of Pennsylvania, 1993.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "\u2033Hierarchical Mixtures of Experts and the EM algorithm\u2033",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [
+                            "I"
+                        ],
+                        "last": "Jordan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Jacobs",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Technical Reports A. I. Memo",
+                "volume": "1440",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Jordan, M. I., and Jacobs, R. A., \u2033Hierarchical Mixtures of Experts and the EM algorithm\u2033, Technical Reports A. I. Memo No. 1440, Massachusetts Institute of Technology, 1993",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "\u2033Hierarchical Classifying Documents Using very few Words\u2033",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Koller",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Sahami",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "ICML-1997: Proceedings of the 14 th International Conference on Machine Learning",
+                "volume": "",
+                "issue": "",
+                "pages": "170--178",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Koller, D., and Sahami, M., \u2033Hierarchical Classifying Documents Using very few Words\u2033, in ICML-1997: Proceedings of the 14 th International Conference on Machine Learning, 1997, pages 170-178.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "\u2033Reuters-21578 Text Categorization Test Collection Distribution\u2033",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Lewis",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lewis, D. D., \u2033Reuters-21578 Text Categorization Test Collection Distribution\u2033, in AT&T Labs -Research, 1996.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "\u2033Document classification on neural networks using only positive examples\u2033, ACM SIGIR",
+                "authors": [
+                    {
+                        "first": "L",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Manevitz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Yousef",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "304--306",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Manevitz, L. M., and Yousef, M., \u2033Document classification on neural networks using only positive examples\u2033, ACM SIGIR, 2000, pages 304-306.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "\u2033Feature Selection, Perceptron Learning, and a Usability Case Study for Text Categorization\u2033",
+                "authors": [
+                    {
+                        "first": "H",
+                        "middle": [
+                            "T"
+                        ],
+                        "last": "Ng",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "B"
+                        ],
+                        "last": "Goh",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Low",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the 20 th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "67--73",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ng, H. T., Goh, W. B., and Low, K. L., \u2033Feature Selection, Perceptron Learning, and a Usability Case Study for Text Categorization\u2033, in Proceedings of the 20 th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, 1997, pages 67-73.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "\u2033Combining Machine Learning and Hierarchical Indexing Structure for Text Categorization\u2033",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Ruiz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Srinivasan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of the 10 th ASIS/SIGCR Workshop on Classification Research",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ruiz, M. E., and Srinivasan, P., \u2033Combining Machine Learning and Hierarchical Indexing Structure for Text Categorization\u2033, in Proceedings of the 10 th ASIS/SIGCR Workshop on Classification Research, 1999.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "\u2033Hierarchical Neural Networks for Text Categorization\u2033",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Ruiz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Srinivasan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of the 22nd ACM SIGIR International Conference on Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "281--282",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ruiz, M. E. and Srinivasan, P., \u2033Hierarchical Neural Networks for Text Categorization\u2033, in Proceedings of the 22nd ACM SIGIR International Conference on Information Retrieval, 1999, pages 281-282.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "\u2033Learning Routing Queries in a Query Zone\u2033",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Singhal",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Mitra",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Buckley",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the 20 th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "25--32",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Singhal, A., Mitra, M., and Buckley, C., \u2033Learning Routing Queries in a Query Zone\u2033, in Proceedings of the 20 th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, 1997, pages 25-32.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Maximizing text-mining performance",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Weiss",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Apte",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Damerau",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Johnson",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Oles",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Goetz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Hampp",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "IEEE Intelligent Systems",
+                "volume": "",
+                "issue": "4",
+                "pages": "63--69",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Weiss, S.M., Apte, C., Damerau, F.J., Johnson, D.E., Oles, F.J., Goetz, T., Hampp, T., \"Maximizing text-mining performance\", IEEE Intelligent Systems, Volume: 14 Issue: 4, July-Aug, 1999, pages 63-69.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF0": {
+                "text": "jchiang@mail.ncku.edu.tw",
+                "content": "<table><tr><td>\u5f88 \u591a \u5728 \u505a \u6587 \u4ef6 \u5206 \u985e \u7684 \u65b9 \u6cd5 \u4e2d \uff0c \u4f8b \u5982 \u4f7f \u7528 \u898f \u5247 \u5eab (rule-based) \u3001 \u77e5 \u8b58 \u5eab \u5716\u4e00\u4e2d\u7684\u865b\u7dda\u7bad\u982d\u90e8\u4efd\u5247\u662f\u6574\u500b\u6e2c\u8a66\u6d41\u7a0b\uff0c\u8d77\u521d\u4e5f\u662f\u5c07\u4e00\u65b0\u6587\u4ef6\u7d93\u904e\u4e00\u9023\u4e32 (tree structure)\uff0c\u5176\u5167\u90e8\u7bc0\u9ede\u662f\u9598\u9580\u3001\u6a39\u8449\u7bc0\u9ede\u662f\u5c08\u5bb6\u3002\u5716\u4e8c\u5c31\u662f\u6211\u5011\u63d0\u51fa\u7684\u4e00\u500b \u795e\u7d93\u5143\u9593\u7684\u6b0a\u91cd\uff0c\u6b64\u51fd\u6578\u7576\u81ea\u8b8a\u6578\u8da8\u5411\u6b63\u8ca0\u7121\u9650\u5927\u6642\uff0c\u51fd\u6578\u503c\u8da8\u8fd1\u65bc\u5e38\u6578\uff0c\u5176\u51fd 4. \u5be6\u9a57\u7d50\u679c\u8207\u5206\u6790 \u5be6\u9a57\u7d50\u679c\u8207\u5206\u6790 \u5be6\u9a57\u7d50\u679c\u8207\u5206\u6790 \u5be6\u9a57\u7d50\u679c\u8207\u5206\u6790 4.2 \u7d50\u679c</td></tr><tr><td>(knowledge-based)\u3001\u6216\u6a23\u672c\u5eab(instance-based)\uff0e\uff0e\uff0e\u7b49\uff0c\u90fd\u662f\u4f9d\u8cf4\u5927\u91cf\u7684\u6a23\u672c\u4f86 \u7684\u524d\u5e8f\u8655\u7406\u5f8c\uff0c\u518d\u4f9d\u7279\u5fb5\u96c6\u8f49\u63db\u6210\u5411\u91cf\u5f62\u5f0f\uff0c\u6700\u5f8c\u900f\u904e\u968e\u5c64\u5f0f\u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44\uff0c \u4e94\u5c64\u7684\u968e\u5c64\u5f0f\u6a21\u7d44\u67b6\u69cb\u5716\u3002 \u6578\u503c\u57df\u5728[0,1]\u4e4b\u9593\u3002 4.1 \u8cc7\u6599\u96c6 \u5728\u8a55\u4f30\u6211\u5011\u7684\u6a21\u7d44\u6548\u80fd\u4e4b\u524d\uff0c\u6211\u5011\u8981\u5148\u91dd\u5c0d\u6211\u5011\u7684\u6a21\u7d44\u63d0\u51fa\u5169\u500b\u554f\u984c\uff1a1)</td></tr><tr><td>\u6c7a\u5b9a\u548c\u6587\u4ef6\u6709\u95dc\u7684\u898f\u5247\u6216\u77e5\u8b58\u3002\u4e00\u822c\u800c\u8a00\uff0c\u9019\u4e9b\u6a23\u672c\u96c6\u5408\u5fc5\u9808\u7531\u90a3\u4e9b\u5c0d\u61c9\u7528\u9818\u57df \u4ee5\u6c7a\u5b9a\u65b0\u6587\u4ef6\u6240\u5c6c\u7684\u985e\u5225\u3002 \u672c\u5be6\u9a57\u6240\u4f7f\u7528\u7684\u6e2c\u8a66\u8cc7\u6599\u96c6\uff0c\u662f\u7531 David D. Lewis [1996]\u548c\u8def\u900f\u793e\u4eba\u54e1\u6240\u5171 \u5728\u540c\u6a23\u4f7f\u7528\u985e\u795e\u7d93\u7db2\u8def\u65b9\u6cd5\u7684\u60c5\u6cc1\u4e0b\uff0c\u6709\u4f7f\u7528\u968e\u5c64\u5f0f\u67b6\u69cb\u548c\u6c92\u6709\u4f7f\u7528\u968e\u5c64\u5f0f\u67b6\u69cb</td></tr><tr><td>\u6709\u6df1\u5165\u8a8d\u8b58\u7684\u5c08\u5bb6\u4f86\u8a02\u5b9a\u8207\u5efa\u7acb\uff0c\u4e5f\u56e0\u6b64\uff0c\u9019\u4e9b\u65b9\u6cd5\u5e38\u5e38\u56e0\u70ba\u76f8\u95dc\u6a23\u672c\u5efa\u7acb\u5f97\u4e0d 3. \u7279\u5fb5\u9078\u53d6\u548c\u8a13\u7df4\u8cc7\u6599\u96c6\u9078\u53d6 \u7279\u5fb5\u9078\u53d6\u548c\u8a13\u7df4\u8cc7\u6599\u96c6\u9078\u53d6 \u7279\u5fb5\u9078\u53d6\u548c\u8a13\u7df4\u8cc7\u6599\u96c6\u9078\u53d6 \u7279\u5fb5\u9078\u53d6\u548c\u8a13\u7df4\u8cc7\u6599\u96c6\u9078\u53d6 \u540c\u6574\u7406\u800c\u6210\u7684\u8def\u900f\u793e\u65b0\u805e\u6027\u6587\u4ef6-Reuters-21578\u3002\u5728\u9019\u500b\u8cc7\u6599\u96c6\u4e2d\uff0c\u7e3d\u5171\u5305\u542b\u4e86 \u7684\u6548\u80fd\u5dee\u7570\u30022)\u6211\u5011\u6240\u63d0\u51fa\u7684\u968e\u5c64\u5f0f\u67b6\u69cb\u548c\u76ee\u524d\u5e7e\u500b\u6709\u540d\u7684\u5206\u985e\u65b9\u6cd5\u6bd4\u8f03\uff0c\u5176\u512a</td></tr><tr><td>\u8db3\u6216\u4e0d\u5b8c\u5168\uff0c\u4f7f\u5f97\u898f\u5247\u6216\u77e5\u8b58\u4e5f\u5c31\u76f8\u5c0d\u5730\u4e0d\u9f4a\u5168\uff0c\u56e0\u6b64\uff0c\u5c31\u7121\u6cd5\u5c0d\u6587\u4ef6\u505a\u5168\u76e4\u6027 \u4e00\u822c\u800c\u8a00\uff0c\u6587\u4ef6\u5927\u90e8\u4efd\u90fd\u662f\u4eba\u5011\u4ee5\u81ea\u7136\u8a9e\u8a00\u6240\u66f8\u5beb\u800c\u6210\u7684\uff0c\u9019\u4e9b\u6587\u4ef6\u4e2d\u7684\u6587 21578 \u7bc7 \u6587 \u4ef6 \uff0c \u5206 \u70ba \u4e94 \u5927 \u985e \u5225 (EXCHANGES, ORGS, PEOPLE, PLACES, \u52a3\u70ba\u4f55\uff1f</td></tr><tr><td>\u7684\u6a23\u672c\u6bd4\u5c0d\uff0c\u4ee5\u81f4\u65bc\u9020\u6210\u4e86\u5206\u985e\u4e0a\u7684\u56f0\u96e3\u3002 \u5b57\u6240\u8981\u8868\u9039\u7684\uff0c\u5247\u662f\u4eba\u5011\u7684\u60f3\u6cd5\u8207\u610f\u898b\u3002\u6211\u5011\u76f8\u4fe1\u5728\u9019\u4e9b\u60f3\u6cd5\u8207\u610f\u898b\u4e2d\uff0c\u4e3b\u8981\u662f TOPICS)\uff0c\u6211\u5011\u53ea\u62ff\u4e94\u5927\u985e\u5225\u4e2d\u7684 TOPICS \u985e\u5225\u505a\u70ba\u5be6\u9a57\u4e4b\u7528\u3002\u5728\u9019\u500b\u985e\u5225\u4e2d\uff0c \u5728\u672c\u5be6\u9a57\u4e2d\u6240\u4f7f\u7528\u7684\u8a55\u4f30\u65b9\u6cd5\uff0c\u70ba\u5728\u8cc7\u8a0a\u64f7\u53d6\u4e2d\u6700\u5e38\u88ab\u5927\u5bb6\u4f7f\u7528\u7684\u6b63\u78ba\u7387</td></tr><tr><td>\u6458\u8981 \u6458\u8981 \u6458\u8981 \u6458\u8981 \u5728\u672c\u7bc7\u8ad6\u6587\u4e2d\uff0c\u4e3b\u8981\u7684\u52d5\u6a5f\u5728\u65bc\u6539\u5584\u76ee\u524d\u6587\u4ef6\u5206\u985e\u7684\u65b9\u6cd5\uff0c\u6211\u5011\u4e0d\u4ee5\u95dc\u9375\u5b57 \u7531\u4e00\u4e9b\u91cd\u8981\u7684\u89c0\u5ff5\u6240\u7d44\u6210\u7684\uff0c\u800c\u6211\u5011\u8a8d\u70ba\u6587\u5b57\u4e2d\u7684\u540d\u8a5e\u5b57\u8a5e\u6700\u80fd\u8868\u9039\u4e00\u500b\u89c0\u5ff5\u7684 \u5305\u542b\u4e86 135 \u500b\u5b50\u985e\u5225\uff0c\u70ba\u4e86\u968e\u5c64\u5f0f\u6a21\u7d44\u7684\u8a13\u7df4\u53ca\u6e2c\u8a66\u7684\u9700\u8981\uff0c\u6211\u5011\u53ea\u9078\u64c7\u5305\u542b\u4e09 (precision)\u3001\u53ec\u56de\u7387(recall)\u548c F1 \u8a55\u4f30\u65b9\u6cd5\u3002</td></tr><tr><td>\u6587\u4ef6\u5206\u985e\u662f\u4e00\u9805\u6c7a\u5b9a\u4e00\u7bc7\u6587\u4ef6\u662f\u5426\u5c6c\u65bc\u4e00\u500b\u6216\u591a\u500b\u5df2\u4e8b\u5148\u5b9a\u7fa9\u597d\u7684\u985e\u5225\u4e4b \u7684\u5b58\u5728\u5426\u4f86\u6c7a\u5b9a\u4e00\u7bc7\u6587\u4ef6\u61c9\u5c6c\u65bc\u90a3\u4e00\u500b\u6216\u591a\u500b\u985e\u5225\u3002\u9032\u4e00\u6b65\u7684\uff0c\u6211\u5011\u63a1\u7528\u4ee5\u985e\u795e \u5f62\u6210\u3002\u56e0\u6b64\uff0c\u5728\u7279\u5fb5\u9078\u53d6\u904e\u7a0b\u4e2d\uff0c\u6211\u5011\u9996\u5148\u4f7f\u7528\u4e86\u7531 Eric Brill [1993]\u6240\u63d0\u51fa\u7684\u8a5e \u7bc7\u6587\u4ef6\u4ee5\u4e0a\u7684\u5b50\u985e\u5225\u505a\u70ba\u6e2c\u8a66\u985e\u5225\u3002\u6700\u5f8c\uff0c\u6211\u5011\u4f7f\u7528\u4e86 96 \u500b\u5b50\u985e\u5225\u300110555 \u7bc7 \u8868\u683c\u4e00\u6240\u793a\uff0c\u662f\u6211\u5011\u6240\u63d0\u51fa\u7684\u968e\u5c64\u5f0f\u65b9\u6cd5\u548c\u6c92\u6709\u4f7f\u7528\u968e\u5c64\u67b6\u69cb\u7684\u65b9\u6cd5\u7684\u6bd4\u8f03</td></tr><tr><td>\u5de5\u4f5c\uff0c\u800c\u81ea\u52d5\u5316\u5206\u985e\u5247\u53ef\u4ee5\u6709\u6548\u5730\u5e6b\u52a9\u5206\u985e\u7684\u8655\u7406\u3002\u5728\u672c\u7bc7\u8ad6\u6587\u4e2d\uff0c\u6211\u5011\u63d0\u51fa\u4e86 \u7d93\u7db2\u8def\u70ba\u57fa\u790e\u7684\u968e\u5c64\u5f0f\u67b6\u69cb\u7684\u6a5f\u5668\u5b78\u7fd2\u7684\u65b9\u6cd5\u4f86\u6c7a\u5b9a\u6587\u4ef6\u7684\u6b78\u5c6c\u3002\u800c\u4e14\uff0c\u7d93\u7531\u9019 \u6027\u5206\u6790\u5668(part-of-speech tagger)\u70ba\u6bcf\u500b\u82f1\u6587\u5b57\u6a19\u793a\u5176\u8a5e\u6027\u8cc7\u8a0a\uff0c\u7136\u5f8c\u9078\u64c7\u540d\u8a5e\u96c6 \u6587\u4ef6\u4f5c\u70ba\u5be6\u9a57\u7528\u7684\u8cc7\u6599\u96c6\u3002 [Manevitz, 2000]\uff0c\u7531\u8868\u683c\u4e2d\uff0c\u6211\u5011\u53ef\u4ee5\u5f88\u6e05\u695a\u5730\u770b\u51fa\u4f86\uff0c\u6211\u5011\u6240\u63d0\u51fa\u7684\u968e\u5c64\u5f0f\u65b9</td></tr><tr><td>\u4e00\u500b\u4ee5\u968e\u5c64\u6df7\u5408\u5f0f\u7684\u5c08\u5bb6\u6a21\u7d44(hierarchical mixture of experts model)\u70ba\u57fa\u790e\u7684\u6587 \u6a23\u5b78\u7fd2\u7684\u65b9\u6cd5\uff0c\u53ef\u4ee5\u4f7f\u6587\u4ef6\u5206\u985e\u7cfb\u7d71\u66f4\u5bb9\u6613\u5730\u61c9\u7528\u5230\u5176\u4ed6\u7684\u9818\u57df\u3002 \u5408\u7684\u95dc\u9375\u5b57\u8a5e\u3002\u63a5\u4e0b\u4f86\u5247\u5fc5\u9808\u4f7f\u7528 stop word \u904e\u6ffe\u5668\u6a21\u7d44\uff0c\u5c07\u4e0a\u8ff0\u6240\u9078\u53d6\u6a19\u793a\u540d \u5c0d\u65bc 96 \u500b\u5b50\u985e\u5225\u7684\u968e\u5c64\u67b6\u69cb\uff0c\u6211\u5011\u4f7f\u7528\u4e86 [\u9673\u5f65\u5448, 2000] \u6240\u63d0\u51fa\u7684\u67b6\u69cb \u6cd5\uff0c\u5927\u5927\u5730\u63d0\u6607\u4e86\u5206\u985e\u7684\u6b63\u78ba\u6027\u3002</td></tr><tr><td>\u4ef6\u5206\u985e\u65b9\u6cd5\u3002\u9019\u500b\u6a21\u7d44\u4f7f\u7528\u4e86\u5206\u5272-\u514b\u670d\u539f\u7406(divide-and-conquer principle)\uff0c\u5728 \u672c\u7bc7\u8ad6\u6587\u9664\u4e86\u7dd2\u8ad6\u5916\uff0c\u7b2c\u4e8c\u7bc0\u5c07\u4ecb\u7d39\u6211\u5011\u6240\u63d0\u7684\u968e\u5c64\u5f0f\u6a21\u7d44\uff0c\u7b2c\u4e09\u7bc0\u5c07\u4ecb\u7d39 \u8a5e\u7684\u95dc\u5efa\u5b57\u8a5e\u4e2d\uff0c\u904e\u6ffe\u4e00\u4e9b\u4e0d\u8db3\u4ee5\u4ee3\u8868\u6587\u4ef6\u672c\u8eab\u7279\u6027\u5b57\u8a5e\uff0c\u4ee5\u907f\u514d\u5728\u63a5\u4e0b\u4f86\u7684\u8655 \u5716\uff0c\u5176\u67b6\u69cb\u5982\u5716\u4e09\u3002\u5b83\u57fa\u672c\u7684\u5efa\u69cb\u6982\u5ff5\u662f\u4f9d\u64da\u6587\u4ef6\u5728\u5404\u985e\u5225\u4e4b\u9593\u7684\u5206\u4f48\u4f86\u5206\u6790\u985e</td></tr><tr><td>\u4e00\u500b\u4e8b\u5148\u5b9a\u7fa9\u597d\u7684\u968e\u5c64\u67b6\u69cb\u4e0b\u5b9a\u7fa9\u8f03\u5c0f\u7684\u5206\u985e\u554f\u984c\uff0c\u800c\u6700\u5f8c\u7684\u5206\u985e\u5668\u5247\u662f\u4f7f\u7528\u985e \u7279\u5fb5\u53ca\u8a13\u7df4\u6a23\u672c\u96c6\u7684\u9078\u53d6\uff0c\u7b2c\u56db\u7bc0\u5247\u91dd\u5c0d\u6211\u5011\u6240\u4f7f\u7528\u7684\u8def\u900f\u793e\u65b0\u805e\u6027\u8cc7\u6599\u96c6\u6240\u505a \u7406\u904e\u7a0b\u4e2d\uff0c\u5f15\u5165\u592a\u591a\u4e0d\u5fc5\u8981\u7684\u96dc\u8a0a(noise)\u3002\u5728\u505a\u5b8c stop word \u7684\u8655\u7406\u5f8c\uff0c\u5176\u4ed6\u5269 \u5225\u9593\u7684\u95dc\u9023\u6027\u6240\u5efa\u7acb\u8d77\u4f86\u7684\u3002 \u8868\u683c\u4e00 \u4f7f\u7528\u968e\u5c64\u5f0f\u67b6\u69cb V.S. \u6c92\u6709\u4f7f\u7528\u968e\u5c64\u5f0f\u67b6\u69cb\u7684\u5e73\u5747\u6548\u80fd\u6bd4\u8f03</td></tr><tr><td>\u795e\u7d93\u7db2\u8def\u4e2d\u7684\u5012\u50b3\u905e\u7db2\u8def\u4f86\u5b8c\u6210\u5206\u985e\u6a5f\u5236\u3002\u53e6\u5916\uff0c\u5728\u7279\u5fb5\u9078\u53d6(feature selection) \u7684\u4e00\u4e9b\u81ea\u52d5\u5316\u6587\u4ef6\u5206\u985e\u5be6\u9a57\u7684\u7d50\u679c\u8207\u5206\u6790\u3002\u6700\u5f8c\uff0c\u6211\u5011\u70ba\u672c\u7bc7\u8ad6\u6587\u63d0\u51fa\u7e3d\u7d50\u3002 \u4e0b\u7684\u540d\u8a5e\u5b57\u96c6\u9084\u4e0d\u80fd\u7b97\u662f\u6700\u5f8c\u60f3\u8981\u7684\u7279\u5fb5\u96c6\u3002\u56e0\u70ba\u6839\u64da\u4eba\u5011\u7684\u5beb\u4f5c\u7fd2\u6163\uff0c\u5c0d\u65bc\u90a3</td></tr><tr><td>\u4e0a \uff0c \u6211 \u5011 \u4e5f \u505a \u4e86 \u4e00 \u4e9b \u6709 \u5225 \u65bc \u50b3 \u7d71 \u65b9 \u6cd5 \u7684 \u6539 \u8b8a \u3002 \u6700 \u5f8c \uff0c \u6211 \u5011 \u4ee5 \u90e8 \u4efd \u8def \u900f \u793e \u4e9b\u51fa\u73fe\u983b\u7387\u592a\u904e\u65bc\u983b\u7e41\u6216\u904e\u65bc\u8ca7\u4e4f\u7684\u5b57\uff0c\u901a\u5e38\u90fd\u6c92\u6709\u592a\u5927\u7684\u7fa9\u610f\u53ca\u91cd\u8981\u6027\uff0c\u5c0d\u65bc</td></tr><tr><td>(Reuters-21578)\u7684\u65b0\u805e\u6027\u6587\u4ef6\u505a\u70ba\u6e2c\u8a66\u8cc7\u6599\uff0c\u5be6\u9a57\u7d50\u679c\u986f\u793a\u6211\u5011\u6240\u63d0\u51fa\u7684\u65b9\u6cd5\u80fd 2. \u968e\u5c64\u5f0f\u6a21\u7d44 \u968e\u5c64\u5f0f\u6a21\u7d44 \u968e\u5c64\u5f0f\u6a21\u7d44 \u968e\u5c64\u5f0f\u6a21\u7d44 \u5716\u4e8c \u672c\u8ad6\u6587\u6240\u63d0\u51fa\u4e4b\u968e\u5c64\u67b6\u69cb\u5716 \u7b26\u5408\u9019\u5169\u7a2e\u60c5\u5f62\u7684\u5b57\u96c6\uff0c\u6211\u5011\u53ef\u4ee5\u7d93\u7531\u5b57\u8a5e\u983b\u7387-\u53cd\u6587\u4ef6\u983b\u7387(term frequency and</td></tr><tr><td>\u6709\u6548\u5730\u6539\u5584\u6587\u4ef6\u5206\u985e\u7684\u6b63\u78ba\u7387\u3002 \u5716\u4e00\u6240\u793a\uff0c\u662f\u6211\u5011\u6240\u63d0\u51fa\u7684\u81ea\u52d5\u5316\u6587\u4ef6\u5206\u985e\u7684\u5b8c\u6574\u6a21\u7d44\u3002\u4e00\u500b\u6587\u4ef6\u5206\u985e\u7cfb\u7d71 \u5716\u4e00 \u672c\u8ad6\u6587\u6240\u63d0\u51fa\u4e4b\u81ea\u52d5\u5316\u6587\u4ef6\u5206\u985e\u6a21\u7d44 inverse document frequency , TFIDF)\u7684\u5206\u6790\u800c\u5c07\u5176\u904e\u6ffe\u6389\uff0c\u5982\u6b64\u8655\u7406\u5f8c\u6240\u5269\u4e0b\u7684</td></tr><tr><td>(text categorization system)\u7684\u4e3b\u8981\u5de5\u4f5c\u6d41\u7a0b\uff0c\u662f\u5148\u7528\u4e00\u7d44\u8a13\u7df4\u6a23\u672c\u96c6\u4f86\u8a13\u7df4\u7cfb\u7d71 \u5728\u6211\u5011\u7684\u6a21\u578b\u4e2d\uff0c\u6bcf\u500b\u9598\u9580\u6240\u8868\u793a\u7684\u662f\u4e00\u4efd\u6587\u4ef6\u7684\u4e00\u822c\u6982\u5ff5\uff0c\u5047\u5982\u6587\u4ef6\u4e2d\u5305 \u90e8\u4efd\uff0c\u6211\u5011\u7a31\u4e4b\u70ba\u7279\u5fb5\u5b57\u8a5e(feature words)\uff0c\u9019\u4e9b\u5b57\u8a5e\u624d\u662f\u6700\u91cd\u8981\u7684\u7cbe\u83ef\u3002</td></tr><tr><td>1. \u7dd2\u8ad6 \u7dd2\u8ad6 \u7dd2\u8ad6 \u7dd2\u8ad6 \u4e2d\u7684\u6587\u4ef6\u5206\u985e\u5668\uff1b\u7136\u5f8c\u518d\u85c9\u7531\u5df2\u8a13\u7df4\u597d\u7684\u5206\u985e\u5668\u5c0d\u6e2c\u8a66\u6a23\u672c\u4e2d\u7684\u65b0\u6587\u4ef6\u505a\u81ea\u52d5\u5316 \u5728\u5716\u4e00\u7528\u865b\u7dda\u65b9\u584a\u6240\u570d\u6210\u7684\uff0c\u5c31\u662f\u6211\u5011\u6240\u63d0\u51fa\u7684\u968e\u5c64\u5f0f\u985e\u795e\u7d93\u7db2\u8def\u6a21\u7d44\uff0c\u5176 \u542b\u8457\u6240\u8868\u793a\u7684\u6982\u5ff5\uff0c\u5247\u7db2\u8def\u7684\u8f38\u51fa\u662f 1\uff0c\u5426\u5247\u70ba 0\u3002\u800c\u5c08\u5bb6\u6240\u8868\u793a\u7684\u662f\u7279\u5b9a\u7684\u985e \u6b64\u5916\uff0c\u5728\u8a13\u7df4\u5206\u985e\u5668\u65b9\u9762\uff0c\u5c0d\u65bc\u540c\u4e00\u985e\u5225\u7684\u6b63\u8ca0\u8a13\u7df4\u6a23\u672c\u9078\u53d6\u4e0a\uff0c\u82e5\u5169\u8005\u7684</td></tr><tr><td>\u8fd1\u5e7e\u5e74\u4f86\uff0c\u96a8\u8457\u7db2\u8def\u6280\u8853\u4e0d\u65b7\u5730\u9032\u6b65\uff0c\u6709\u7528\u7684\u8cc7\u8a0a\u4e5f\u76f8\u5c0d\u5730\u5927\u91cf\u6210\u9577\u4e2d\u3002\u96d6 \u5206\u985e\u7684\u52d5\u4f5c\u3002\u5728\u5716\u4e00\u7684\u5be6\u7dda\u7bad\u982d\u90e8\u4efd\u662f\u6574\u500b\u6587\u4ef6\u5206\u985e\u7684\u8a73\u7d30\u8a13\u7df4\u904e\u7a0b\uff0c\u9996\u5148\u6c7a\u5b9a \u8a73\u7d30\u7684\u67b6\u69cb\u5982\u5716\u4e8c\uff0c\u6b64\u6a21\u7d44\u7684\u4e3b\u8981\u7684\u9748\u611f\u662f\u4f86\u81ea\u65bc Jordan \u548c Jacobs[1993]\u6240\u63d0\u51fa \u5225[Ruiz, 1999]\u3002\u6240\u6709\u7684\u6587\u4ef6\u90fd\u4ee5\u5411\u91cf\u8868\u793a\u4e4b\u3002\u6574\u500b\u5206\u985e\u5de5\u4f5c\u662f\u7531\u6839\u7bc0\u9ede(root node) \u9078\u53d6\u5dee\u8ddd\u904e\u5927\uff0c\u9020\u6210\u904e\u5ea6\u5730\u4e0d\u5e73\u5747\uff0c\u5f88\u6709\u53ef\u80fd\u6703\u9020\u6210\u5206\u985e\u5668\u5728\u5b78\u7fd2\u4e0a\u7684\u8aa4\u5dee\uff0c\u4ee5</td></tr><tr><td>\u7136\u7db2\u8def\u4e0a\u8209\u624b\u53ef\u5f97\u7684\u8cc7\u8a0a\u65b9\u4fbf\u4eba\u5011\u5c0d\u8cc7\u8a0a\u7684\u53d6\u5f97\u8207\u50b3\u905e\uff0c\u4f46\u662f\u7576\u7db2\u8def\u8cc7\u8a0a\u91cf\u6108\u4f86 \u4e00\u7d44\u5df2\u7531\u5c08\u5bb6\u5206\u985e\u597d\u7684\u6a23\u672c\u96c6\uff0c\u5f9e\u6b64\u6a23\u672c\u96c6\u4e2d\uff0c\u7d93\u904e\u4e00\u9023\u4e32\u7684\u524d\u8655\u7406\u7a0b\u5e8f\u5f8c\uff0c\u9078 \u958b\u59cb\uff0c\u5047\u5982\u9598\u9580\u7684\u8f38\u51fa\u503c\u70ba\u771f\uff0c\u5247\u7b2c\u4e8c\u5c64\u7684\u7bc0\u9ede\u90fd\u6703\u88ab\u555f\u52d5\uff0c\u5982\u6b64\u7684\u7a0b\u5e8f\u6301\u7e8c\u81f3 \u7684\u968e\u5c64\u5f0f\u6df7\u5408\u7684\u5c08\u5bb6\u6a21\u578b(hierarchical mixture of experts , HME model)\u3002HME \u6a21 \u81f4\u65bc\u9020\u6210\u6700\u5f8c\u5206\u985e\u4e0a\u7684\u932f\u8aa4\u3002\u56e0\u6b64\uff0c\u5c0d\u65bc\u8a13\u7df4\u6a23\u672c\u7684\u9078\u53d6\u4e5f\u662f\u4e0d\u53ef\u5ffd\u8996\u7684\u5de5\u4f5c\u4e4b</td></tr><tr><td>\u6108\u5927\u6642\uff0c\u5982\u4f55\u6709\u6548\u3001\u4e14\u5feb\u901f\u5730\u53d6\u5f97\u6709\u7528\u7684\u8cc7\u8a0a\uff0c\u4fbf\u6210\u70ba\u975e\u5e38\u91cd\u8981\u7684\u4e8b\u60c5\u3002\u6b64\u6642\uff0c \u64c7\u4e00\u7d44\u6700\u80fd\u4ee3\u8868\u53ca\u8b58\u5225(identification)\u6b64\u985e\u5225\u7684\u7279\u5fb5\u96c6(feature set)\u3002\u4e26\u4ee5\u5411\u91cf\u65b9\u5f0f \u5f0f\u662f\u4ee5\u5206\u5272-\u514b\u670d\u539f\u7406(divide-and-conquer principle)\u70ba\u57fa\u790e\uff0c\u5176\u4e3b\u8981\u7684\u60f3\u6cd5\u662f\u5c07\u4e00 \u5b83\u5230\u9054\u4e00\u500b\u6a39\u8449\u7bc0\u9ede\u3002 \u4e00\u3002\u5728\u9019\u4e00\u65b9\u9762\uff0c\u6211\u5011\u63a1\u7528\u4e86\u7531 Ruiz [1999]\u6240\u63d0\u51fa\u7684\u2033\u985e\u5225\u5340(category zone)\u2033\u7684</td></tr><tr><td>\u6587\u4ef6\u5206\u985e(text categorization)\u6280\u8853\uff0c\u5373\u900f\u904e\u6f14\u7b97\u6cd5\u5206\u6790\u4e00\u96fb\u5b50\u6587\u4ef6\u5f8c\uff0c\u5c07\u5176\u5206\u914d \u8868\u793a\u4e4b\uff0c\u5982\u6b64\u5c31\u53ef\u5f97\u5230\u4e00\u500b\u4ee5\u7279\u5fb5\u5411\u91cf\u8868\u793a\u7684\u6a23\u672c\u7d44\uff0c\u800c\u5728\u968e\u5c64\u5f0f\u985e\u795e\u7d93\u7db2\u8def\u6a21 \u500b\u5927\u554f\u984c\u5206\u5272\u6210\u82e5\u5e72\u500b\u5bb9\u6613\u89e3\u6c7a\u7684\u5c0f\u554f\u984c\uff0c\u7136\u5f8c\u518d\u7d50\u5408\u9019\u4e9b\u5c0f\u554f\u984c\u7684\u89e3\u7b54\uff0c\u4ee5\u5f97 \u5c0d\u65bc\u9598\u9580\u548c\u5c08\u5bb6\u7db2\u8def\uff0c\u7531\u65bc\u985e\u795e\u7d93\u7db2\u8def\u4e2d\u7684\u5012\u50b3\u905e\u7db2\u8def(back-propagation , \u6982\u5ff5\u4f86\u9078\u53d6\u8a13\u7df4\u6a23\u672c\u96c6\uff0c\u5176\u57fa\u672c\u505a\u6cd5\u70ba\u9078\u53d6\u5c6c\u65bc\u6b64\u985e\u5225\u7684\u6587\u4ef6\u70ba\u6b63\u6a23\u672c\uff0c\u800c\u9078\u53d6 \u8868\u683c\u4e8c\u6240\u793a\uff0c\u5247\u662f\u6211\u5011\u6240\u63d0\u51fa\u7684\u65b9\u6cd5\u548c\u5169\u500b\u8457\u540d\u7684\u5206\u985e\u65b9\u6cd5\u7684\u6bd4\u8f03-\u6c7a\u7b56\u6a39</td></tr><tr><td>(assign)\u7d66\u4e00\u6216\u591a\u500b\u985e\u5225(categories)\uff0c\u4fbf\u626e\u6f14\u8457\u5176\u4e2d\u91cd\u8981\u7684\u89d2\u8272\u3002 \u7d44\u4e2d\uff0c\u4e3b\u8981\u662f\u5e0c\u671b\u80fd\u900f\u904e\u6bcf\u4e00\u500b\u6a23\u672c\u7d44\u4f86\u8a13\u7df4\u5176\u6240\u5c6c\u7684\u5206\u985e\u5668\uff0c\u4f7f\u5176\u80fd\u5f88\u6b63\u78ba\u5730 \u5230\u4e00\u822c\u5316\u7684\u89e3\u7b54\u3002\u800c\u5728\u5206\u985e\u4e00\u500b\u6e1b\u5c11\u7bc4\u570d\u4e0a(reduced domain)\uff0cHME \u6a21\u578b\u662f\u7d93\u7531 BP Network)\u5177\u6709\u5b78\u7fd2\u6b63\u78ba\u7387\u9ad8\u3001\u7406\u8ad6\u7c21\u660e[Zurada, 1992]\u3002\u56e0\u6b64\uff0c\u6211\u5011\u6c7a\u5b9a\u4f7f\u7528 \u6700\u9760\u8fd1\u6b64\u985e\u5225\u3001\u537b\u4e0d\u5c6c\u65bc\u6b64\u985e\u5225\u7684\u6587\u4ef6\u505a\u70ba\u8ca0\u6a23\u672c\u3002\u9019\u6a23\u7684\u6982\u5ff5\uff0c\u6700\u65e9\u662f\u4f86\u81ea\u65bc (decision tree)</td></tr><tr><td>\u50b3\u7d71\u7684\u6587\u4ef6\u5206\u985e\u5de5\u4f5c\u90fd\u662f\u7531\u67d0\u500b\u9818\u57df\u7684\u4eba\u985e\u5c08\u5bb6(human experts in domain) \u5c07\u6bcf\u4e00\u500b\u6a23\u672c\u5206\u5230\u6b63\u78ba\u7684\u985e\u5225\u53bb\u3002\u7d93\u904e\u4e00\u9023\u4e32\u7684\u53cd\u8986\u5b78\u7fd2\u5f8c\uff0c\u6211\u5011\u5f97\u5230\u4e00\u7d44\u5df2\u8a13 \u5c07\u8f38\u5165\u7a7a\u9593(input space)\u5283\u5206\u6210\u4e00\u5de2\u72c0\u3001\u9806\u5e8f\u7684\u5340\u57df\uff0c\u7136\u5f8c\u8a13\u7df4\u7279\u5b9a\u7684\u8f03\u5c0f\u5206\u985e \u4e09\u5c64\u7684\u5012\u50b3\u905e\u985e\u795e\u7d93\u7db2\u8def\uff0c\u5176\u8f38\u5165\u5c64\u5305\u542b\u4e86 N \u500b\u7279\u5fb5\uff0c\u96b1\u85cf\u5c64\u5305\u542b\u4e86(2N/3)\u500b\u7bc0 Singhal \u7b49\u4eba[1997]\u70ba\u6587\u4ef6\u7e5e\u9001(text routing)\u6240\u63d0\u51fa\u4f86\u7684\u60f3\u6cd5\u3002</td></tr><tr><td>\u6240\u5b8c\u6210\u3002\u4f46\u662f\uff0c\u96a8\u8457\u6587\u4ef6\u6578\u91cf\u5feb\u901f\u5730\u6210\u9577\uff0c\u5c0d\u65bc\u5c08\u5bb6\u800c\u8a00\uff0c\u9019\u6a23\u7684\u5de5\u4f5c\u5c31\u8b8a\u5f97\u66f4 \u56f0\u96e3\u4e86\u3002\u5728\u9019\u7a2e\u60c5\u6cc1\u4e0b\uff0c\u6587\u4ef6\u7684\u81ea\u52d5\u5206\u985e\u5c31\u986f\u5f97\u66f4\u52a0\u91cd\u8981\u4e86\u3002 \u5668\uff0c\u4ee5\u6b64\u6c42\u5f97\u4e00\u500b\u5206\u985e\u554f\u984c\u7684\u7b54\u6848\u3002HME \u6a21\u578b\u5305\u542b\u5169\u500b\u57fa\u672c\u7684\u5143\u4ef6\uff1a\u9598\u9580\u7db2\u8def \u9ede\uff0c\u800c\u8f38\u51fa\u5c64\u70ba\u55ae\u4e00\u500b\u7bc0\u9ede\u3002\u800c\u5728\u795e\u7d93\u5143\u7684\u67b6\u69cb\u4e2d\uff0c\u6211\u5011\u4f7f\u7528 S \u5f62\u51fd\u6578(sigmoid \u7df4\u597d\u3001\u5177\u6709\u76f8\u7576\u8fa8\u8b58\u7a0b\u5ea6\u7684\u5206\u985e\u5668\uff0c\u4ee5\u4f9b\u6e2c\u8a66\u968e\u6bb5\u6642\u4f7f\u7528\u3002 (gating networks)\u548c\u5c08\u5bb6\u7db2\u8def(expert networks)\u3002\u9019\u4e9b\u5143\u4ef6\u7684\u7d50\u69cb\u985e\u4f3c\u65bc\u6a39\u72c0\u7d50\u69cb function)\u4f5c\u70ba\u8f49\u63db\u51fd\u6578\u3002\u6b64\u51fd\u6578\u5177\u6709\u5fae\u5206\u5bb9\u6613\u7684\u512a\u9ede\uff0c\u53ef\u914d\u5408\u964d\u68af\u5ea6\u6cd5\u5247\u4f86\u8abf\u6574 \u5716\u4e09 \u5728 TOPICS \u4e2d\uff0c96 \u500b\u5b50\u985e\u5225\u7684\u968e\u5c64\u5f0f\u67b6\u69cb\u5716</td></tr></table>",
+                "html": null,
+                "type_str": "table",
+                "num": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1005.json

@@ -0,0 +1,823 @@
+{
+    "paper_id": "O01-1005",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:43.503514Z"
+    },
+    "title": "Using Chi-square Testing in Modeling Confusion Characteristics for Robust Phonetic Set Generation",
+    "authors": [
+        {
+            "first": "Yeou-Jiunn",
+            "middle": [],
+            "last": "Chen",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Industrial Technology Research Institute",
+                "location": {
+                    "settlement": "Hsinchu",
+                    "country": "Taiwan, R.O.C. ("
+                }
+            },
+            "email": "chenyj@itri.org.tw"
+        },
+        {
+            "first": "Chung-Hsien",
+            "middle": [],
+            "last": "Wu",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Cheng Kung University",
+                "location": {
+                    "settlement": "Tainan",
+                    "country": "Taiwan, R.O.C"
+                }
+            },
+            "email": "chwu@csie.ncku.edu.tw"
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "A phonetic representation of a language is used to describe the corresponding pronunciation and synthesize the acoustic model of any vocabulary. In order to obtain better phonetic representation, context-dependent units are used to model co-articulation effects between phones and have been broadly in speech recognition. However, this representation generally increases the number of recognition units. A phonetic representation with smaller phonetic units such as SAMPA-C for Mandarin Chinese can be applied to reduce the number of recognition units. Nevertheless, smaller phonetic units such as SAMPA-C will contain confusion characters and generally degrade the recognition performance. In this paper, a statistical method based on chi-square testing is used to investigate the confusion characteristics among phonetic units and develop a more reliable phonetic set, named modified SAMPA-C. Finally, experiments on continuous Mandarin telephone speech recognition were conducted. Experimental results show an encouraging improvement on recognition performance can be obtained. In addition, the proposed approaches represent a good compromise between the demands of accurate acoustic modeling.",
+    "pdf_parse": {
+        "paper_id": "O01-1005",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "A phonetic representation of a language is used to describe the corresponding pronunciation and synthesize the acoustic model of any vocabulary. In order to obtain better phonetic representation, context-dependent units are used to model co-articulation effects between phones and have been broadly in speech recognition. However, this representation generally increases the number of recognition units. A phonetic representation with smaller phonetic units such as SAMPA-C for Mandarin Chinese can be applied to reduce the number of recognition units. Nevertheless, smaller phonetic units such as SAMPA-C will contain confusion characters and generally degrade the recognition performance. In this paper, a statistical method based on chi-square testing is used to investigate the confusion characteristics among phonetic units and develop a more reliable phonetic set, named modified SAMPA-C. Finally, experiments on continuous Mandarin telephone speech recognition were conducted. Experimental results show an encouraging improvement on recognition performance can be obtained. In addition, the proposed approaches represent a good compromise between the demands of accurate acoustic modeling.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "From the viewpoint of speech recognition, a phonetic representation is functionally defined by the mapping of the fundamental phonetic units of a language to describe the corresponding pronunciation and synthesize the acoustic model of any vocabulary. In the past years, contextdependent units have been broadly used to model the co-articulation effects such as triphone models, which consider both left and right phonemes at the same time. However, this representation generally increases the number of recognition units. Approaches for designing a smaller number of phonetic units are needed in the context-dependent based recognition.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In recent years, many phoneme-based phonetic representations have been used such as International Phonetic Alphabet (IPA) [1] , Speech Assessment Methods Phonetic Alphabet (SAMPA) [2] , and SAMPA for Chinese (SAMPA-C) [3] . Among these representations, SAMPA-C is more flexible and consistent than other phoneme-based phonetic representations for Mandarin Chinese. However, in SAMPA-C, several phonetic units with short duration are not easy to be distinguished and therefore degrade the recognition performance.",
+                "cite_spans": [
+                    {
+                        "start": 122,
+                        "end": 125,
+                        "text": "[1]",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 180,
+                        "end": 183,
+                        "text": "[2]",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 218,
+                        "end": 221,
+                        "text": "[3]",
+                        "ref_id": "BIBREF2"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "For Mandarin speech, the confusion characteristics can be found and analyzed in syllabledependent, subsyllable-dependent, or phoneme-dependent situation. In a training database, syllabledependent confusion characteristics are difficult to extract due to the sparse data problem. In contrast, the inconsistent phoneme segment in the training data is also not suitable to detect the phone-dependent confusion characteristics. The misdetected phones will result in misrecognition of syllables/subsyllables. Consequently, the phone-dependent confusion characteristic is not helpful for the analysis and representation of confusion characteristics of SAMPA-C based Mandarin speech recognizer. Therefore, the subsyllable is chosen as a compromising unit for the analysis of subsyllable-dependent confusion characteristics.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, based on the statistical hypothesis, the \u03c7 2 (chi-square) testing [4] is an alternative test for evaluating dependence, which does not assume normally distributed probabilities. The underlying principle is to compare the observed frequencies with the expected frequencies. For investigating the effects of the confusion characteristics, the \u03c7 2 statistic is used to examine the consistencies of two probabilistic distributions and the statistical decision criteria are applied to evaluate the statistical evidence for the confusion degree of two subsyllables. According to the analysis result, a less confusable phonetic set, namely modified SAMPA-C, is applied to develop a new Mandarin speech recognizer and compared to the original SAMPA-C.",
+                "cite_spans": [
+                    {
+                        "start": 81,
+                        "end": 84,
+                        "text": "[4]",
+                        "ref_id": "BIBREF3"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "The architecture for constructing the recognition model is shown in Fig. 1 and can be divided into two processes: development process and evaluation process. In the development process, an acoustic training database is collected and classified statistically for establishing SAMPA-C based recognition models. By analyzing the output distributions of confusion models, the confusion characteristics are extracted and used to generate the modified SAMPA-C. Moreover, using decision tree, the context-dependent models are generated for evaluating the performance. In the evaluation process, two continuous Mandarin speech recognition systems are developed and used to evaluate the syllable recognition rates using SAMPA-C and modified SAMPA-C HMM-based recognition models, respectively. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 68,
+                        "end": 74,
+                        "text": "Fig. 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "To accurately represent the confusion characteristics of Mandarin speech, the subsyllables are used as the basic units in the analysis process and extracted from the recognition outputs generated by the SAMPA-C based syllable recognizer. In this analytic procedure, 50 context-independent leftto-right HMMs with 4 states and 12 mixtures are built as the basic recognition models. 1551 utterances generated by 80 speakers in Mandarin Speech Database Across Taiwan (MAT) are used for advance analysis. In the following tests, the effect of the confusion characteristics between every two subsyllables is considered.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Analysis of Confusion Characteristics",
+                "sec_num": "2."
+            },
+            {
+                "text": "To clarify the subsyllable-dependent confusion characteristics, the training and misrecognized data are used and divided into several categories, which are defined as subsyllable attributes (SA).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "For each SA, the numbers of occurrences and misrecognitions generated by the recognizer are accumulated. Then, these two corresponding frequency distributions of the training and misrecognized data, treated as SA distributions, are utilized to quantitatively analyze the confusion degree by using the \u03c7 2 testing. The \u03c7 2 value, which is greater than a threshold of the predefined significance level, implies that the SA distributions can be regarded as different. Accordingly, several subsyllables are treated as confusable and need further discrimination. The formula to calculate the \u03c7 2 value is defined as follows.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "( )( ) 2 1 N i i i i i i M E M E E \u03c7 = \u2212 \u2212 = \u2211 (1)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "where N is the number of SAs, M i is the number of misrecognitions of the i-th SA and E i is the expected value of M i and can be defined as",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "1 1 N j j i i N j j M E W W = = = \u2211 \u2211 (2)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "where W i is the number of appearances of the i-th SA.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "The effects of confusion characteristics are analyzed and extracted from the recognition outputs generated by the SAMPA-C based syllable recognizer. Table I and Table II show two SA distributions of INITIALs and FINALs represented by SAMPA-C, respectively. It is clear that \"d\" and \"V:\" has the largest number of appearances in INITIALs and FINALs. However, the tendency of \"dC\" and \"IM\" was misrecognized frequently more than that of \"d\" and \"V:\", respectively. \"dC\", \"IM\", \"d\", and \"V:\" are the Mandarin syllables represented by SAMPA-C. As a result for a Mandarin speech recognizer, the confusion characteristics seems to strongly depend on the subsyllables. Next, since insufficient training data happen for some SAs, the \u03c7 2 testing conditions might not be satisfied. Thus, the following two conditions in each SA have to be considered [5] .",
+                "cite_spans": [
+                    {
+                        "start": 841,
+                        "end": 844,
+                        "text": "[5]",
+                        "ref_id": "BIBREF4"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 149,
+                        "end": 169,
+                        "text": "Table I and Table II",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "(1) The percentage of the expected value over five is above 80%.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "(2) All expected values are more than one.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "In Table I and Table II ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 3,
+                        "end": 23,
+                        "text": "Table I and Table II",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "2-1 Testing for subsyllable-dependent confusion characteristics",
+                "sec_num": null
+            },
+            {
+                "text": "According to the previous analysis, the misrecognition happens in some specific SAs. In general, the misrecognition is caused by the incorrect pronunciation or the confusable phonetic set.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-2 Examination of confusable phonetic set",
+                "sec_num": null
+            },
+            {
+                "text": "The incorrect pronunciation is due to inarticulacy such as the retroflexion in Mandarin speech. For examples, the \"tS\" and \"IN\" is usually pronounced as \"ts\" and \"IM\" in INITIALs and FINALs, respectively. Thus, in this paper, the confusion characteristic of each recognition units in the SAMPA-C based recognizer has to be examined and the phonetic set should be redefined. Table III shows some examples of SA distributions of confusions for recognition units in SAMPA-C. The upper two measures show the \u03c7 2 values are greater than 5% of the significance level and the phoneme will cause the subsyllable-dependent confusion according to the \u03c7 2 testing. On the other hand, the lower two measures show the \u03c7 2 values are smaller than 5% of the significance level and these subsyllables possess less confusion characteristic. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 374,
+                        "end": 383,
+                        "text": "Table III",
+                        "ref_id": "TABREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "2-2 Examination of confusable phonetic set",
+                "sec_num": null
+            },
+            {
+                "text": "Given a subsyllable A, the subsyllable-dependent confusion characteristic between subsyllables A and B can be analyzed in Table IV , which show the four possible outcomes for a given trial. The confusion relationship between subsyllables A and B can be shown in Fig. 2 .",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 122,
+                        "end": 130,
+                        "text": "Table IV",
+                        "ref_id": "TABREF3"
+                    },
+                    {
+                        "start": 262,
+                        "end": 268,
+                        "text": "Fig. 2",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "2-3 Determination of confusable phones",
+                "sec_num": null
+            },
+            {
+                "text": "According to this representation, the \u03c7 2 testing serves as a way to quantify the confusion between these two distributions. Hence, based on the four outcomes in Table IV , the \u03c7 2 testing can be applied to determine the degree of confusion between subsyllables A and B and is given by",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 162,
+                        "end": 170,
+                        "text": "Table IV",
+                        "ref_id": "TABREF3"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "2-3 Determination of confusable phones",
+                "sec_num": null
+            },
+            {
+                "text": "( )( ) 2 ij ij ij ij cells ij f E f E E \u03c7 \u2212 \u2212 = \u2211 (3)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-3 Determination of confusable phones",
+                "sec_num": null
+            },
+            {
+                "text": "where f ij is the observed frequency. E ij is the expected frequency and defined as",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-3 Determination of confusable phones",
+                "sec_num": null
+            },
+            {
+                "text": "0 0 2 0 1 j ij i k k f E f f = = \u2211 (4)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-3 Determination of confusable phones",
+                "sec_num": null
+            },
+            {
+                "text": "where f i0 is the totals of the i-th row and f 0j is the totals of the j-th column. If the value in Table IV is small, Yate's correction method is used to estimate a robust \u03c7 2 value [6] . Therefore, the confusable phone, which causes the subsyllable-dependent confusion, can be found. Table V shows some examples of confusion measure. In this table (a) and (b) have high confusion contrast to (c) and (d).",
+                "cite_spans": [
+                    {
+                        "start": 183,
+                        "end": 186,
+                        "text": "[6]",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 100,
+                        "end": 108,
+                        "text": "Table IV",
+                        "ref_id": "TABREF3"
+                    },
+                    {
+                        "start": 286,
+                        "end": 293,
+                        "text": "Table V",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "2-3 Determination of confusable phones",
+                "sec_num": null
+            },
+            {
+                "text": "Accordingly, subsyllable \"d+C\" and subsyllable \"U+N\" are likely confused with subsyllable \"C\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-3 Determination of confusable phones",
+                "sec_num": null
+            },
+            {
+                "text": "and subsyllable \"i+U+N\", respectively.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-3 Determination of confusable phones",
+                "sec_num": null
+            },
+            {
+                "text": "Significance Level In the first experiment, the SAMPA-C based recognizer and the modified SAMPA-C based recognizer were built for the comparison of recognition performance. In these systems, the contextindependent models were adopted and the subsyllable recognition rates of INITIALs and FINALs for the two systems are listed in Table VII . duration is easy to be misrecognized. Therefore, the confusion between INITIALs can be discriminated using the modified SAMPA-C and the recognition performance can be improved significantly.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 329,
+                        "end": 338,
+                        "text": "Table VII",
+                        "ref_id": "TABREF4"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "P(R A |O A ) A B",
+                "sec_num": null
+            },
+            {
+                "text": "Moreover, another phonetic representation set is also developed for evaluating the confusion characteristics analysis. This phonetic representation with 58 fundamental subsyllables [7] [8] [9] was adopted in this experiment. With the same training database, the distributions of misrecognition for subsyllable \"dC\", \"C\", \"dZ\", and \"d\" are shown in Fig. 3 . The subsyllable \"dC\" is usually misrecognized to \"C\". However, the subsyllable \"C\" is not usually misrecognized to \"dC\". It is difficult to detect the confusion characteristic of subsyllable \"dC\". In our approach, the \u03c7 2 value of \"dC\" compared with other subsyllables is shown in Fig. 4 . The confusion characteristic of subsyllable \"dC\" can be detected. For the significance level, the subsyllable \"C\" usually confused with subsyllable \"dC\".",
+                "cite_spans": [
+                    {
+                        "start": 181,
+                        "end": 184,
+                        "text": "[7]",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 185,
+                        "end": 188,
+                        "text": "[8]",
+                        "ref_id": "BIBREF7"
+                    },
+                    {
+                        "start": 189,
+                        "end": 192,
+                        "text": "[9]",
+                        "ref_id": "BIBREF8"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 348,
+                        "end": 354,
+                        "text": "Fig. 3",
+                        "ref_id": "FIGREF1"
+                    },
+                    {
+                        "start": 638,
+                        "end": 644,
+                        "text": "Fig. 4",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "P(R B |O A ) P(R A |O B ) P(R B |O B )",
+                "sec_num": null
+            },
+            {
+                "text": "In the next experiment, the context-dependent models were applied for evaluation and the experimental results are shown in Fig. 5 . It is clear that the modified SAMPA-C can achieve an encouraging recognition performance, which is better than that obtained using the SAMPA-C.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 123,
+                        "end": 129,
+                        "text": "Fig. 5",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "P(R B |O A ) P(R A |O B ) P(R B |O B )",
+                "sec_num": null
+            },
+            {
+                "text": "Especially, for the context-dependent models, the confusion between syllables can be efficiently discriminated and the recognition performance can also be improved. Table VIII shows the experimental results and the modified SAMPA-C based approach outperformed the other two types. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 165,
+                        "end": 175,
+                        "text": "Table VIII",
+                        "ref_id": "TABREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "P(R B |O A ) P(R A |O B ) P(R B |O B )",
+                "sec_num": null
+            },
+            {
+                "text": "In this paper, the confusion characteristics for Mandarin speech using SAMPA-C were analyzed. The confusion characteristics generated with respect to confusable phonetic set can be discriminated by incorporating a statistical categorical data analysis method without any model assumption. Redefining the phonetic set, the effect of the confusion characteristics can be reduced and the recognition performance can be improved significantly. Hence, a modified SAMPA-C is proposed to provide a corresponding phonetic representation for building more reliable recognition models. Experimental results show that the proposed approaches give an encouraging improvement.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions",
+                "sec_num": "5."
+            },
+            {
+                "text": "For the portability to other languages, the proposed procedure can be easily applied to detect the confusion phonetic units of that language. Accordingly, a more reliable phonetic set for that language can be obtained.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions",
+                "sec_num": "5."
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Mathews' Chinese-English Dictionary, Caves, 13th printing",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Mathews",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1975,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R. H. Mathews, Mathews' Chinese-English Dictionary, Caves, 13th printing, 1975.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "EAGGLES Handbook on Spoken Language Systems(DRAFT) -SAMPA computer readable phonetic alphabet",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Wells",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "J. Wells, EAGGLES Handbook on Spoken Language Systems(DRAFT) -SAMPA computer readable phonetic alphabet, <http:// www.phon.ucl.ac.uk/home/sampa/home.htm>, 1997.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Phonetic modeling in the Philips Chinese continuous-speech recognition system",
+                "authors": [
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Seide",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "N",
+                        "middle": [
+                            "J C"
+                        ],
+                        "last": "Wang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proc. of ISCSLP'98",
+                "volume": "",
+                "issue": "",
+                "pages": "54--59",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "F. Seide, and N. J. C. Wang, \"Phonetic modeling in the Philips Chinese continuous-speech recognition system\", Proc. of ISCSLP'98, 1998, pp. 54-59.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Categorical Data Analysis",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Agresti",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "A. Agresti, Categorical Data Analysis, John Wiley & Sons, 1990.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Some methods for strengthening of common tests",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [
+                            "G"
+                        ],
+                        "last": "Cochran",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1954,
+                "venue": "J. of the International Biometric Society",
+                "volume": "",
+                "issue": "",
+                "pages": "417--451",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "W. G. Cochran, \"Some methods for strengthening of common tests\", J. of the International Biometric Society, 1954, pp. 417-451.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Principles of Multivariate Analysis",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Krzanowski",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1988,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "W. J. Krzanowski, Principles of Multivariate Analysis. Oxford University Press, New York, 1988.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "An RNN-based pre-classification method for fast continuous Mandarin speech recognition",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Liao",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Chiang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "IEEE Transactions on Speech and Audio Processing",
+                "volume": "6",
+                "issue": "1",
+                "pages": "86--90",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "S. H. Chen, Y. F. Liao, S. M. Chiang, and S. Chang, \"An RNN-based pre-classification method for fast continuous Mandarin speech recognition\", IEEE Transactions on Speech and Audio Processing, Vol. 6, No. 1, 1998, pp. 86-90.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Isolated Mandarin based-syllable recognition based upon the segmental probability model",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [
+                            "Y"
+                        ],
+                        "last": "Lyc",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "I",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Hong",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Shen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "Y"
+                        ],
+                        "last": "Lee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "L",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Lee",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "IEEE Transactions on Speech and Audio Processing",
+                "volume": "6",
+                "issue": "3",
+                "pages": "293--299",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R. Y. Lyc, I. C. Hong, J. L. Shen, M. Y. Lee, and L. S. Lee, \"Isolated Mandarin based-syllable recognition based upon the segmental probability model\", IEEE Transactions on Speech and Audio Processing, Vol. 6, No. 3, 1998, pp. 293-299.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Integration of phonetic and prosodic information for robust utterance verification",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Wu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Yan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "IEE Proceedings-Vision, Image and Signal Processing",
+                "volume": "147",
+                "issue": "",
+                "pages": "55--61",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C. H. Wu, Y. J. Chen, and G. L. Yan, \"Integration of phonetic and prosodic information for robust utterance verification\", IEE Proceedings-Vision, Image and Signal Processing, Vol. 147, 2000, pp. 55-61.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "uris": null,
+                "text": "Confusion relationship of subsyllables A and B",
+                "type_str": "figure",
+                "num": null
+            },
+            "FIGREF1": {
+                "uris": null,
+                "text": "Design of the Modified SAMPA-CBased on the analysis of confusion characteristics, several confusion subsyllables caused by the confusable phonetic representation can be extracted. The confusable phonetic representation can be automatically detected using the above process. In our experimental results, the automatic speech recognition based on SAMPA-C cannot model the rapid variation between subsyllables. This is because that the confusion always occurs in the short duration between two subsyllables and the phonetic units representing the short phones cannot model this short duration well. Accordingly, a longer phonetic representation similar to subsyllable units is adopted to eliminate the confusion between two confusable subsyllables. These unsuitable phonetic units are manually analyzed. Each unit is concatenated with other phonetic unit to form a new, longer phonetic unit. The testing process is performed on the new representation iteratively. Finally, a modified SAMPA-C phonetic set, which suitably represent Chinese pronunciation is obtained and listed inTable VI. The original SAMPA-C phonetic set is also listed inTable VIfor comparison. The phonetic units with boldface are the newly defined units. For example, the new phonetic unit \"G\" is defined by concatenating the phonetic units \"d\" and \"C.\" The total number of phonetic units in the modified SAMPA-C becomes 52 compared to 45 in the original SAMPA-C.",
+                "type_str": "figure",
+                "num": null
+            },
+            "FIGREF3": {
+                "uris": null,
+                "text": "Distributions of error rate for subsyllables (a) dC, (b) C, (c) dZ, and (d) d. \u03c7 2 value of \"dC\" compared with other subsyllables Syllable error rates with respect to SAMPA-C and modified SAMPA-C based recognition system. In order to evaluate the performance of different phonetic representations, we conducted experiments on three continuous syllable recognition model types. Three forms of subsyllablic units -right-context dependent INITIAL/FINAL (RCD-IF), SAMPA-C based tri-phones, and modified SAMPA-C based tri-phones were conducted to evaluate the syllable recognition rates (SRR).",
+                "type_str": "figure",
+                "num": null
+            },
+            "TABREF1": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td>INITIAL</td><td colspan=\"3\">NULL b</td><td>p</td><td>m</td><td>f</td><td>d</td><td>t</td><td>n</td><td>l</td><td>g</td><td>k</td><td/></tr><tr><td colspan=\"2\">Number of appearances</td><td>141</td><td>65</td><td>46</td><td>47</td><td>17</td><td>227</td><td>71</td><td>87</td><td>97</td><td>59</td><td/><td/></tr><tr><td colspan=\"2\">Number of misrecognition</td><td>49</td><td>27</td><td>11</td><td>10</td><td>6</td><td>57</td><td>38</td><td>23</td><td>26</td><td>21</td><td/><td/></tr><tr><td>INITIAL</td><td/><td>h</td><td>dC</td><td>tC</td><td>C</td><td>dZ</td><td>tS</td><td>S</td><td>R</td><td>dz</td><td>ts</td><td>s</td><td/></tr><tr><td colspan=\"2\">Number of appearances</td><td>80</td><td>54</td><td>51</td><td>49</td><td>62</td><td>56</td><td>72</td><td>52</td><td>57</td><td>49</td><td/><td/></tr><tr><td colspan=\"2\">Number of misrecognition</td><td>18</td><td>51</td><td>31</td><td>19</td><td>32</td><td>51</td><td>38</td><td>11</td><td>36</td><td>44</td><td/><td/></tr><tr><td colspan=\"13\">Table II. SA distributions of FINALs represented by SAMPA-C, 2 \u03c7 value = 97, p \u2264 0.05</td><td/></tr><tr><td>FINAL</td><td colspan=\"2\">NULL a:</td><td>O:</td><td>V:</td><td>ai</td><td>ei</td><td>aU</td><td colspan=\"5\">ou aM @M aN VN</td><td>r</td></tr><tr><td>Number of appearances</td><td>38</td><td>72</td><td>8</td><td colspan=\"2\">194 60</td><td>40</td><td>61</td><td>53</td><td>69</td><td>52</td><td>59</td><td>56</td><td>8</td></tr><tr><td>Number of misrecognition</td><td>11</td><td>32</td><td>3</td><td>33</td><td>24</td><td>14</td><td>36</td><td>29</td><td>13</td><td>20</td><td>18</td><td>14</td><td>2</td></tr><tr><td>FINAL</td><td>i:</td><td>ja:</td><td>jE</td><td colspan=\"7\">jai jaU jou jEM IM jaN IN</td><td colspan=\"3\">u: wa: wO:</td></tr><tr><td>Number of appearances</td><td>41</td><td>15</td><td>37</td><td>4</td><td>30</td><td>29</td><td>47</td><td>34</td><td>25</td><td>43</td><td>58</td><td>21</td><td>47</td></tr><tr><td>Number of misrecognition</td><td>21</td><td>11</td><td>11</td><td>2</td><td>6</td><td>11</td><td>25</td><td>25</td><td>6</td><td>22</td><td>13</td><td>8</td><td>22</td></tr><tr><td>FINAL</td><td colspan=\"7\">wai wei waM w@M waN wVN y:</td><td colspan=\"4\">yE yEM yM yN</td><td/><td/></tr><tr><td>Number of appearances</td><td>23</td><td>57</td><td>58</td><td>38</td><td>27</td><td>53</td><td>17</td><td>23</td><td>24</td><td>14</td><td>16</td><td/><td/></tr><tr><td>Number of misrecognition</td><td>8</td><td>9</td><td>23</td><td>20</td><td>12</td><td>4</td><td>11</td><td>11</td><td>12</td><td>4</td><td>8</td><td/><td/></tr></table>",
+                "text": ", the \u03c7 2 values are 164 and 97 for INITIALs and FINALs, respectively. It is clear that the \u03c7 2 value is greater than 5% of the significance level. Therefore, the analyzed results show significant evidences that the confusion characteristics of INITIALs and FINALs can be regarded as subsyllable-dependent.Table I. SA distributions of INITIALs represented by SAMPA-C, 2\u03c7 value = 164, p \u2264 0.05",
+                "html": null
+            },
+            "TABREF2": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td colspan=\"5\">concatenating (+) phonetic units in SAMPA-C</td></tr><tr><td>Subsyllable</td><td>d</td><td>d+C</td><td>d+Z</td><td>d+z</td></tr><tr><td>Num. of appearances</td><td>227</td><td>54</td><td>62</td><td>57</td></tr><tr><td>Num. of misrecognition</td><td>57</td><td>51</td><td>32</td><td>36</td></tr><tr><td/><td/><td colspan=\"3\">\u03c7 2 value = 55, p \u2264 0.05</td></tr><tr><td/><td>(a)</td><td/><td/><td/></tr><tr><td>Subsyllable</td><td colspan=\"3\">y+E y+E+M y+M</td><td>y+N</td></tr><tr><td>Num. of appearances</td><td>23</td><td>14</td><td>14</td><td>16</td></tr><tr><td>Num. of misrecognition</td><td>11</td><td>4</td><td>4</td><td>8</td></tr><tr><td/><td/><td colspan=\"3\">\u03c7 2 value = 1.65, p \u2265 0.05</td></tr><tr><td/><td>(b)</td><td/><td/><td/></tr></table>",
+                "text": "Comparison of SA distributions of syllables represented by",
+                "html": null
+            },
+            "TABREF3": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td/><td colspan=\"8\">Table V. Examples of confusion measure (number of appearances)</td></tr><tr><td/><td/><td colspan=\"3\">Recognition Result</td><td/><td/><td/><td>Recognition Result</td></tr><tr><td/><td/><td>d+C</td><td/><td>C</td><td/><td/><td/><td>d+C</td><td>d</td></tr><tr><td>Observations</td><td>d+C C</td><td>3 3</td><td/><td>23 30</td><td colspan=\"2\">Observations</td><td>d+C d</td><td>3 0</td><td>12 170</td></tr><tr><td/><td colspan=\"4\">\u03c7 2 value = 0.00265, p \u2265 0.05</td><td/><td/><td colspan=\"2\">\u03c7 2 value = 23.16, p \u2264 0.05</td></tr><tr><td/><td>(a)</td><td/><td/><td/><td/><td/><td>(c)</td></tr><tr><td/><td/><td colspan=\"3\">Recognition Result</td><td/><td/><td/><td>Recognition Result</td></tr><tr><td/><td/><td>U+N</td><td colspan=\"2\">i+U+N</td><td/><td/><td/><td>U+N</td><td>V+N</td></tr><tr><td>Observations</td><td>U+N i+U+N</td><td>49 7</td><td/><td>0 8</td><td colspan=\"2\">Observations</td><td>U+N V+N</td><td>49 2</td><td>3 42</td></tr><tr><td/><td colspan=\"4\">\u03c7 2 value = 0.00146, p \u2265 0.05</td><td/><td/><td colspan=\"2\">\u03c7 2 value = 76.98, p \u2264 0.05</td></tr><tr><td/><td>(b)</td><td/><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td/><td/><td colspan=\"2\">Recognition Result</td><td/></tr><tr><td/><td/><td/><td/><td/><td>R A</td><td>R B</td><td/></tr><tr><td/><td colspan=\"3\">Observations</td><td>O A O B</td><td>P(R A |O A ) P(R A |O B )</td><td colspan=\"2\">P(R B |O A ) P(R B |O B )</td></tr></table>",
+                "text": "Four possible outcomes for a given trial",
+                "html": null
+            },
+            "TABREF4": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td>Modified</td><td>Examples by</td><td>Modified</td><td>Examples by</td></tr><tr><td>SAMPA-C</td><td>PINYIN</td><td>SAMPA-C</td><td>PINYIN</td></tr><tr><td>G(d+C)</td><td>GIN (\u6676 jing1)</td><td>z(d+z)</td><td>zI: (\u5b50 zi3)</td></tr><tr><td>Q(t+C)</td><td>Qi: (\u4e03 qi1)</td><td>c(t+s)</td><td>cu@M (\u6751 cun1)</td></tr><tr><td>X(C)</td><td>XiaU (\u5c0f xiao3)</td><td>aN(a+N)</td><td>laN (\u72fc lang2)</td></tr><tr><td>Z(d+Z)</td><td>ZUN (\u4e2d zhong1)</td><td>aM(a+M)</td><td>maM (\u6162 man4)</td></tr><tr><td>C(t+S)</td><td>Ca: (\u8336 cha2)</td><td>iU(I+U)</td><td>XiUN (\u5144 xiong)</td></tr><tr><td>4. Experimental Results</td><td/><td/><td/></tr><tr><td colspan=\"4\">In the experiment setup, a Mandarin Speech Across Taiwan (MAT) telephone speech database,</td></tr><tr><td colspan=\"4\">pronounced by 160 speakers (81 males, 79 females), with 8,237 files (sampling rate of 8kHz) was</td></tr><tr><td colspan=\"4\">employed. Another speech database with 500 utterances was also collected and used as the testing</td></tr><tr><td colspan=\"4\">data. In the following experiments, 12 Mel-Frequency Cepstrum Coefficient (MFCC), 12 delta</td></tr><tr><td colspan=\"4\">MFCC, one delta log energy, and one delta delta log energy are extracted as a 26-dimension feature</td></tr><tr><td>vector.</td><td/><td/><td/></tr></table>",
+                "text": "Modified SAMPA-C and the examples with the corresponding Chinese characters and PinYin",
+                "html": null
+            },
+            "TABREF5": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td colspan=\"3\">Recognition rates using SAMPA-C and modified SAMPA-C, respectively</td></tr><tr><td/><td colspan=\"2\">SAMPA-C Modified SAMPA-C</td></tr><tr><td colspan=\"2\">INITIAL 55.86%</td><td>75.08%</td></tr><tr><td>FINAL</td><td>66.53%</td><td>67.26%</td></tr><tr><td colspan=\"3\">For Mandarin speech, the confusion effects of INITIALs are more obvious than that of</td></tr><tr><td colspan=\"3\">FINALs. Due to the channel distortion of telephone network, the unvoiced INITIAL part with short</td></tr></table>",
+                "text": "",
+                "html": null
+            },
+            "TABREF6": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td/><td>RCD IF</td><td>SAMPA-C</td><td>Modified SAMPA-C</td></tr><tr><td/><td/><td>Tri-phones</td><td>Tri-phones</td></tr><tr><td>No. of Nodes</td><td>675</td><td>754</td><td>812</td></tr><tr><td>SRR</td><td>46.12%</td><td>43.23%</td><td>50.23%</td></tr></table>",
+                "text": "Syllable recognition rates using RCD IF, SAMPA-C based tri-phones, and modified SAMPA-C based tri-phones",
+                "html": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1006.json

@@ -0,0 +1,1007 @@
+{
+    "paper_id": "O01-1006",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:41.052050Z"
+    },
+    "title": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method",
+    "authors": [
+        {
+            "first": "Jau-Hung",
+            "middle": [],
+            "last": "Chen",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "Advanced Technology Center, Computer and Communication Research Laboratories",
+                "institution": "Industrial Technology Research Institute",
+                "location": {
+                    "addrLine": "Chutung 310",
+                    "country": "Taiwan"
+                }
+            },
+            "email": "chenjh@itri.org.tw"
+        },
+        {
+            "first": "Yung-An",
+            "middle": [],
+            "last": "Kao",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "Advanced Technology Center, Computer and Communication Research Laboratories",
+                "institution": "Industrial Technology Research Institute",
+                "location": {
+                    "addrLine": "Chutung 310",
+                    "country": "Taiwan"
+                }
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "In a text-to-speech (TTS) conversion system based on the time-domain pitch-synchronous overlap-add (TD-PSOLA) method, accurate estimation of pitch periods and pitch marks is necessary for pitch modification to assure an optimal quality of the synthetic speech. In general, there are two major issues on pitch marking: pitch detection and location determination. In this paper, an adaptable filter, which serves as a bandpass filter, is proposed for pitch detection to transform the voiced speech into a sine-like wave. Based on the sine-like wave, a peak-valley decision method is investigated to determine the appropriate part (positive part and negative part) of the voiced speech for pitch mark estimation. At each pitch period, two possible peaks/valleys are searched and the dynamic programming is performed to obtain the pitch marks. Experimental results indicate that our proposed method performed very well if correct pitch information is estimated.",
+    "pdf_parse": {
+        "paper_id": "O01-1006",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "In a text-to-speech (TTS) conversion system based on the time-domain pitch-synchronous overlap-add (TD-PSOLA) method, accurate estimation of pitch periods and pitch marks is necessary for pitch modification to assure an optimal quality of the synthetic speech. In general, there are two major issues on pitch marking: pitch detection and location determination. In this paper, an adaptable filter, which serves as a bandpass filter, is proposed for pitch detection to transform the voiced speech into a sine-like wave. Based on the sine-like wave, a peak-valley decision method is investigated to determine the appropriate part (positive part and negative part) of the voiced speech for pitch mark estimation. At each pitch period, two possible peaks/valleys are searched and the dynamic programming is performed to obtain the pitch marks. Experimental results indicate that our proposed method performed very well if correct pitch information is estimated.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "In past years, the approach of concatenative synthesis has been adopted by many text-to-speech (TTS) systems [1] - [6] . The concatenative synthesis uses real recorded speech segments as the synthesis units and concatenates them together during synthesis.",
+                "cite_spans": [
+                    {
+                        "start": 109,
+                        "end": 112,
+                        "text": "[1]",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 115,
+                        "end": 118,
+                        "text": "[6]",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Also, the time-domain pitch-synchronous overlap-add (TD-PSOLA) [6] method has been employed to perform prosody modification. This method modifies the prosodic features of the synthesis unit according to the target prosodic information. Generally, the prosodic information of the speech includes pitch (the fundamental frequency), intensity, and duration, etc. For a synthesis scheme based on TD-PSOLA method, it is necessary to obtain a pitch mark for each pitch period in order to assure an optimal quality of the synthetic speech. The pitch mark is a reference point for the overlap of the speech signals.",
+                "cite_spans": [
+                    {
+                        "start": 63,
+                        "end": 66,
+                        "text": "[6]",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "It is useful to have a speech synthesizer with various voices for speech synthesis.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Sometimes it is also important for a service-providing company to have a synthesizer with the voice of its own employee or the speaker of its favorite. For conventional TTS systems, however, it is a professional but tedious job to create a new voice. Recently, corpus-based TTS systems have been appreciated which use a large amount of speech segments. Some approaches selected the speech segments as the candidates of synthesis units. Establishing the synthesis units includes speech segmentation, pitch estimation, pitch marking, and so on. However, pitch marking is very labor-intensive among them if there involved no automatic mechanism.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In general, there are two major issues on pitch marking: pitch detection and location determination. Compared to pitch detection [7] - [14] , few papers have been presented for pitch marking [15] [16] , which is also a difficult problem because of the great variability of the speech signals. Moulines et al. [15] proposed a pitch-marking algorithm based on the detection of abrupt changes at glottal closure instants. At each period, they assumed that the speech waveform could be represented by the concatenation of the response of two all-pole systems. On the other hand, Kobayashi et al. [16] used dyadic wavelet for pitch marking. The glottal closure instant was detected by searching for a local peak in the wavelet transform of the speech waveform.",
+                "cite_spans": [
+                    {
+                        "start": 129,
+                        "end": 132,
+                        "text": "[7]",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 135,
+                        "end": 139,
+                        "text": "[14]",
+                        "ref_id": "BIBREF13"
+                    },
+                    {
+                        "start": 191,
+                        "end": 195,
+                        "text": "[15]",
+                        "ref_id": "BIBREF14"
+                    },
+                    {
+                        "start": 196,
+                        "end": 200,
+                        "text": "[16]",
+                        "ref_id": "BIBREF15"
+                    },
+                    {
+                        "start": 309,
+                        "end": 313,
+                        "text": "[15]",
+                        "ref_id": "BIBREF14"
+                    },
+                    {
+                        "start": 592,
+                        "end": 596,
+                        "text": "[16]",
+                        "ref_id": "BIBREF15"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we propose a pitch-marking method based on an adaptable filter and a peak-valley estimation method. The block diagram is shown in Fig. 1 . The input signals are constrained to the voiced speech because only the periodic parts are interested. We introduce an adaptable filter, which serves as a bandpass filter, to transform the voiced speech into a sine-like wave. The autocorrelation method is then used to estimate the pitch periods on the sine-like wave. Also, a peak-valley decision method is presented to determine which part of the voiced speech is suitable for pitch mark estimation. The positive part (the speech with positive amplitude) and the negative part (the speech with negative amplitude) are investigated in this method. This is motivated from Fig. 2 ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 145,
+                        "end": 151,
+                        "text": "Fig. 1",
+                        "ref_id": "FIGREF0"
+                    },
+                    {
+                        "start": 776,
+                        "end": 782,
+                        "text": "Fig. 2",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "The proposed adaptable filter serves as a bandpass filter in which its pass band is from 50 Hz to the detected fundamental frequency, up to 500 Hz, of the voiced speech. The adaptable filter is achieved by the following three steps.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Autocorrelation Method",
+                "sec_num": null
+            },
+            {
+                "text": "Step 1. It computes the FFT (Fast Fourier Transform) to transform the voiced speech into the frequency domain.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Autocorrelation Method",
+                "sec_num": null
+            },
+            {
+                "text": "Step 2. The fundamental frequency, f 0 , is detected by searching the first peak of the spectral contour.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Autocorrelation Method",
+                "sec_num": null
+            },
+            {
+                "text": "Step 3. The IFFT (Inverse FFT) is invoked over the passband between 50 Hz and f 0 to obtain the filtered speech. An example of the adaptable filter is displayed in Fig. 2 . Panel (a) and (b) shows the waveforms of the original speech and the filtered speech, respectively. It can be seen that the filtered speech is generally a sine-like wave that reveals clear periodicity than that on the original speech waveform. For a frame in the middle of the voiced speech, the spectral contour is depicted in panel (d). Note that the frequency axis is not linearly plotted for the reason of inspecting the first spectral peak. The first peak was found at 168 Hz, which is the fundamental frequency. Finally, the pitch periods are obtained by analyzing the filtered speech using the conventional autocorrelation method.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 164,
+                        "end": 170,
+                        "text": "Fig. 2",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Autocorrelation Method",
+                "sec_num": null
+            },
+            {
+                "text": "From observations, we found that the voiced speech, s [\u2022] , is synchronous with the filtered speech, o [\u2022] , either at peaks or at valleys. For the case illustrated in Fig. 2 (a) ",
+                "cite_spans": [
+                    {
+                        "start": 54,
+                        "end": 57,
+                        "text": "[\u2022]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 103,
+                        "end": 106,
+                        "text": "[\u2022]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 168,
+                        "end": 178,
+                        "text": "Fig. 2 (a)",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Pitch Mark Determination Using a Peak-Valley Decision Method and Dynamic Programming 3-1 Peak-Valley Decision",
+                "sec_num": "3."
+            },
+            {
+                "text": "where the symbols are defined as follows: ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Mark Determination Using a Peak-Valley Decision Method and Dynamic Programming 3-1 Peak-Valley Decision",
+                "sec_num": "3."
+            },
+            {
+                "text": "Once the adoption of the peak or valley has been decided, say peak, the positions of ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-2 Pitch mark determination Based on Dynamic Programming",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": ") , ( ) , ( ) 1 ( k j g P L L k j d i k i ij i + \u2212 \u2212 = \u2212 , for i=2,\u2026,PN (3) \uf8fe \uf8fd \uf8fc \uf8f3 \uf8f2 \uf8f1 + + = \u2212 \u2212 (2) ) 2 , ( (1), ) 1 , ( min ) ( 1 1 i i i i i A j d A j d j A , for i=2,3,\u2026,PN",
+                        "eq_num": "(4)"
+                    }
+                ],
+                "section": "3-2 Pitch mark determination Based on Dynamic Programming",
+                "sec_num": null
+            },
+            {
+                "text": "where PN is the total number of pitch period and j, k=1,2. In Equation 3 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-2 Pitch mark determination Based on Dynamic Programming",
+                "sec_num": null
+            },
+            {
+                "text": "The penalty function is introduced here due to the preference of the highest peak.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-2 Pitch mark determination Based on Dynamic Programming",
+                "sec_num": null
+            },
+            {
+                "text": "The search path of the dynamic programming is illustrated in Fig. 3 . The peak locations (pitch marks) can be obtained by back tracing the peak sequence corresponding to the smallest value of A i (1) and A i (2) . An example of the results of pitch marking is shown in Fig. 2(c) . Similar procedures described above can be applied to the case of \"valley\". For the voiced speech, the waveforms along with the pitch marks obtained from our pitch-marking program were visually displayed. The pitch marks were then checked and corrected by an experienced person through a friendly interface. For the evaluation of the experiments, we obtained 436 sets of human-labeled pitch marks, denoted as H, which comprises 23868 pitch marks.",
+                "cite_spans": [
+                    {
+                        "start": 208,
+                        "end": 211,
+                        "text": "(2)",
+                        "ref_id": "BIBREF1"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 61,
+                        "end": 67,
+                        "text": "Fig. 3",
+                        "ref_id": "FIGREF7"
+                    },
+                    {
+                        "start": 269,
+                        "end": 278,
+                        "text": "Fig. 2(c)",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "3-2 Pitch mark determination Based on Dynamic Programming",
+                "sec_num": null
+            },
+            {
+                "text": "The results of the peak-valley decision were verified by human judgment on visual displays. A success rate of 99.1% is obtained (4 of the 436 results were disagreed). For the female speaker, we found that 97.2% of the voiced segments reveal clear periodicity on the negative parts.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "4-2 Performance of the pitch marking method",
+                "sec_num": null
+            },
+            {
+                "text": "The proposed method generated 23860 pitch marks, denoted as I, without any duplication. The success rate of the pitch marking method is defined as follows: (6) As shown in Table 1 , a success rate of 97.2% is obtained (baseline), in contrast with the 95% and 97% success rates of the methods of [15] and [16] , respectively. However, we found that most of the errors are resulted from the incorrect results of pitch detection.",
+                "cite_spans": [
+                    {
+                        "start": 156,
+                        "end": 159,
+                        "text": "(6)",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 295,
+                        "end": 299,
+                        "text": "[15]",
+                        "ref_id": "BIBREF14"
+                    },
+                    {
+                        "start": 304,
+                        "end": 308,
+                        "text": "[16]",
+                        "ref_id": "BIBREF15"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 172,
+                        "end": 179,
+                        "text": "Table 1",
+                        "ref_id": "TABREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "4-2 Performance of the pitch marking method",
+                "sec_num": null
+            },
+            {
+                "text": "Most of the pitch errors are due to large changes of pitch locating at the boundaries of the voiced speech. Providing correct pitch information, our method leads to a success rate of 99.5%. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "4-2 Performance of the pitch marking method",
+                "sec_num": null
+            },
+            {
+                "text": "In this paper, a preliminary work on pitch marking has been proposed. We present the adaptable filter combined with the autocorrelation method for pitch detection. On the other hand, a peak-valley decision method is introduced to select either the positive or the negative parts for evaluation of pitch mark. Also, a dynamic-programming-based pitch mark determination method is demonstrated where two peaks/valleys are searched at each period. In the experiments, our pitch-marking method achieves 97.2% success rate.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions",
+                "sec_num": "5"
+            }
+        ],
+        "back_matter": [
+            {
+                "text": "This paper is a partial result of Project 3XS1B11 conducted by ITRI under sponsorship of the Ministry of Economic Affairs, Taiwan, R.O.C.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Acknowledgement",
+                "sec_num": null
+            }
+        ],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "A diphone synthesis based on time-domain prosodic modifications of speech",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Hamon",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Moulines",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Charpentier",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1989,
+                "venue": "Proc ICASSP",
+                "volume": "",
+                "issue": "",
+                "pages": "238--241",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Hamon, C., E. Moulines, and F. Charpentier, \"A diphone synthesis based on time-domain prosodic modifications of speech,\" in Proc ICASSP, 1989, pp.238-241.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Speech segment network approach for optimization of synthesis unit set",
+                "authors": [
+                    {
+                        "first": "N",
+                        "middle": [],
+                        "last": "Iwahashi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Sagisaka",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Computer Speech and Language",
+                "volume": "",
+                "issue": "",
+                "pages": "335--352",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Iwahashi, N. and Y. Sagisaka, \"Speech segment network approach for optimization of synthesis unit set,\" Computer Speech and Language, 1995, pp.335-352.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Issues in text-to-speech conversion for Mandarin",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Shih",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Sproat",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Computational Linguistics and Chinese Language Processing",
+                "volume": "1",
+                "issue": "",
+                "pages": "37--86",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Shih, C. L. and R. Sproat, \"Issues in text-to-speech conversion for Mandarin,\" in Computational Linguistics and Chinese Language Processing, vol.1, 1996, pp.37-86.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "An RNN-based prosodic information Synthesizer for Mandarin text-to-speech",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Hwang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Wang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "IEEE Trans. on Speech and Audio Processing",
+                "volume": "6",
+                "issue": "3",
+                "pages": "226--239",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, S. H., S. H. Hwang and Y. R. Wang, \"An RNN-based prosodic information Synthesizer for Mandarin text-to-speech,\" IEEE Trans. on Speech and Audio Processing, Vol. 6, No. 3, 1998, pp. 226-239.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Corpus-based Mandarin speech synthesis with contextual syllabic units based on phonetic properties",
+                "authors": [
+                    {
+                        "first": "F",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Chou",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [
+                            "Y"
+                        ],
+                        "last": "Tseng",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proc. ICASSP",
+                "volume": "",
+                "issue": "",
+                "pages": "893--896",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chou, F. C. and C. Y. Tseng, \"Corpus-based Mandarin speech synthesis with contextual syllabic units based on phonetic properties,\" in Proc. ICASSP, 1998, pp.893-896.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Diphone synthesis using an overlap-add technique for speech waveforms concatenation",
+                "authors": [
+                    {
+                        "first": "F",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Charpentier",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "G"
+                        ],
+                        "last": "Stella",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1986,
+                "venue": "Proc. ICASSP",
+                "volume": "",
+                "issue": "",
+                "pages": "2015--2020",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Charpentier, F. J. and M. G. Stella, \"Diphone synthesis using an overlap-add technique for speech waveforms concatenation,\" in Proc. ICASSP, 1986, pp. 2015-2020.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "A Comparative performance study of several pitch detection algorithms",
+                "authors": [
+                    {
+                        "first": "L",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Rabiner",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Cheng",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Rosenberg",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1976,
+                "venue": "IEEE Trans. Acoust., Speech, Signal Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "399--417",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Rabiner, L. R., M. J. Cheng, A. E. Rosenberg, and C. A. McGonegal, \"A Comparative performance study of several pitch detection algorithms,\" IEEE Trans. Acoust., Speech, Signal Processing, vol. ASSP-24, 1976, pp. 399-417.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "On the use of autocorrelation analysis for pitch detection",
+                "authors": [
+                    {
+                        "first": "L",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Rabiner",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1977,
+                "venue": "IEEE Trans. Acoust., Speech, Signal Processing",
+                "volume": "25",
+                "issue": "",
+                "pages": "24--33",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Rabiner, L. R., \"On the use of autocorrelation analysis for pitch detection,\" IEEE Trans. Acoust., Speech, Signal Processing, Vol. ASSP-25, 1977, pp. 24-33.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Cepstrum pitch determination",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Noll",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1967,
+                "venue": "J. Acoust. Soc. Amer",
+                "volume": "47",
+                "issue": "",
+                "pages": "293--309",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Noll, A. M., \"Cepstrum pitch determination,\" J. Acoust. Soc. Amer., Vol. 47, 1967, pp. 293-309.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "The SIFT algorithm for fundamental frequency estimation",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Markel",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1972,
+                "venue": "IEEE Trans. Audio Electroacoust",
+                "volume": "20",
+                "issue": "",
+                "pages": "367--377",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Markel, J. D., \"The SIFT algorithm for fundamental frequency estimation,\" IEEE Trans. Audio Electroacoust., Vol. Au-20, 1972, pp. 367-377.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Pitch detection with a neural-net classifier",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Barnard",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Cole",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "P"
+                        ],
+                        "last": "Vea",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Alleva",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "IEEE Trans. On Signal Processing",
+                "volume": "39",
+                "issue": "2",
+                "pages": "298--307",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Barnard, E., R. A. Cole, M. P. Vea, and F. A. Alleva, \"Pitch detection with a neural-net classifier,\" IEEE Trans. On Signal Processing, vol. 39, No. 2, 1991, pp. 298-307.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "A comparison of a wavelet functions for pitch detection of speech signals",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Kadambe",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Boudreaux-Bartels",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "Proc. ICASSP",
+                "volume": "",
+                "issue": "",
+                "pages": "449--452",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Kadambe, S., G. F. Boudreaux-Bartels, \"A comparison of a wavelet functions for pitch detection of speech signals,\" in Proc. ICASSP, 1991, pp.449-452.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "Colored L-l filters and their application in speech pitch detection",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Barner",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "IEEE Trans. On Signal Processing",
+                "volume": "48",
+                "issue": "9",
+                "pages": "2601--2606",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Barner, K. E., \"Colored L-l filters and their application in speech pitch detection,\" IEEE Trans. On Signal Processing, Vol. 48, No. 9, 2000, pp. 2601-2606.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Pitch tracking and tone features for Mandarin speech recognition",
+                "authors": [
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Seide",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Proc. ICASSP",
+                "volume": "",
+                "issue": "",
+                "pages": "1523--1526",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Huang, H. and F. Seide, \"Pitch tracking and tone features for Mandarin speech recognition,\" in Proc. ICASSP, 2000, pp.1523-1526.",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "A real-time French text-to-speech system generating high-quality synthetic speech",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Moulines",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Emerard",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Larreur",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Le Saint Milon",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "L",
+                        "middle": [
+                            "Le"
+                        ],
+                        "last": "Faucheur",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Marty",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Charpentier",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Sorin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Proc. ICASSP",
+                "volume": "",
+                "issue": "",
+                "pages": "309--312",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Moulines, E., F. Emerard, D. Larreur, J. L. Le Saint Milon, L. Le Faucheur, F. Marty, F. Charpentier, and C. Sorin, \"A real-time French text-to-speech system generating high-quality synthetic speech,\" in Proc. ICASSP, 1990, pp.309-312.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "Wavelet analysis used in text-to-speech synthesis",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Kobayashi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Sakamoto",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Saito",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Hashimoto",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Nishimura",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Suzuki",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "IEEE Trans. on Circuits and Systems-II",
+                "volume": "45",
+                "issue": "8",
+                "pages": "1125--1129",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Kobayashi, M., M. Sakamoto, T. Saito, Y. Hashimoto, M. Nishimura, and K. Suzuki, \"Wavelet analysis used in text-to-speech synthesis,\" IEEE Trans. on Circuits and Systems-II, Analog and Digital Signal Processing, Vol. 45, No. 8, 1998, pp. 1125-1129.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "num": null,
+                "uris": null,
+                "text": "(a), which displays an example of waveform having the negative part reveals explicit periodicity. In general, it could synthesize better speech quality if the pitch marks are labeled at the positions of extreme points (peaks and valleys) of the speech. At each pitch period, two possible peaks/valleys are searched. Finally, the pitch marks are obtained by the dynamic programming by calculating the pitch distortion. Block diagram of the proposed pitch-marking method.",
+                "type_str": "figure"
+            },
+            "FIGREF1": {
+                "num": null,
+                "uris": null,
+                "text": "Results of the adaptable filter and pitch mark determination. (a) Waveform of the voiced speech with explicit periodicity on the negative part. (b) Waveform of the filtered speech. (c) Detected pitch marks. (d) Spectral contour (note that the frequency axis is not linearly plotted).",
+                "type_str": "figure"
+            },
+            "FIGREF2": {
+                "num": null,
+                "uris": null,
+                "text": "b), they are synchronous at valleys having explicit periodicity instead of those at peaks.As a result, the pitch marks could be easily determined at the negative part than those at the positive part. In the following, peak-valley decision method calculates two costs bysumming the amplitudes of s[m], where m represents the position of the local extreme point of o[\u2022] over each pitch period:",
+                "type_str": "figure"
+            },
+            "FIGREF3": {
+                "num": null,
+                "uris": null,
+                "text": "Cost estimated at the peaks of o[\u2022]. valley C : Cost estimated at the valleys of o[\u2022]. peak N : Total number of the peaks of o[\u2022]. valley N : Total number of the valleys of o[\u2022]. ] [n Pos peak : Position of the n-th peak of o[\u2022]. ] [n Pos valley : Position of the n-th valley of o[\u2022]. The peak-valley decision is made as follows: If peak C > valley C then the positive part (peak) of s[\u2022] is adopted for the evaluation of pitch mark. Otherwise, the negative part (valley) of s[\u2022] is adopted.",
+                "type_str": "figure"
+            },
+            "FIGREF4": {
+                "num": null,
+                "uris": null,
+                "text": "pitch marks are determined by picking the peaks of s[\u2022]. For the i-th pitch period, P i , two highest peaks in the corresponding voiced speech are searched. Suppose the highest and the second highest peaks are located at L i1 and L i2 , respectively. It might occur that the second one is absent. For this case, we let L i2 = L i1 . For all the detected peaks, the determination of pitch mark is then performed based on dynamic programming. The distortion of pitch period, d i (j,k), and its accumulation, A i (j), are defined as follows:",
+                "type_str": "figure"
+            },
+            "FIGREF7": {
+                "num": null,
+                "uris": null,
+                "text": "Illustration of the peak-picking search path of the dynamic programming.",
+                "type_str": "figure"
+            },
+            "FIGREF8": {
+                "num": null,
+                "uris": null,
+                "text": "continuous speech database was established which provides the basic synthesis units of our Mandarin Chinese TTS system. This database is composed of 70 phrases and their lengths are between 4 to 6 Chinese characters. It includes an amount of 436 tonal syllables comprising the required 413 basic synthesis units. A native female speaker read them in normal speaking style. The speech signals were then digitized by a 16-bit A/D converter at a 44.1k Hz sampling rate. The syllable segmentation was manually done in order to obtain the precise boundaries of voiced speech and unvoiced speech. The total duration of the 436 voiced speech is about 2.1 minutes. For each syllable, the voiced speech was used to test the proposed methods. The frame size used in the adaptable filter was set to 4096 speech samples (92.8 ms).",
+                "type_str": "figure"
+            },
+            "TABREF1": {
+                "html": null,
+                "text": "Success rate of the pitch-marking method.",
+                "num": null,
+                "content": "<table><tr><td>Condition</td><td colspan=\"2\">Baseline Using correct pitch</td></tr><tr><td>Success rate</td><td>97.2%</td><td>99.5%</td></tr></table>",
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1007.json

@@ -0,0 +1,76 @@
+{
+    "paper_id": "O01-1007",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:32.377180Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O01-1007",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "= 9 QDPHO\\ 1 L 6 T 3 L = = = LW PXVW VDWLVI\\ \u2211 = = 1 L L ^D LM _ L M 1",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "M F = O O Y M PD[ DUJ = V 1 O \u2264 \u2264 6HW F E = DQG F E Y Y = 6WHS &RS\\ WKH FXUUHQW EHVW SDWK F WR HDFK WHVW VROXWLRQ O V 1 O \u2264 \u2264 )RU HDFK WHVW VROXWLRQ O V 1 O \u2264 \u2264 JHQHUDWH WZR UDQGRP LQWHJHUV U DQG U 1 U \u2264 \u2264 1 U \u2264 \u2264 U U \u2260 *",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "+ = O O W W ,I V O 7 W > VHW = O W ,I P O L < VHW + = L",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {},
+        "ref_entries": {
+            "TABREF0": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td colspan=\"2\">JOREDO RSWLPXP</td><td/><td/><td/><td/></tr><tr><td colspan=\"7\">,Q 6HFWLRQ RI WKLV SDSHU WKH GHILQLWLRQ RI +00 LV JLYHQ WKHQ WKH WDEX VHDUFK DOJRULWKP LV</td></tr><tr><td colspan=\"7\">GHVFULEHG LQ 6HFWLRQ 7KH 76+00 WUDLQLQJ DOJRULWKP LV SUHVHQWHG LQ 6HFWLRQ 6LPXODWLRQ UHVXOWV</td></tr><tr><td colspan=\"7\">DUH VKRZQ LQ 6HFWLRQ DQG FRQFOXVLRQV DUH JLYHQ LQ 6HFWLRQ</td></tr><tr><td colspan=\"7\">+,''(1 0$5.29 02'(/</td></tr><tr><td colspan=\"7\">+00 LV D SUREDELOLW\\ PRGHO XVHG WR UHSUHVHQW WKH VWDWLVWLF SURSHUW\\ RI WKH VWRFKDVWLF SURFHVV DQG</td></tr><tr><td colspan=\"7\">LV FKDUDFWHUL]HG E\\ WKH PRGHO SDUDPHWHUV 7KH VWRFKDVWLF SURFHVV LQ VSHHFK UHFRJQLWLRQ LV WKH ILQLWH</td></tr><tr><td colspan=\"7\">OHQJWK VWRFKDVWLF VHTXHQFHV FDOOHG REVHUYDWLRQ V\\PERO GHQRWHG E\\</td><td>=</td><td>4</td><td>4</td><td>4</td><td>ZKHUH LV WKH</td></tr><tr><td colspan=\"7\">GLPHQVLRQ RI WKH REVHUYDWLRQ V\\PERO 2QH +00 ZLWK VWDWHV $ $ $ FDQ EH FKDUDFWHUL]HG E\\</td></tr><tr><td colspan=\"2\">WKH SDUDPHWHU VHW</td><td>=</td><td colspan=\"2\">{</td><td>\u00f9</td><td>}</td><td>ZKHUH</td></tr><tr><td>=</td><td>[</td><td colspan=\"2\">]</td><td colspan=\"3\">LV WKH LQLWLDO GLVWULEXWLRQ ,W LV XVHG WR GHVFULEH WKH SUREDELOLW\\ GLVWULEXWLRQ RI WKH</td></tr><tr><td colspan=\"7\">REVHUYDWLRQ V\\PERO LQ WKH LQLWLDO PRPHQW ZKHQ</td></tr></table>",
+                "text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
+                "html": null
+            },
+            "TABREF1": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td>7DEX 6HDUFK $OJRULWKP</td><td/></tr><tr><td colspan=\"2\">\u2211 = 1 M ^E LN _ L 1 N 0`LV WKH REVHUYDWLRQ V\\PERO SUREDELOLW\\ GLVWULEXWLRQ PDWUL[ LQ = D LM JHQHUDWH WKH LQLWLDO VROXWLRQV FDOFXODWH WKH FXUUHQW VROXWLRQ DQG WKH EHVW VROXWLRQ PHPRU\\ /HW { } 8 1 W = WR EH WKH VHW RI WKH WHVW VROXWLRQV OHW { } 1 F F F F = DQG</td></tr><tr><td colspan=\"2\">WKH GLVFUHWH +00 ,WV HOHPHQW DW URZ L FROXPQ N LV WKH SUREDELOLW\\ E LN RI REVHUYDWLRQ V\\PERO ZLWK ZKLOH WHUPLQDWLRQ FULWHULRQ QRW UHDFKH\u011c { } 1 E E E = EH WKH EHVW VROXWLRQ RI FXUUHQW LWHUDWLRQ DQG WKH EHVW VROXWLRQ RI DOO LQGH[ N HPLWWHG E\\ FXUUHQW VWDWH L DQG PXVW VDWLVI\\ WKH IROORZLQJ FRQGLWLRQ \u2211 = = 0 N LN E LWHUDWLRQV UHVSHFWLYHO\\ OHW ' W ^Y Y \u00ab Y 1V` Y F DQG Y E GHQRWH WKH VHW RI REMHFWLYH IXQFWLRQ YDOXHV IRU E JHQHUDWH WKH WHVW VROXWLRQV LQ WKH QHLJKERUKRRG RI WKH FXUUHQW VROXWLRQV WHVW VROXWLRQV WKH REMHFWLYH IXQFWLRQ YDOXH IRU WKH EHVW VROXWLRQ RI FXUUHQW LWHUDWLRQ DQG WKH REMHFWLYH FDOFXODWH WKH FRUUHVSRQGLQJ REMHFWLYH YDOXHV IXQFWLRQ YDOXH IRU WKH EHVW VROXWLRQ RI DOO LWHUDWLRQV UHVSHFWLYHO\\ ZKHUH Y O LV WKH REMHFWLYH IXQFWLRQ XSGDWH WKH FXUUHQW VROXWLRQ DQG WKH EHVW VROXWLRQ YDOXH IRU VROXWLRQ O V 1 O \u2264 \u2264 7KH DOJRULWKP LV JLYHQ DV IROORZV XSGDWH WKH WDEX OLVW 6WHS 6HW WKH WDEX OLVW VL]H V 7 WKH QXPEHU RI WHVW VROXWLRQV V 1 DQG WKH RSWLPXP QXPEHU RI LWHUDWLRQV</td></tr><tr><td>P O 6HW WKH LWHUDWLRQ FRXQWHU VROXWLRQV { } 8 1 W =</td><td>= L DQG LQVHUWLRQ SRLQW RI WKH WDEX OLVW UDQGRPO\\ FDOFXODWH WKH FRUUHVSRQGLQJ REMHFWLYH YDOXHV = O 1 W *HQHUDWH V</td></tr><tr><td>7+( 76+00 $/*25,7+0 '</td><td/></tr><tr><td/><td>DQG UHVSHFWLYHO\\</td></tr><tr><td colspan=\"2\">$ VROXWLRQ RI WKLV DOJRULWKP LV GHILQHG DV 8 O FRQVLVWLQJ RI D VHW RI UHDO QXPEHUV OLNH WKH RQH VKRZQ `LV WKH WUDQVLWLRQ SUREDELOLW\\ GLVWULEXWLRQ PDWUL[ ,WV HOHPHQW DW URZ L LQ )LJ 7KH SUREDELOLW\\ 5 Q _ Q RI WKH +00 VROXWLRQ\u00ac Q ZKLFK JHQHUDWHV WKH WUDLQLQJ REVHUYDWLRQ FROXPQ M LV WKH SUREDELOLW\\ D LM RI WUDQVLWLRQ IURP FXUUHQW VWDWH L WR QH[W VWDWH M QDPHO\\ D LM 3 T W 6 M _ T W 6 L VHTXHQFHV 0 4 4 4</td></tr><tr><td colspan=\"2\">LW PXVW VDWLVI\\ WKH IROORZLQJ FRQGLWLRQ</td></tr><tr><td>IROORZV</td><td/></tr></table>",
+                "text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
+                "html": null
+            },
+            "TABREF2": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td/><td/><td/><td/><td/><td/><td/><td>HQHUDWH</td></tr><tr><td colspan=\"8\">WKH QHZ WHVW VROXWLRQV E\\ VZDSSLQJ</td><td>O</td><td>U</td><td>DQG</td><td>U</td><td>&amp;DOFXODWH WKH FRUUHVSDQGLQJ REMHFWLYH</td></tr><tr><td colspan=\"2\">YDOXHV</td><td>Y</td><td>Y</td><td>Y</td><td>1</td><td>8</td><td>IRU WKH QHZ WHVW VROXWLRQV</td></tr><tr><td colspan=\"8\">6WHS 6RUW LQ LQFUHDVLQJ RUGHU )6WHS ,I 8 1 Y Y Y E F Y Y &lt; VHW E F = DQG E F Y Y = ,QVHUW WKH VZDSSHG LQGLFHV RI WKH FXUUHQW EHVW VROXWLRQ</td></tr><tr><td>F</td><td colspan=\"7\">LQWR WKH WDEX OLVW 6HW WKH LQVHUWLQJ SRLQW RI WKH WDEX OLVW</td></tr></table>",
+                "text": "URP WKH EHVW WHVW VROXWLRQ WR WKH ZRUVW WHVW VROXWLRQ LI WKH WHVW VROXWLRQ LV D QRQWDEX VROXWLRQ RU LW LV D WDEX VROXWLRQ EXW LWV REMHFWLYH YDOXH LV ODUJHU WKDQ WKH EHVW YDOXH RI DOO LWHUDWLRQV Y E DVSLUDWLRQ OHYHO WKHQ FKRRVH WKLV VROXWLRQ DV WKH FXUUHQW EHVW VROXWLRQ F DQG FKRRVH LWV REMHFWLYH YDOXH DV WKH FXUUHQW EHVW REMHFWLYH YDOXH Y F JR WR VWHS RWKHUZLVHWU\\ WKH QH[W WHVW VROXWLRQ ,I DOO WHVW VROXWLRQV DUH WDEX VROXWLRQV WKHQ JR WR VWHS",
+                "html": null
+            },
+            "TABREF3": {
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td>5()(5(1&amp;(</td></tr><tr><td>6,08/$7,216 / 5 5DELQHU \u00b3$ WXWRULDO RQ +LGGHQ 0DUNRY 0RGHOV DQG 6HOHFWHG DSSOLFDWLRQV LQ VSHHFK</td></tr><tr><td>= UHFRJQLWLRQ\u00b43URFHHGLQJ RI ,((( 9RO 1R SS a V 7 WKH WKUHVKROG RI WKH SUREDELOLW\\ = WK 3 WKH QXPEHU RI WKH LWHUDWLRQ , P -RVHSK 3LFRQH \u00b3&amp;RQWLQXRXV VSHHFK UHFRJQLWLRQ XVLQJ +LGGHQ 0DUNRY 0RGHOV\u00b4,((( $663 0DJ 9RO 1R SS a 6SHHFK $XGLR 3URFHVVLQJ 9RO 1R SS a H[WKH WDEX OLVW ' %XUVKWHLQ \u00b35REXVW SDUDPHWULF PRGHOLQJ RI GXUDWLRQ LQ +LGGHQ 0DUNRY 0RGHOV\u00b4,((( 7UDQV RQ )LJ $ ILYH VWDWHV OHIWULJKW PRGHO</td></tr><tr><td>WKH QXPEHU RI WKH VROXWLRQV LQ HDFK LWHUDWLRQ1 6 7KH LQLWLDO PRGHO SDUDPHWHUV DUH FUHDWHG UDQGRPO\\ ) -HOLQHN \u00b3&amp;RQWLQXRXV VSHHFK UHFRJQLWLRQ E\\ VWDWLVWLFDO PHWKRGV\u00b43URFHHGLQJ RI ,((( 9RO 1R</td></tr><tr><td>DQG DUH QRUPDOL]HG WR VDWLVI\\ WKH HTXDWLRQ DQG SS a</td></tr><tr><td>,Q HDFK H[SHULPHQW WKH +00 WUDLQLQJ XVLQJ WKH IRUZDUGEDFNZDUG DOJRULWKP ZLOO EH WHUPLQDWHG 6 ( /HYLQVRQ / 5 5DELQHU DQG 0 0 6RQGKL \u00b3$Q LQWURGXFWLRQ WR WKH DSSOLFDWLRQ RI WKH WKHRU\\ RI $ $</td></tr><tr><td>ZKHQ WKH LQFUHDVH RI WKH DYHUDJH ORJ SUREDELOLW\\ OHVV WKDQ DQG 76+00 WUDLQLQJ ZLOO EH WHUPLQDWHG DIWHU LWHUDWLRQV SUREDELOLVWLF IXQFWLRQV RI D 0DUNRY SURFHVV WR DXWRPDWLF VSHHFK UHFRJQLWLRQ\u00b47KH %HOO 6\\VWHP VW URZ RI QG URZ RI VW URZ RI 7HFKQLFDO -RXUQDO $SULO SS a PDWUL[ $ PDWUL[ $ PDWUL[ %</td></tr><tr><td>,Q WKLV SDSHU ZH FRPSDUHG WKH +00V WUDLQHG E\\ WKH WDEX VHDUFK DOJRULWKP DQG WKH IRUZDUG ) *ORYHU DQG 0 /DJXQD \u00b37DEX VHDUFK\u00b4.OXZHU $FDGHPLF 3XEOLVKHUV</td></tr><tr><td>EDFNZDUG DOJRULWKP UHVSHFWLYHO\\ 6LPXODWLRQ UHVXOWV DUH VKRZQ LQ 7DEOH 7KH\\ DUH PDGH XS RI WZR , , -, , , ----</td></tr><tr><td>SDUWV V 3 DQG G 3 V 3 GHQRWHV WKH DYHUDJH ORJ SUREDELOLW\\ RI WKH +00 JHQHUDWHG E\\ WKH WUDLQLQJ</td></tr><tr><td>REVHUYDWLRQ VHTXHQFHV RI WKLV +00 DQG G 3 GHQRWHV WKH DYHUDJH ORJ SUREDELOLW\\ RI WKH +00 JHQHUDWHG )LJ 7KH VWULQJ UHSUHVHQWDWLRQ RI +00</td></tr><tr><td>E\\ WKH RWKHU WUDLQLQJ REVHUYDWLRQ VHTXHQFHV RI WKH RWKHU +00V</td></tr><tr><td>L 7DEOH 7KH FRPSDULVRQ RI DYHUDJH ORJ SUREDELOLW\\ REWDLQHG ZLWK WZR DOJRULWKPV DQG JR WR VWHS RWKHUZLVH UHFRUG WKH EHVW SDWK LQGH[ DQG WHUPLQDWH WKH DOJRULWKP 76 )RUZDUG%DFNZDUG ([SHULPHQW $V DOJRULWKPV 3 8 3 / 3 8 3 /</td></tr></table>",
+                "text": "SHULPHQWV DUH FRQGXFWHG WR YDOLGDWH WKH DOJRULWKP SURSRVHG LQ WKLV SDSHU :H UHFRUGHG HDFK ZRUG \u00b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
+                "html": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1008.json

@@ -0,0 +1,182 @@
+{
+    "paper_id": "O01-1008",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:34.543705Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O01-1008",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "!\"#$ !\"#$ !\"#$ !\"#$ :\u00f36\u00ee.%Z=>7=#.*4 . ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "-",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "?./0)* )* )* )*v v v v \u00b2A \u00b2A \u00b2A \u00b2A. . . .\u00b33 \u00b33",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u00dc>&FL\u00bd+\u00b2A\u00a1?:.\u00b3\u00b4\u00ba\u00bbq\u00bbOP ^.\u00dd7\u00bc@\u00b8>Dw:\u00bd\u00be7\u00bfm\u00c0",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u00c0\u00f3\u00ac\u00f2\u00bc\u00c1\u00c2\u00be\u00a8s.w\u00df\u00c3\u00c4\u00bc\u00c1\u00c2\u00be\u00f8&\u00c57",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Aa\u00d0\u00a1?]AEB\u00be\u00f26\u00ee\u00b4=\u00a5{\u00c7\u00f8\u00ed\u00b3 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u2022OP\u00c8q\u00a5X \u00a8!\u00eauV \u00c9\u00c77mae\u00f8uV 6_\u00dc{\u00ca\u00ed_\u00cbR\u00c77\u00b3\u00aa *0-! m u V \u00ed _ \u00b3 \u00aa [ R : * , ,;-\u00c96 :\u00ea\u00cc\u00ed_t\u00bd\u00ca\u00b3\u00aa\u00ea\u00cc\u00ed_ t\u00fa\u00f2\u00cd\u00ac7A\u00ce*445:\u00ce\u00a9R -\u00dcq\u00fd\u00b3\u00aa\u00cbR\u00c770 *0.-\u00b4! muV\u00ed_0[R : * , ,;-\u00cf * -R*\u00b4 \u00b3\u00aa\u00a5tOPA\u00d0 R \u00b3\u00b42_t *3 34$-R> ( ) ( ) 0 \u00d7 \u2211 = = * - \u00c9l\u00a2>. : % \u00a4l\u00eb_ \u00a4l\u00ed_ q\u00a5 \u00a7 _tR ! 0 0 0 0 = + + + \u00bd>\u00a1?:ae\u00f8% \u00a866 \u00d1]mn OP v v v",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u00dc \u00c7 7 v & \u00a5 ae \u00e8 \u00f2 \u00c7 303 * _`R =303 U />303 -\u00ea!&F>\u00f07 \u00fcv \u00fd H v ] : \u00fe R =303 \u00ea ! \u00f5 \u00ed _ M a \u00a4 l v v v v\u00f8 \u00f8 \u00f8 \u00f8? ? ? ? \u00a9 \u00a9 \u00a9 \u00a9R R R R > > > >& & & &F F F F \u00d4 \u00d4 \u00d4 \u00d4 v v v v\u00f7 \u00f7 \u00f7 \u00f7? ? ? ? \u00a9 \u00a9 \u00a9 \u00a9R R R R > > > >& & & &F F F F \u00d4 \u00d4 \u00d4 \u00d4 />303 \u00ea!\u00ed_M!\u00ff#\u00fe\u00f3# />303 \u00ea!\u00ed_M\u00bdOP\u00f4\u00ea\u00ef\u00fe\u00bd \u00ed_\u00b3\u2022]\u00c2\u00fe \u00b3\u2022nTR\u00eb_\u00d4v\u00fdGvv-\u00d5Rv\u00f7Uv\u00f8&\u00b5 \u00b6 >&F\u00ea!\u00d4R\u00f5.v \u00ac& \u00ac& \u00ac& \u00ac&\u00b3\u2022 \u00b3\u2022 \u00b3\u2022 \u00b3\u2022DE; DE; DE; DE;",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": ".\u00b3\u00b46\u00eeae\u00f8BC% o$9;\u00c8\u00f2DE #0 :*.\u00b3\u00b4ae\u00f8-@ A :#0B#0B%@ \u5c0d.\u00b3\u00b4:\u00a97\u00ed_\u5747\u505a\u5982\u3198\u6aa2\u67e5>",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u00cb\u00ed_RH?R #0 c\u00be\u00b3\u00b4 \u82e5 c\u00be\u00b3\u00b4:\u00ed_t 9:A*\u00d5\u00b4\u00b3\u00b4:%7nR\u00ed_-",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u5247 \u00fb\u00ed_K]\u00bd\u00ed\u00b3\u2022 T\u00eb \u7d50\u675f v7lvR]DE\u00ac&\u00b3\u2022L\u00d4> 3,-./012 3,-./012 3,-./012 3,-./012 DEBC\u00be;\u00d1|%BDE& BC\"o\u00b3\u2022ae v v v v\u00fd \u00fd \u00fd \u00fd? ? ? ?&\u00b5 \u00b6 &\u00b5 \u00b6 &\u00b5 \u00b6 &\u00b5 \u00b6> > > >& & & &F F F Fl l l l v v v v v v v v--- -?\u00a8v ?\u00a8v ?\u00a8v ?\u00a8v\u00f7 \u00f7 \u00f7 \u00f7Kv Kv Kv Kv\u00f8&\u00b5 \u00b6 \u00f8&\u00b5 \u00b6 \u00f8&\u00b5 \u00b6 \u00f8&\u00b5 \u00b6> > > >& & & &F F F F \u00d4 \u00d4 \u00d4 \u00d4 \u00f8BC\u00f2WEv7lv.\u00a3$%7\u00fdBCv",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": ".\u00dd7\u00bc@\u00b8#>\u00f2\u00f3\u00a1?SEaTUuVaB Cv\u00f3q\u00bc@\u00b8\u00fb\u00c77.&F>\u00f07",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u00efl6\u00ee^ae\u00ed_\u00d5R. 3 i\u00fe* -#q : ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": ";DE:BC +>DEBC!a\u00b8\u00b9\u00e0>R\u00c77\u00b4LZo$_ '(\u00b8\u00b9\u00a6K>DEBC\u00be; \u00a6 \u00a6 \u00a6 \u00a6 \u00b81 1 1 \u00b9 \u00a6\u00b8\u00b9OP\u00ea\u00ef^\\V.#\u00d1\u00d2 \u00a6v#ROPa\u00dc\u00a97v:\u00ea%\u00ed\u00b3\u2022\\uVl \u00dal7vL\u00dc]\\uVAE7\\uV9\u00b9\u00f9\u00bduV\u00f2\u00c7\u00b9^\u00cd \u00ac 7 \u00aa \u00ab 3 03 \u00dc ]\u00b8\u00b9 R e \u00a6 _ 4 5 ! 3 03 :%\" %\u00bb\u00d4 v v v v7 7 7 7?\u00ac& ?\u00ac& ?\u00ac& ?\u00ac&\u00b3\u2022 \u00b3\u2022 \u00b3\u2022 \u00b3\u2022DE DE DE DE \u00d4 \u00d4 \u00d4 \u00d4\u00a2 \u00a2 \u00a2 \u00a2 \u00ac&\u00ed\u00b3\u2022 &0=0 v v v v",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": ".\u00b3R% \u00b3\u00b4\u00f36\u00ee'R % ! *0.-\u00a8s :.F&c\u00be\u00b3\u00b4\"c#F!Z!\u00f8*\u00d5\u00f8- \u00f3&#is\u00d4.vae\u00f8 &0=0 *\u00d5ae\u00f8-OP\u00be; l>\u00bdZ!$\u00b0R &0=0B%!8 &0=0B \u00b4Z !G67\u00a5Fv#&\u00bblv\u00d5#FGH\u00d4\u00a2 <# \u00cc\u00cd\u00e7+_ K (( Z=a( !% %%& \u00cc\u00cd'r \u00de\u00df \u00b6#9()/0\u00f5^ % % (( \u00cc\u00cd 1**Ua (( /0\u00f2\u00e7+_ /+K*_\u00f7Z=\u00c9. :\u00ea# /+R <E %\u00ea56\u00be; v#+R`a\u00dc\u00da>\u00dbi_R \" \u00b3\u2022\u00a8\u00a97\u00b3\u2022:_\u00b1, -.]lHG\u00d2\u00a5\u00be<\u00d4q\u00b3\u2022:\u00be<%\u00ed_@} R R %\u00b47\"!\u00f5 \" \u00a5* \" \u00b3\u2022B \u00a5\u00be<-@} v v v v#& #& #& #&?FGH ?FGH ?FGH ?FGH \u00d4 \u00d4 \u00d4 \u00d4\u00a2 \u00a2 \u00a2 \u00a2 v v v v#+ #+ #+ #+? ? ? ? > > > >@} @} @} @} OP/[^7\u00fd.v9RC\u00e7\u00a29\u00b9r\u00dcq\u00fdv @ }\u00bel7\u00a5.>+9/0`aA6]@}e C\u00e7\u00a2:@}9\u00b9rOP \u00a8\u00c7@}\u00a5t9R _\u00c7\u00c8 r@}\u00c7\u2022\u00f87_\u00b3\u00b4\u2022\u00f8_t\u00f3s % \u00f5 \" _t0ae\u00a4Z\u00a5>0\u2022 l?17\u00a59 \u00a7>\u00a9%7.>R _`UC\u00e7\u00a2:\u00d7\u00af\u00da \u00d72\u00da9\u00b9r]@}_`l> ? 2? \u00b9r\u00d4C\u00e7\u00a2\u00d7\u00af\u00da\u2022\u00f8R \"% _\u00d72\u00da\u2022\u00f8R % _\u00b3\u00b4\u00d7\u00af\u00daR\u00b4.>%734 v#\u00d2#> \u00d2\u00ac <# /0L\u00ea!AE#",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "; ; ; ; < < < <\u00da \u00da \u00da \u00da A<A$/ <##A ##<# &33 &. \u00bbVa \" !*!- %* \" -!*- PVa \" *!-%* -!*- % 127 \" \"*!-% * -!*- 12& \" * !-* -!*- 12+ \" \"*-* -!*- \u00a4\u00f8 \u00a4\u00f8 \u00a4\u00f8 \u00a4\u00f8? ? ? ? LMN LMN LMN LMN \u00d4 \u00d4 \u00d4 \u00d4 \u00d7 \u00d7 \u00d7 \u00d7= = = =K K K K>? >? >? >?\u00da \u00da \u00da \u00da A<A$/ <##A ##<# &33 &. \u00bbVa !% \" * -\"* -*! - PVa !% \" * -* -*! - % 127 !% \"*-\"*- *! - % 12& !% \"%*-!*!\"- *! - 12+ !% \"*-*! - *! - \u00a4\u00d2\u00f5\u00a4\u00f8\u00e2\u00ee12+#\u00ea%\u00be;:\u00d4\u00bb\u00d3[\\VW % U cd^[VWb\u00f5 \u00dd\\\u00b3RC\u00e7\u00a2R/[ ,\" \u00a5[\\VW:_`% U ! .;9\u00b9rfgC\u00e7\u00a2hr#\u00c2+dQ7 \u00dc <# LMN12+45\u00d4\u00c29 <# LMN*\u00bbVa-",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "12+\u00bb \u00bb \u00bb \u00bbV V V Va a a a \u00d4\u00b9\u00f9_\u00f7 \u00d4\u00b9\u00f9_\u00f7 \u00d4\u00b9\u00f9_\u00f7 \u00d4\u00b9\u00f9_\u00f7 A # AA AD DA DD 7\u00b389 % !* - \"*%!- *%- * - :;< \"! !% *- *%- *- !* - =K>?",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\"! \"%*!-*%-\"*-%*!-:;<=>!\" :;<=>!\" :;<=>!\" :;<=>!\" \u00a8aAE6$\u00ea\u00d7. <# \u00be;OP\u00d1\u00d22\u00dc:3 6 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u00bdX\u00bf\u00c0 \u00b6#\u00fb\u00fc~\u00bf\u00c0\u00e2\u00e3N~ ~ ~ ~ ~ v",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {},
+        "ref_entries": {
+            "FIGREF0": {
+                "type_str": "figure",
+                "text": "!\" .. /,=.0@,))2,<)*%%%-5$ A1 )",
+                "num": null,
+                "uris": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1009.json

@@ -0,0 +1,231 @@
+{
+    "paper_id": "O01-1009",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:38.651715Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "x\u00c4!\u00a8\u00edjt\u00d6\u00db\u00eb\u00de\u00c45\u00fc\u00d69\u00eb)\u00ffy\u00c3H\u00ff\u00a9 8<\u00eb\u00fc-5#Cx\u00bd\u00ac!&]\u00d6\u00db\u00ea\u00ea5\u00e2\u00ac\u00bd\u00ee\u00cb\u0229 \u00e9#CN'\u00ee&\u00d6\u00db|R \u00bf\u00eb\u00de|RV\u00eb\u00d1qC'x9 FN\u00db*Y\u00b8\u00d6\u00db#\u00ebW\u00c9|R\u00d6V\u00ed#CV\u00eb\u00d1q\u00d27 \u00b1#C\u00a3*'\u00ee\u00c9[x\u00ae*g\u00fcn\u00d1q\u00b0#C\u00d6\u00db\u2022Y\u00ed#C\u00d6\u00db z\u00f8y\u00d6\u00bd\u00da\u00eb\u00b8\u00d1q\u00d27x\u00d6\u00bd\u00da\u00d6V\u00eb\u00bfs\u00bd\u00daX]\u00ae \u00b0N\u00ebW\u00de[T\u00cag\u00b4\u00c45F\u00e0Bxx(\u00fb\u00edz\u00f8y\u00d6\u00ee \u00ed\u00eb\u00d1q\u00efoy\u00aaY\u00b8\u00a8\u00ed\u00f2\u2022R\u00eb\u00acf&\u00a8\u00ed*\u00be |\u00eb\u00fb\u00eb\u00de\u00f6xL|;\u00f3\u00edL\u00f8y\u2022D\u00eb\u00d1q\u00d27\u00f0Y8\u00d8 \u2022D\u00d6V\u00ebW?\u00d4yHwx\u00d6\u00dbecM,g17|\u00ebi\u00eb\u00b0 \u00ec!X\u00ae|\u2022Y\u00ed &\u00d6 \u00d6\u00db\u00b8xdpsqvt.cbtfe\u00eb)y\u00bf\u00fag\u00d69\u00d9\u00ee7\u00d69\u00eb)xP5 \u00cdDivsdi-Nfsdfs:4Difo:5Ivboh:6\u00ed C!\u00a8\u00edjt\u00d6\u00dbyxF\u00ac\u2022*\u00ebWy5\u00fc\u00d69\u00eb)n,\u2022 *jogsbtusvduvsf\u00edx\u00ffHS\u00ef\u00d6\u00db\u00eb\u00eci\u00ebMPC\u00ebMpoepo.Mvoe 23 \u00ac\u00bd\u00eb)7\u00aa\u00f5+\u00ed \u00c4^\u00eb)jt\u00d6\u00db 8w\u00c4\u00ebjt\u00d6\u00db Tjojdb Dpsqvt \u00eby\u00d1 zx\u00b0\u00f5!\u00a8\u00ed>\u00d6jt\u00d6\u00db\u00ed$ 2::5 \u00fa\u00dd\u00c9eFf\u00da]\u00d9\u00c3\u00eb)c \u00f0'\u00b1\u00deco*\u00c4\u00aa\u00da\u00af\u00eb\u00de\u00b0\u00a8\u00cc\u00aa\u00f2H'1o5||\u00ed2::8 \u00fa\u00c9\" \u00eb)\u00d6\u00db 4/1 w\u00b9\u00edi\u00cd$\u00db7\u00cc\u00aa\u00ed\u00d1q\u00b1\u00ff\u00de 3114 \u00fa\u00bfsi \u00a6\u00db\u00cc\u00aa\u00ed C#x\u00c4jt!\u00a8\u00ed\u00d6\u00db\u00eb<k\u00d6\u00da\u00ec\u00d6\u00f5R)<\u00d6 %\u00bc\u00ec\u00b8\u00d6\u00e0!\u00ec:!\u00a8\u00ed22*\u00ec\u2022\u00ca\u00ed\u00ac\u00b1C#\u00edD\u00c4\u00eb$\u00a9 %\u00d4u\u2022w\u00eb\u00d1qxm'\u00f12\u00b1\u00d1q'\u00c5\u00fbY\u00b8\u00d6\u00b2 9F\u00eb\u00b4ae<\u00fe\u00c3\u00ccx\u00ebx(\u00fb\u00fb'{\u00d4\u00ee\u00dex \u00c7\u00f2\u00bb\u00c7\u00ed\u00e4\u00a3'\u00b89F#\u00f0\u00f5F\u00d6|R\u00ee7\u00e3\u00b4\u00c3\u00d8\u00eb\u00cc\u00f0 \u2022\u00f0\u2022Y\u00e0-9\u00eb_ o\u00fb\u00ed)3*\u00ac\u00bd\u00d6\u00bd\u00da\u00ec\u00bf\u00ec\u00e0!\u00ec\u00ca\u00a3 \u00d1'\u00fb:'\u2022R\u00ebW\u00fb\u2022R\u00d6V\u00eb\u00c5|R\u00ed\u00ed)4*oG\u00d1 qc\u00db\u00b25@\u00c9*Difo-Mjv:3\u00d6:'\u00a8\u00ed\u00eb\u00fcS$ x\u00b4\u00c4y\u00ef\u00de\u00ebc\u00aa*\u00bbS\u00acfU\u00eb-o\u00e1oZL\u00ac\u00bdg :'\u00d4\u00e0\u00ed5\u2022|\u00eb$'\u00ee!\u00a8\u00edK]S\u00eb| \u2022wrD\u00e3[\u00ff\u00fbf\u00eb\u00eei\u00ebc\u00aay\u00a8\u00ed \u00f0]\u00d9-]\u00ed\u00e4[Y\u00ee\u00de|R\u00c7]\u00eb\u00cc\u00de\u2022|o*\u01de \u00ac\u00bd\u00ee\u00cb\u00eb\u00de\u00a8\u00ed\u00fb\u00c3'd\u00ed \u00d9\u00eb\u00d1q\u00d8\u00b85NC#x&]jt\u00d6\u00db\u00ebo\u00e158\u00ca7\u00b1;",
+    "pdf_parse": {
+        "paper_id": "O01-1009",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "x\u00c4!\u00a8\u00edjt\u00d6\u00db\u00eb\u00de\u00c45\u00fc\u00d69\u00eb)\u00ffy\u00c3H\u00ff\u00a9 8<\u00eb\u00fc-5#Cx\u00bd\u00ac!&]\u00d6\u00db\u00ea\u00ea5\u00e2\u00ac\u00bd\u00ee\u00cb\u0229 \u00e9#CN'\u00ee&\u00d6\u00db|R \u00bf\u00eb\u00de|RV\u00eb\u00d1qC'x9 FN\u00db*Y\u00b8\u00d6\u00db#\u00ebW\u00c9|R\u00d6V\u00ed#CV\u00eb\u00d1q\u00d27 \u00b1#C\u00a3*'\u00ee\u00c9[x\u00ae*g\u00fcn\u00d1q\u00b0#C\u00d6\u00db\u2022Y\u00ed#C\u00d6\u00db z\u00f8y\u00d6\u00bd\u00da\u00eb\u00b8\u00d1q\u00d27x\u00d6\u00bd\u00da\u00d6V\u00eb\u00bfs\u00bd\u00daX]\u00ae \u00b0N\u00ebW\u00de[T\u00cag\u00b4\u00c45F\u00e0Bxx(\u00fb\u00edz\u00f8y\u00d6\u00ee \u00ed\u00eb\u00d1q\u00efoy\u00aaY\u00b8\u00a8\u00ed\u00f2\u2022R\u00eb\u00acf&\u00a8\u00ed*\u00be |\u00eb\u00fb\u00eb\u00de\u00f6xL|;\u00f3\u00edL\u00f8y\u2022D\u00eb\u00d1q\u00d27\u00f0Y8\u00d8 \u2022D\u00d6V\u00ebW?\u00d4yHwx\u00d6\u00dbecM,g17|\u00ebi\u00eb\u00b0 \u00ec!X\u00ae|\u2022Y\u00ed &\u00d6 \u00d6\u00db\u00b8xdpsqvt.cbtfe\u00eb)y\u00bf\u00fag\u00d69\u00d9\u00ee7\u00d69\u00eb)xP5 \u00cdDivsdi-Nfsdfs:4Difo:5Ivboh:6\u00ed C!\u00a8\u00edjt\u00d6\u00dbyxF\u00ac\u2022*\u00ebWy5\u00fc\u00d69\u00eb)n,\u2022 *jogsbtusvduvsf\u00edx\u00ffHS\u00ef\u00d6\u00db\u00eb\u00eci\u00ebMPC\u00ebMpoepo.Mvoe 23 \u00ac\u00bd\u00eb)7\u00aa\u00f5+\u00ed \u00c4^\u00eb)jt\u00d6\u00db 8w\u00c4\u00ebjt\u00d6\u00db Tjojdb Dpsqvt \u00eby\u00d1 zx\u00b0\u00f5!\u00a8\u00ed>\u00d6jt\u00d6\u00db\u00ed$ 2::5 \u00fa\u00dd\u00c9eFf\u00da]\u00d9\u00c3\u00eb)c \u00f0'\u00b1\u00deco*\u00c4\u00aa\u00da\u00af\u00eb\u00de\u00b0\u00a8\u00cc\u00aa\u00f2H'1o5||\u00ed2::8 \u00fa\u00c9\" \u00eb)\u00d6\u00db 4/1 w\u00b9\u00edi\u00cd$\u00db7\u00cc\u00aa\u00ed\u00d1q\u00b1\u00ff\u00de 3114 \u00fa\u00bfsi \u00a6\u00db\u00cc\u00aa\u00ed C#x\u00c4jt!\u00a8\u00ed\u00d6\u00db\u00eb<k\u00d6\u00da\u00ec\u00d6\u00f5R)<\u00d6 %\u00bc\u00ec\u00b8\u00d6\u00e0!\u00ec:!\u00a8\u00ed22*\u00ec\u2022\u00ca\u00ed\u00ac\u00b1C#\u00edD\u00c4\u00eb$\u00a9 %\u00d4u\u2022w\u00eb\u00d1qxm'\u00f12\u00b1\u00d1q'\u00c5\u00fbY\u00b8\u00d6\u00b2 9F\u00eb\u00b4ae<\u00fe\u00c3\u00ccx\u00ebx(\u00fb\u00fb'{\u00d4\u00ee\u00dex \u00c7\u00f2\u00bb\u00c7\u00ed\u00e4\u00a3'\u00b89F#\u00f0\u00f5F\u00d6|R\u00ee7\u00e3\u00b4\u00c3\u00d8\u00eb\u00cc\u00f0 \u2022\u00f0\u2022Y\u00e0-9\u00eb_ o\u00fb\u00ed)3*\u00ac\u00bd\u00d6\u00bd\u00da\u00ec\u00bf\u00ec\u00e0!\u00ec\u00ca\u00a3 \u00d1'\u00fb:'\u2022R\u00ebW\u00fb\u2022R\u00d6V\u00eb\u00c5|R\u00ed\u00ed)4*oG\u00d1 qc\u00db\u00b25@\u00c9*Difo-Mjv:3\u00d6:'\u00a8\u00ed\u00eb\u00fcS$ x\u00b4\u00c4y\u00ef\u00de\u00ebc\u00aa*\u00bbS\u00acfU\u00eb-o\u00e1oZL\u00ac\u00bdg :'\u00d4\u00e0\u00ed5\u2022|\u00eb$'\u00ee!\u00a8\u00edK]S\u00eb| \u2022wrD\u00e3[\u00ff\u00fbf\u00eb\u00eei\u00ebc\u00aay\u00a8\u00ed \u00f0]\u00d9-]\u00ed\u00e4[Y\u00ee\u00de|R\u00c7]\u00eb\u00cc\u00de\u2022|o*\u01de \u00ac\u00bd\u00ee\u00cb\u00eb\u00de\u00a8\u00ed\u00fb\u00c3'd\u00ed \u00d9\u00eb\u00d1q\u00d8\u00b85NC#x&]jt\u00d6\u00db\u00ebo\u00e158\u00ca7\u00b1;",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "DQG 5 / 0HUFHU \u00b3,QWURGXFWLRQ WR WKH 6SHFLDO ,VVXH RQ &RPSXWDWLRQDO /LQJXLVLWFV 8VLQJ /DUJH &RUSRUD\u00b4&RPSXWDWLRQDO /LQJXLVWLFV 9RO 1R SS",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "&KXUFK . : DQG 5 / 0HUFHU \u00b3,QWURGXFWLRQ WR WKH 6SHFLDO ,VVXH RQ &RPSXWDWLRQDO /LQJXLVLWFV 8VLQJ /DUJH &RUSRUD\u00b4&RPSXWDWLRQDO /LQJXLVWLFV 9RO 1R SS",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "&KHQ .HKMLDQQ 6KLQJKXDQ /LX /LSLQJ &KDQJ DQG <HK+DR &KLQ \u00b3$ 3UDFWLFDO 7DJJHU IRU &KLQHVH &RUSRUD\u00b43URFHHGLQJV RI 52&/,1* 9",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "&KHQ .HKMLDQQ 6KLQJKXDQ /LX /LSLQJ &KDQJ DQG <HK+DR &KLQ \u00b3$ 3UDFWLFDO 7DJJHU IRU &KLQHVH &RUSRUD\u00b43URFHHGLQJV RI 52&/,1* 9,, SS",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "WUDFW\u00b4&RPSXWDWLRQDO /LQJXLVWLFV 9RO 1R SS",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "WUDFW\u00b4&RPSXWDWLRQDO /LQJXLVWLFV 9RO 1R SS",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "LYDVLORJORX 9 \u00b37UDQVODWLQJ &ROORFDWLRQV IRU %LOLQJXDO /H[LFRQV $ 6WDWLVWLFDO $SSURDFK\u00b4&RPSXWDWLRQDO /LQJXLVWLFV 9RO 1R",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [
+                            "Hrzq"
+                        ],
+                        "last": "Udqn 0f",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Dqg +dw",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "UDQN 0F.HRZQ .U DQG +DW]LYDVLORJORX 9 \u00b37UDQVODWLQJ &ROORFDWLRQV IRU %LOLQJXDO /H[LFRQV $ 6WDWLVWLFDO $SSURDFK\u00b4&RPSXWDWLRQDO /LQJXLVWLFV 9RO 1R",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "QIRUPDWLRQ DQG 5HODWLYH )UHTXHQF\\ &RXQW\u00b43URFHHGLQJV RI 52&/,1* 9, 1DQWRX 7DLZDQ 52& 6HS SS 6SURDW 5LFKDUG DQG 6KLQ &KLOLQ \u00b3$ 6WDWLVWLFDO 0HWKRG )RU )LQGLQJ :RUG %RXQGDULHV ,Q &KLQHVH 7H[W\u00b4&RPSXWHU 3URFHVVLQJ RI &KLQHVH 2ULHQWDO /DQJXDJH 9RO 1R 0DUFK",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "&kdqj -Lqj 6klq \u00b3$xwrpdwlf /H",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "&KDQJ -LQJ 6KLQ \u00b3$XWRPDWLF /H[LFRQ $FTXLVLWLRQ DQG 3UHFLVLRQ5HFDOO 0D[LPL]DWLRQ IRU 8QWDJJHG 7H[W &RUSRUD\u00b41DWLRQDO 7VLQJ+XD 8QLYHUVLWK 3KG WKHVLV :X 0 : DQG 6X . < \u00b3&RUSXVEDVHG $XWRPDWLF &RPSRXQG ([WUDFWLRQ ZLWK 0XWXDO ,QIRUPDWLRQ DQG 5HODWLYH )UHTXHQF\\ &RXQW\u00b43URFHHGLQJV RI 52&/,1* 9, 1DQWRX 7DLZDQ 52& 6HS SS 6SURDW 5LFKDUG DQG 6KLQ &KLOLQ \u00b3$ 6WDWLVWLFDO 0HWKRG )RU )LQGLQJ :RUG %RXQGDULHV ,Q &KLQHVH 7H[W\u00b4&RPSXWHU 3URFHVVLQJ RI &KLQHVH 2ULHQWDO /DQJXDJH 9RO 1R 0DUFK",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "RUG ,GHQWLILFDWLRQ IRU 0DQGDULQ &KLQHVH 6HQWHQFHV \u00b1 $ 8QLILFDWLRQ $SSURDFK\u00b4&RPSXWHU 3URFHVVLQJ RI &KLQHVH 2ULHQWDO /DQJXDJHV 9RO 1R 0DUFK",
+                "authors": [
+                    {
+                        "first": "<hk &klqj/Rqj Dqg /Hh +lv-Ldq",
+                        "middle": [],
+                        "last": "\u00b35xoh%dvhg",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "<HK &KLQJ/RQJ DQG /HH +LV-LDQ \u00b35XOH%DVHG :RUG ,GHQWLILFDWLRQ IRU 0DQGDULQ &KLQHVH 6HQWHQFHV \u00b1 $ 8QLILFDWLRQ $SSURDFK\u00b4&RPSXWHU 3URFHVVLQJ RI &KLQHVH 2ULHQWDO /DQJXDJHV 9RO 1R 0DUFK",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "/LQ 0< &KDQJ 7 + DQG 6X . < \u00b3$ SUHOLPLQDU\\ VWXG\\ RQ XQNQRZQ ZRUG SUREOHP LQ &KLQHVH ZRUG VHJPHQWDWLRQ\u00b43URFHHGLQJV RI 52& &RPSXWDWLRQDO /LQJXLVWLFV &RQIHUHQFH 7DLZDQ SS",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "/LQ 0< &KDQJ 7 + DQG 6X . < \u00b3$ SUHOLPLQDU\\ VWXG\\ RQ XQNQRZQ ZRUG SUREOHP LQ &KLQHVH ZRUG VHJPHQWDWLRQ\u00b43URFHHGLQJV RI 52& &RPSXWDWLRQDO /LQJXLVWLFV &RQIHUHQFH 7DLZDQ SS",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "RUG 'HWHFWLRQ IRU &KLQHVH E\\ D &RUSXVEDVHG /HDUQLQJ 0HWKRG\u00b43URFHHGLQJV RI 52&/,1* ; 7DLSHL 7DLZDQ 52& SS &KHQ .HK -LDQQ DQG /LX 6KLQJ +XDQ \u00b3:RUG ,GHQWLILFDWLRQ IRU 0DQGDULQ &KLQHVH 6HQWHQFHV\u00b43URFHHGLQJV RI &2/,1* YRO",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "&khq",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Hk -Ldqq %dl 0lqj +rqj \u00b38qnqrzq",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "&KHQ .HK -LDQQ %DL 0LQJ +RQJ \u00b38QNQRZQ :RUG 'HWHFWLRQ IRU &KLQHVH E\\ D &RUSXVEDVHG /HDUQLQJ 0HWKRG\u00b43URFHHGLQJV RI 52&/,1* ; 7DLSHL 7DLZDQ 52& SS &KHQ .HK -LDQQ DQG /LX 6KLQJ +XDQ \u00b3:RUG ,GHQWLILFDWLRQ IRU 0DQGDULQ &KLQHVH 6HQWHQFHV\u00b43URFHHGLQJV RI &2/,1* YRO , SS",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "RUGV\u00b43URFHHGLQJV RI WK 1DWXUDO /DQJXDJH 3URFHVVLQJ 3DFLILF 5LP 6\\PSRVLXP1/356 \u00b6 SS",
+                "authors": [
+                    {
+                        "first": "&khq &-0+ %dl .-&khq \u00b3&dwhjru\\ *xhvvlqj Iru &klqhvh",
+                        "middle": [],
+                        "last": "8qnqrzq",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "&KHQ &-0+ %DL .-&KHQ \u00b3&DWHJRU\\ *XHVVLQJ IRU &KLQHVH 8QNQRZQ :RUGV\u00b43URFHHGLQJV RI WK 1DWXUDO /DQJXDJH 3URFHVVLQJ 3DFLILF 5LP 6\\PSRVLXP1/356 \u00b6 SS",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "\u00b3m\u00ae \u00ca\u00eb\u00b0?\u00b1\u00b4&.,3 7HFKQLFDO 5HSRUW QR T[\" \u00b3m\u00d2\u00c3\u00b1SrV\u00d4\u00b0%\u00e1\u00b1\u00b4&.,3 7HFKQLFDO 5HSRUW QR",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "T[\" \u00b3m\u00ae \u00ca\u00eb\u00b0?\u00b1\u00b4&.,3 7HFKQLFDO 5HSRUW QR T[\" \u00b3m\u00d2\u00c3\u00b1SrV\u00d4\u00b0%\u00e1\u00b1\u00b4&.,3 7HFKQLFDO 5HSRUW QR",
+                "links": null
+            }
+        },
+        "ref_entries": {}
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1010.json

@@ -0,0 +1,1930 @@
+{
+    "paper_id": "O01-1010",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:08:49.873744Z"
+    },
+    "title": "Design, Compilation and Processing of CUCall: A Set of Cantonese Spoken Language Corpora Collected Over Telephone Networks",
+    "authors": [
+        {
+            "first": "W",
+            "middle": [
+                "K"
+            ],
+            "last": "Lo",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "The Chinese University of Hong Kong",
+                "location": {}
+            },
+            "email": "wklo@ee.cuhk.edu.hk"
+        },
+        {
+            "first": "P",
+            "middle": [
+                "C"
+            ],
+            "last": "Ching",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "The Chinese University of Hong Kong",
+                "location": {}
+            },
+            "email": "pcching@ee.cuhk.edu.hk"
+        },
+        {
+            "first": "Tan",
+            "middle": [],
+            "last": "Lee",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "The Chinese University of Hong Kong",
+                "location": {}
+            },
+            "email": "tanlee@ee.cuhk.edu.hk"
+        },
+        {
+            "first": "Helen",
+            "middle": [],
+            "last": "Meng",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "The Chinese University of Hong Kong",
+                "location": {}
+            },
+            "email": "hmmeng@se.cuhk.edu.hk"
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "The design and compilation of the CUCall telephone speech corpora is described in this paper. Speech database is an indispensable resource for research and development of state-of-the-art spoken language technology. These speech recognition systems rely greatly on a huge amount of well-designed and appropriately processed speech data for parameters training. On the other hand, as telephony applications are becoming more demanding and complicated, natural language interface is gaining more popularity than the traditional touch tone operation. Therefore, large telephone speech databases are required for such system building. Separate speech corpora are needed for telephone systems since there exist significant differences due to the channel difference. In this paper, we will describe the design and processing of a set of spoken language corpora for Cantonese that are collected over fixed line as well as mobile telephone networks. The corpora are intended as a versatile set of training data for general purpose application systems that adopt a statistical approach to spoken language processing. The designed set of corpora will be made up of over 1000 speaker calls.",
+    "pdf_parse": {
+        "paper_id": "O01-1010",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "The design and compilation of the CUCall telephone speech corpora is described in this paper. Speech database is an indispensable resource for research and development of state-of-the-art spoken language technology. These speech recognition systems rely greatly on a huge amount of well-designed and appropriately processed speech data for parameters training. On the other hand, as telephony applications are becoming more demanding and complicated, natural language interface is gaining more popularity than the traditional touch tone operation. Therefore, large telephone speech databases are required for such system building. Separate speech corpora are needed for telephone systems since there exist significant differences due to the channel difference. In this paper, we will describe the design and processing of a set of spoken language corpora for Cantonese that are collected over fixed line as well as mobile telephone networks. The corpora are intended as a versatile set of training data for general purpose application systems that adopt a statistical approach to spoken language processing. The designed set of corpora will be made up of over 1000 speaker calls.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "Speech data collected over telephone network is an essential resource for telephone based spoken language systems. The increasing penetration of remote system or service access over telephone networks has created a great driving force for collecting a huge amount of telephone speech data from a large speaker population and for different languages.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "Since the current state-of-the-art speech recognition techniques are statistically based, the availability of annotated data is particularly important. In general, the greater the amount and coverage of the data, the better the speech applications developed. In order to build a spoken language system over telephone network, the speech data has to be collected over telephone network and properly transcribed. The goal of this work 1 is to collect and compile a set of general purpose Cantonese telephone speech data from a large group of people of both genders. With the availability of this set of corpora, the rapid growth in the spoken language applications over telephone networks for the Cantonese speaking community is made possible.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "Over the past decades, many telephone based spoken language systems have been developed with great success. They all take advantage of the existence of several spoken language corpora compiled in recent years. Examples include the Jupiter from MIT [25] , HMIHY 2 from AT&T [17] and the European Union projects such as ACCeSS [27] and ARISE [28] etc. Nowadays, there are quite a large number of companies that make use of simple automatic telephone service systems to reduce the cost of employing human operators. Many of them have upgraded or wish to upgrade their touch-tone based system to speech enabled versions. It is obvious that continuous efforts are needed to enhance these services via speech technologies as much as possible.",
+                "cite_spans": [
+                    {
+                        "start": 248,
+                        "end": 252,
+                        "text": "[25]",
+                        "ref_id": "BIBREF25"
+                    },
+                    {
+                        "start": 273,
+                        "end": 277,
+                        "text": "[17]",
+                        "ref_id": "BIBREF17"
+                    },
+                    {
+                        "start": 325,
+                        "end": 329,
+                        "text": "[27]",
+                        "ref_id": "BIBREF27"
+                    },
+                    {
+                        "start": 340,
+                        "end": 344,
+                        "text": "[28]",
+                        "ref_id": "BIBREF28"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "For building telephone speech recognition systems, there has long been a great demand on Cantonese telephone speech data. This work is an initial effort to collect a set of Cantonese spoken language corpora over telephone network. It is targeted to provide some versatile data for public use. It aims to enrich the infrastructure for spoken language technology by providing the speech community with well-designed corpora in Cantonese.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "The compiled database will enable the integration of Cantonese speech technology to many of the existing telephone based interactive systems.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1"
+            },
+            {
+                "text": "There has been much effort in spoken language corpora development over the past decades.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "1-1 Background",
+                "sec_num": null
+            },
+            {
+                "text": "These include the TIMIT [8] , Resource Management [15] , Wall Street Journal [14] , Air Travel Information Service [16] etc. from the United States. In Europe, there are the EUROM1 [21] and SpeechDat [3] etc. They contain microphone data and telephone data as well. From the early adaptation of microphone corpora to network versions like NTIMIT [5] and the collection of real-world telephone data such as MACROPHONE [1] , CALLHOME [30] , SpeechDat [3] , POLYPHONE [29] etc., there is an abundant amount of data available for the western languages. The availability of these telephone corpora has successfully helped drive the research and development of telephone-based speech technologies of these languages.",
+                "cite_spans": [
+                    {
+                        "start": 24,
+                        "end": 27,
+                        "text": "[8]",
+                        "ref_id": "BIBREF8"
+                    },
+                    {
+                        "start": 50,
+                        "end": 54,
+                        "text": "[15]",
+                        "ref_id": "BIBREF15"
+                    },
+                    {
+                        "start": 77,
+                        "end": 81,
+                        "text": "[14]",
+                        "ref_id": "BIBREF14"
+                    },
+                    {
+                        "start": 115,
+                        "end": 119,
+                        "text": "[16]",
+                        "ref_id": "BIBREF16"
+                    },
+                    {
+                        "start": 181,
+                        "end": 185,
+                        "text": "[21]",
+                        "ref_id": "BIBREF21"
+                    },
+                    {
+                        "start": 200,
+                        "end": 203,
+                        "text": "[3]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 346,
+                        "end": 349,
+                        "text": "[5]",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 417,
+                        "end": 420,
+                        "text": "[1]",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 432,
+                        "end": 436,
+                        "text": "[30]",
+                        "ref_id": "BIBREF30"
+                    },
+                    {
+                        "start": 449,
+                        "end": 452,
+                        "text": "[3]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 465,
+                        "end": 469,
+                        "text": "[29]",
+                        "ref_id": "BIBREF29"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "1-1 Background",
+                "sec_num": null
+            },
+            {
+                "text": "For Asian languages, there has been limited investment spent on corpora development.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "1-1 Background",
+                "sec_num": null
+            },
+            {
+                "text": "Much effort came from Japan, for example those reported in [6, 7, 13] . For Chinese language, speech database collection has only started relatively recently. More widely used databases include microphone speech corpora such as the USTC95 [19] , HKU96 [2, 24] , HKU99 [4] , CMSC [22] and others [23] ; and telephone speech corpora such as MAT-160 and MAT-2000 [18, 20, 31] . These telephone data become valuable resources for many voice-activated telephony applications development.",
+                "cite_spans": [
+                    {
+                        "start": 59,
+                        "end": 62,
+                        "text": "[6,",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 63,
+                        "end": 65,
+                        "text": "7,",
+                        "ref_id": "BIBREF7"
+                    },
+                    {
+                        "start": 66,
+                        "end": 69,
+                        "text": "13]",
+                        "ref_id": "BIBREF13"
+                    },
+                    {
+                        "start": 239,
+                        "end": 243,
+                        "text": "[19]",
+                        "ref_id": "BIBREF19"
+                    },
+                    {
+                        "start": 252,
+                        "end": 255,
+                        "text": "[2,",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 256,
+                        "end": 259,
+                        "text": "24]",
+                        "ref_id": "BIBREF24"
+                    },
+                    {
+                        "start": 268,
+                        "end": 271,
+                        "text": "[4]",
+                        "ref_id": "BIBREF4"
+                    },
+                    {
+                        "start": 279,
+                        "end": 283,
+                        "text": "[22]",
+                        "ref_id": "BIBREF22"
+                    },
+                    {
+                        "start": 295,
+                        "end": 299,
+                        "text": "[23]",
+                        "ref_id": "BIBREF23"
+                    },
+                    {
+                        "start": 360,
+                        "end": 364,
+                        "text": "[18,",
+                        "ref_id": "BIBREF18"
+                    },
+                    {
+                        "start": 365,
+                        "end": 368,
+                        "text": "20,",
+                        "ref_id": "BIBREF20"
+                    },
+                    {
+                        "start": 369,
+                        "end": 372,
+                        "text": "31]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "1-1 Background",
+                "sec_num": null
+            },
+            {
+                "text": "Among the many Chinese dialects, Cantonese is one of the most popular Chinese dialects used in the southern China. Development of spoken language corpora has just started within the past decade [9, 10, 11] . It began with some small-scale corpus collection for specific projects. There is great shortage in Cantonese speech corpora to drive the growth and advancement of Cantonese speech technologies.",
+                "cite_spans": [
+                    {
+                        "start": 194,
+                        "end": 197,
+                        "text": "[9,",
+                        "ref_id": "BIBREF9"
+                    },
+                    {
+                        "start": 198,
+                        "end": 201,
+                        "text": "10,",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 202,
+                        "end": 205,
+                        "text": "11]",
+                        "ref_id": "BIBREF11"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "1-1 Background",
+                "sec_num": null
+            },
+            {
+                "text": "In 1997, the development of CUCorpora 3 [9, 11, 12] was initiated at the Chinese University of Hong Kong. CUCorpora is the first large-scale Cantonese spoken language corpora that are made available for public access. It is designed to cover both phonetically based content and common task oriented and application-specific content. The present work on telephone speech data compilation is a momentous extension of this effort. The vast variation of operator network protocols in Hong Kong 4 yet enrich the content of the corpora. Since the speakers will have to call our server to activate the data collection process, the resultant corpora are thus code named CUCall. The availability of the invaluable CUCall will undoubtedly nourish the booming technologies to a greater extent.",
+                "cite_spans": [
+                    {
+                        "start": 40,
+                        "end": 43,
+                        "text": "[9,",
+                        "ref_id": "BIBREF9"
+                    },
+                    {
+                        "start": 44,
+                        "end": 47,
+                        "text": "11,",
+                        "ref_id": "BIBREF11"
+                    },
+                    {
+                        "start": 48,
+                        "end": 51,
+                        "text": "12]",
+                        "ref_id": "BIBREF12"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "1-1 Background",
+                "sec_num": null
+            },
+            {
+                "text": "The paper is organized as follows. The design of the corpora materials will be described in detail first in Section 2. The design selection of the major parts of data will be elaborated.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "1-2 Paper organization",
+                "sec_num": null
+            },
+            {
+                "text": "After that, actual collection process is presented. From the recording system setup down to the collection process, every detail of the process will be given. In Section 4, the postprocessing of the captured data will be explained. The validation, transcription as well as the organization procedures are described. We will then provide some initial analysis on the designed corpora materials. Finally, conclusions are made in Section 6.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "1-2 Paper organization",
+                "sec_num": null
+            },
+            {
+                "text": "The design of the CUCall has been based on our previous experience with CUCorpora.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Corpora Design and Organization",
+                "sec_num": "2"
+            },
+            {
+                "text": "The concepts behind stay the same. Like CUCorpora, CUCall comprises of linguistically oriented and application-specific data. In CUCall, we take a step forward to include spontaneous conversations and short paragraphs data. These will altogether make up two major parts in the corpora: Sentences The sentences are chosen to be phonetically rich in the sense that they constitute complete coverage of bi-phone class context. The selection of sentences was detailed in [9, 11] . It was implemented as a semi-automatic process where human intervention is included to decide on the readability of the automatically selected sentences.",
+                "cite_spans": [
+                    {
+                        "start": 467,
+                        "end": 470,
+                        "text": "[9,",
+                        "ref_id": "BIBREF9"
+                    },
+                    {
+                        "start": 471,
+                        "end": 474,
+                        "text": "11]",
+                        "ref_id": "BIBREF11"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Corpora Design and Organization",
+                "sec_num": "2"
+            },
+            {
+                "text": "The short paragraphs attempt to emphasize more on the variations of the speaking behaviour and characteristics. For short paragraphs, the selection is solely based on the readability of the paragraphs without taking into consideration of the phonetic content. It aims to enrich the sentence data as well as provide data that bears very different speaking style. Table 1 shows the amount of data for each of these types. Short paragraphs While the short paragraphs can enrich the phonetic coverage as mentioned in 2-1-1, the data collected in this part is believed to be very different from that of the stand alone sentences. There are many different speaking phenomena being exaggerated when people reading a section of long text materials. These include correction, hesitation, breathing, long pause etc. Therefore, these recorded materials can also serve the purpose of representing another kind of speaking style in addition to enriching the phonetic content of the sentence corpus.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 362,
+                        "end": 369,
+                        "text": "Table 1",
+                        "ref_id": "TABREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "Spontaneous conversation In the CUCall corpora, a new type of speech data to be collected is the spontaneous conversation type of utterances. These data are collected with the aim to obtain the characteristics of various speakers when prompted to speak in an unprepared manner. There are expected delay, hesitation, correction and skipped words etc. In addition, there are also many colloquials, pronunciations and agrammatical sentences that will not be found in normal read speech. These will provide us with invaluable data for the study of the variation of speaking characteristics under different situations.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "The design of \"prompts\" for this part of data collection has been carefully planned.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "It is implemented as a single round dialogue between the speaker and the system. Since the speakers are free to answer anything to the prompts, the phonetic content is uncontrollable. The major consideration here is to ensure that there is a high proportion of speakers capable of responding to the prompts. Due to the lengthy nature of the recording process, some speakers are expected to skip these prompts intentionally while some may be too enthusiastic to give very long answers. Several points are considered during the design of prompts:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "1. The prompts must be simple enough that \"spontaneous\" response is possible. Calculation, memory recall or questions requiring accuracy are not suitable.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "2. The prompts must have different answers from different speakers so as to increase the variations of the collected data. It would be even better if the same speaker will",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "give different answers at different time.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "3. The responses to the prompts may be either long or short. 4 . Both for legal purpose and encouraging speakers to answer, the content of the answer must be irrelevant to privacy of the speakers.",
+                "cite_spans": [
+                    {
+                        "start": 61,
+                        "end": 62,
+                        "text": "4",
+                        "ref_id": "BIBREF4"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "Based on the considerations mentioned above, we have carefully designed six prompts.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "These prompts are carried out at the end of each of the collection sessions. It is done this way because by that time, the speaker will be more familiar with the recording process. This will then reduce the probability of making mistake since unprepared types of responses are usually \"error\" prone. In Figure 1 , the six prompts are listed for reference (in English translation, because of the colloquial nature of the Cantonese prompt, not all words are writtable in characters ). Kong together with the navigation commands adopted from the CUCMD [9, 11] corpus.",
+                "cite_spans": [
+                    {
+                        "start": 549,
+                        "end": 552,
+                        "text": "[9,",
+                        "ref_id": "BIBREF9"
+                    },
+                    {
+                        "start": 553,
+                        "end": 556,
+                        "text": "11]",
+                        "ref_id": "BIBREF11"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 303,
+                        "end": 311,
+                        "text": "Figure 1",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "These phrases cover the financial domain, navigation commands, as well as major local places. They could be used when building command based speech applications for the related domains. Table 2 lists the amount of corresponding type of phrases. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 186,
+                        "end": 193,
+                        "text": "Table 2",
+                        "ref_id": "TABREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Short paragraphs",
+                "sec_num": null
+            },
+            {
+                "text": "The data collection is facilitated by using an automatic call centre type telephone server system. The overall set-up is shown in Figure 3 . This server system allows the speakers to call in and then read the provided materials. It is also equipped with the usual navigating features with a touch-tone telephone system.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 130,
+                        "end": 138,
+                        "text": "Figure 3",
+                        "ref_id": "FIGREF11"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Data Collection Process",
+                "sec_num": "3"
+            },
+            {
+                "text": "The telephone server is a cluster of computers with one file server and two computer telephony servers (see Figure Figure 4 ). The file server has a large 64 GB harddisk and is directly connected to the two telephony servers over a 100 Mbps isolated ethernet. 5 The computer telephony servers are equipped with a Dialogic D/41-ESC four port telephony cards for telephone network connection.",
+                "cite_spans": [
+                    {
+                        "start": 260,
+                        "end": 261,
+                        "text": "5",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 108,
+                        "end": 123,
+                        "text": "Figure Figure 4",
+                        "ref_id": "FIGREF3"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "3-1 Telephone Server",
+                "sec_num": null
+            },
+            {
+                "text": "There are eight ports available on the Dialogic D/41-ESC card, but only two ports are used. This is sufficient for our current scale of speech collection. Also, additional ports may be used as backup during system maintenance or occasional system breakdown.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-1 Telephone Server",
+                "sec_num": null
+            },
+            {
+                "text": "Furthermore, we can also even out the potential analog channel discrepancies among the different ports by intermittently changing the answering ports over the course of data collection.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-1 Telephone Server",
+                "sec_num": null
+            },
+            {
+                "text": "The actual collection process was implemented in several steps:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-2 Collection Process",
+                "sec_num": null
+            },
+            {
+                "text": "1. Preparation of the reading materials;",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-2 Collection Process",
+                "sec_num": null
+            },
+            {
+                "text": "2. Distribution of the reading materials;",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-2 Collection Process",
+                "sec_num": null
+            },
+            {
+                "text": "3. Accepted speakers call to the telephone server. Speakers call The speakers will make call to our telephone server at any time they so wish. The server would answer the calls whenever it is idle. The speakers are then requested to jot down a generated serial number for bookkeeping purpose. After that, our server program will prompt the speaker by the item numbers on the prompt sheet and then wait for the speakers' speech data with an automatic silence detector. After the speakers have read the prompted item (or time-out if the speakers do not say anything), the data is immediately stored on to the server's hard disk. This prompting process repeats until the last item is finished. The server then reminds the speakers to fill out the questionnaire and hang up subsequently.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-2 Collection Process",
+                "sec_num": null
+            },
+            {
+                "text": "Questionnaire return After the agents have collected the prompt sheet, the serial number and questionnaire results are entered into our database for bookkeeping and analysis purposes. Up to this point the collection process is completed and the data are kept for later post-processing.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "3-2 Collection Process",
+                "sec_num": null
+            },
+            {
+                "text": "The most important part of a spoken language corpora development process is the postprocessing of the collected speech data. The collected data need to be accurately annotated with necessary labels and organized properly for easy distribution and usage.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Post-Processing of Data",
+                "sec_num": "4"
+            },
+            {
+                "text": "Based on our previous experience from developing the CUCorpora [9] , we have carefully designed the post-processing procedure for the telephone speech data. Phonemic transcription of the validated data A major effort in spoken language corpora development is annotation. This is the most important and labour intensive process. In our case, all of the validated data will be transferred using cassette tapes to our contracted professional transcribers. They will listen to the recording tapes and provide",
+                "cite_spans": [
+                    {
+                        "start": 63,
+                        "end": 66,
+                        "text": "[9]",
+                        "ref_id": "BIBREF9"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Post-Processing of Data",
+                "sec_num": "4"
+            },
+            {
+                "text": "Cantonese phonemic transcriptions to all data or mark them as noise wherever applicable. Those successfully transcribed data will then be accompanied by the corresponding phonemic transcription when distributed.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Post-Processing of Data",
+                "sec_num": "4"
+            },
+            {
+                "text": "Partitioning and distribution of the collected data The transcribed data will then be partitioned according to the different parts (e.g. digit strings, short phrases, sentences, spontaneous conversation etc.). The partitioned data will be organized into different directories according to different speakers. The phonemic transcription will also be provided in the form of LSHK 6 transcription symbols. These organized directories of speech data and transcription will be printed on to compact disk for distribution.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Post-Processing of Data",
+                "sec_num": "4"
+            },
+            {
+                "text": "In this section, some statistical information of the designed corpora reading materials will be presented. Although there are many expected discrepancies from the actual data that are collected, these statistics can still give an overview of the characteristics of the designed corpora. The discrepancies between the designed materials and the recorded data are mainly due to the reason that there are many speakers who read colloquial and 'lazy' pronunciations, mis-read of materials (e.g. insertion, deletion and substitution of words), and mis-use of the recording systems (e.g. start reading before the recording actually started, stop reading before all of the materials are read, etc.). These could only be analyzed after all data have been transcribed. Detailed statistical analysis of the actual collected data will be released after the information has been prepared. it is observed that the content of the corpora is reasonably distributed. While there are some frequently occurred syllables and also some rarely occurred syllables, the majority of the syllable occurrences lie in the middle range. This could then enable us to obtain a normal distribution for the syllables in these parts of the corpora.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Data Analysis",
+                "sec_num": "5"
+            },
+            {
+                "text": "For the application-specific corpora, information shown in Table 3 can give us an idea of what is being collected for the database. We have some randomly generated digit strings of various lengths. They should cover most of the common applications where digit strings are needed to be recognized. These may include getting identity card number, telephone number, credit card numbers etc. The 7-digit, 8-digit, 16-digit strings together with the single digits are targeted for these applications. However, since digit strings are so general that continuous digit string data can definitely be applied to other areas of applications.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 59,
+                        "end": 66,
+                        "text": "Table 3",
+                        "ref_id": "TABREF3"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Data Analysis",
+                "sec_num": "5"
+            },
+            {
+                "text": "The other application-specific data collected in this corpora are phrases of various kinds (see Section 2). The phrases from the various different domains are mixed and shuffled for each of the speakers so as to increase the variation in the collection data.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Data Analysis",
+                "sec_num": "5"
+            },
+            {
+                "text": "From Table 3 , it may be found that the acoustic coverage of the phrase part is not as good as that of sentences and paragraphs. Since these data are designed for use in the designated domains, phonetic coverage is not the major concern during corpus design.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 5,
+                        "end": 12,
+                        "text": "Table 3",
+                        "ref_id": "TABREF3"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Data Analysis",
+                "sec_num": "5"
+            },
+            {
+                "text": "Nevertheless, the base syllable coverage for these phrase is not far deviated from the complete Cantonese syllable inventory.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Data Analysis",
+                "sec_num": "5"
+            },
+            {
+                "text": "Regarding the amount of data in the corpora, a rough estimation has been made. Up We are currently post-processing these data and they will be made available for public release in the near future when the data are processed.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Data Analysis",
+                "sec_num": "5"
+            },
+            {
+                "text": "In this paper, the design and data collection process for a telephone spoken language corpora is presented. Details about the post-processing and preliminary analysis of the data are given. Based on the previous experience in microphone speech data collection, this work is extended to collect telephone speech data so as to provide sufficient materials for the building of statistical spoken language systems. The corpora are again divided into two parts: phonetically oriented data and application-specific data. In this work, we have further extend our previous design to include also short paragraph for encompassing speaking characteristics when people reading long materials. Furthermore, we have also included some free-form open questions or prompts for obtaining speaking characteristics in spontaneous speech. Spontaneous speech presents new challenges to speech recognition and the collected data is a valuable resource for investigating possible solutions.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions",
+                "sec_num": "6"
+            },
+            {
+                "text": "This project is developed with the support from the Innovation and Technology Fund (AF/96/99). We are grateful to industrial sponsors: Group Sense Limited and SmarTone Mobile Communication Limited. We would also like to thank the Hong Kong Blind Union for helping us transcribes the telephone speech data. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Acknowledgements",
+                "sec_num": "7"
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "\u00a2 \u00a1 \u00a4 \u00a3 \u00a6 \u00a5 \u00a7 \u00a9 \u00a5 \u00a5 \" ! $ # & % ( ' \u00a7 \u00a5 0 ) 1 ) 3 2 5 4 1 6 8 7 9 3 @ B A D C F E H G P I R Q S T I \" U V I \" C F W X @ \u1ef2 b a d c e I \" Q g f R I \" Q i h p r q G s W D t q 5 u U 5 @ F v",
+                        "eq_num": "q"
+                    }
+                ],
+                "section": "Acknowledgements",
+                "sec_num": "7"
+            },
+            {
+                "text": "http://dsp.ee.cuhk.edu.hk/speech/cucall.html 2 How may I help you? is the service offered by AT&T.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "http://dsp.ee.cuhk.edu.hk/speech 4 Hong Kong has a large number of mobile network operators offering different kinds of network services using different protocols. This includes the GSM900, GSM1800, TDMA, CDMA.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "This is intentionally set up to improve the security and robustness of the systems. The cluster of computer connected in their own network could eliminate the interference of possible network traffic from other irrelevant processes.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Linguistic Society of Hong Kong.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "MACROPHONE, an American English telephone speech corpus for the polyphone project",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Bernstein",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Taussig",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Godfrey",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "Proceedings of 1994 IEEE International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "1",
+                "issue": "",
+                "pages": "81--84",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "J. Bernstein, K. Taussig, and J. Godfrey, \"MACROPHONE, an American English telephone speech corpus for the polyphone project,\" Proceedings of 1994 IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 1, pp. 81-84, 1994.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Design considerations of a Putonghua database for speech recognition",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Chan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the Conference on Phonetics of the Languages in China",
+                "volume": "",
+                "issue": "",
+                "pages": "13--16",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C. Chan, \"Design considerations of a Putonghua database for speech recognition,\" Proceedings of the Conference on Phonetics of the Languages in China, pp. 13-16, Hong Kong, 1998.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "European speech databases for telephone applications",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Choukri",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "3",
+                "issue": "",
+                "pages": "1771--1774",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Choukri, \"European speech databases for telephone applications,\" Proceedings of 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 3, pp. 1771-1774, 1997.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Training material considerations for task-independent subword modeling: design and other possibilities",
+                "authors": [
+                    {
+                        "first": "Q",
+                        "middle": [],
+                        "last": "Huo",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "B",
+                        "middle": [],
+                        "last": "Ma",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "85--88",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Q. Huo, and B. Ma, \"Training material considerations for task-independent sub- word modeling: design and other possibilities,\" Proceedings of 1999 Oriental CO- COSDA Workshop, pp. 85-88, 1999.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "NTIMIT: a phonetically balanced, continuous speech, telephone bandwidth speech database",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Jankowski",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Kalyanswamy",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Basson",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Spitz",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Proceedings of 1990 IEEE International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "1",
+                "issue": "",
+                "pages": "109--112",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C. Jankowski, A. Kalyanswamy, S. Basson, and J. Spitz, \"NTIMIT: a phonetically balanced, continuous speech, telephone bandwidth speech database,\" Proceedings of 1990 IEEE International Conference on Acoustics, Speech, and Signal Process- ing, vol. 1, pp. 109-112, 1990.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "ATR Japanese speech database as a tool of speech recognition and synthesis",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Kurematsu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Takeda",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Sagisaka",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Katagiri",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Kuwabara",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Shikano",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Speech Communication",
+                "volume": "9",
+                "issue": "",
+                "pages": "357--363",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "A. Kurematsu, K. Takeda, Y. Sagisaka, S. Katagiri, H. Kuwabara, and K. Shikano, \"ATR Japanese speech database as a tool of speech recognition and synthesis,\" Speech Communication, vol. 9, pp. 357-363, Elsevier Science, 1990.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Construction of a large-scale Japanese speech database and its management system",
+                "authors": [
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Kuwabara",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Takeda",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Sagisaka",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Katagiri",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Morikawa",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Watanabe",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1989,
+                "venue": "Proceedings of 1989 IEEE International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "1",
+                "issue": "",
+                "pages": "560--563",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "H. Kuwabara, K. Takeda, Y. Sagisaka, S. Katagiri, S. Morikawa, and T. Watanabe, \"Construction of a large-scale Japanese speech database and its management sys- tem,\" Proceedings of 1989 IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 1, pp. 560-563, 1989.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Speech database development: design and analysis of the acoustic-phonetic corpus",
+                "authors": [
+                    {
+                        "first": "L",
+                        "middle": [],
+                        "last": "Lamel",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Kassel",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Seneff",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1986,
+                "venue": "Proceedings of DARPA Speech Recognition Workshop",
+                "volume": "",
+                "issue": "",
+                "pages": "100--109",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "L. Lamel, R. Kassel, and S. Seneff, \"Speech database development: design and analysis of the acoustic-phonetic corpus,\" Proceedings of DARPA Speech Recogni- tion Workshop, pp. 100-109, 1986.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Spoken language resources for Cantonese speech processing",
+                "authors": [
+                    {
+                        "first": "Tan",
+                        "middle": [],
+                        "last": "Lee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Lo",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Ching",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Helen",
+                        "middle": [],
+                        "last": "Meng",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2001,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Tan Lee, W.K. Lo, P.C. Ching, and Helen Meng, \"Spoken language resources for Cantonese speech processing,\" to appear in Speech Communication, Elsevier Science, 2001.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Cantonese databases developed at CUHK for speech processing",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Lo",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Chow",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Tan",
+                        "middle": [],
+                        "last": "Lee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Ching",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the Conference on Phonetics of the Languages in China",
+                "volume": "",
+                "issue": "",
+                "pages": "77--80",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "W.K. Lo, K.F. Chow, Tan Lee, and P.C. Ching, \"Cantonese databases developed at CUHK for speech processing,\" Proceedings of the Conference on Phonetics of the Languages in China, pp. 77-80, Hong Kong, 1998.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Development of Cantonese spoken language corpora for speech applications",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Lo",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Tan",
+                        "middle": [],
+                        "last": "Lee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Ching",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the First International Symposium on Chinese Spoken Language Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "102--107",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "W.K. Lo, Tan Lee, and P.C. Ching, \"Development of Cantonese spoken language corpora for speech applications,\" Proceedings of the First International Symposium on Chinese Spoken Language Processing, pp. 102-107, Singapore, 1998.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "Sub-syllabic acoustic modeling across Chinese dialects",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Lo",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Helen",
+                        "middle": [],
+                        "last": "Meng",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Ching",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Proceedings of the Second International Symposium on Chinese Spoken Language Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "97--100",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "W.K. Lo, Helen Meng, and P.C. Ching, \"Sub-syllabic acoustic modeling across Chinese dialects,\" Proceedings of the Second International Symposium on Chinese Spoken Language Processing, pp. 97-100, Beijing, 2000.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Japanese large-vocabulary continuous speech recognition using a newspaper corpus and broadcast news",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Ohtsuki",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Matsuoka",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Mori",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Yoshida",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Taguchi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Furui",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Shirai",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Speech Communication",
+                "volume": "28",
+                "issue": "",
+                "pages": "155--166",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "K. Ohtsuki, T. Matsuoka, T. Mori, K. Yoshida, Y. Taguchi, S. Furui, and K. Shi- rai, \"Japanese large-vocabulary continuous speech recognition using a newspaper corpus and broadcast news,\" Speech Communication, vol. 28, pp. 155-166, Elsevier Science, 1999.",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "The design of the Wall Street Journal based CSR corpus",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Paul",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Baker",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "Proceedings of the Fifth DARPA Speech and Natural Language Workshop",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "D. Paul, and J. Baker, \"The design of the Wall Street Journal based CSR corpus,\" Proceedings of the Fifth DARPA Speech and Natural Language Workshop, Morgan Kaufmann, 1992.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "The DARPA 1000-word resource management database for continuous speech recognition",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Price",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Fisher",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Bernstein",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Pallett",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1988,
+                "venue": "Proceedings of 1998 IEEE International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "1",
+                "issue": "",
+                "pages": "651--654",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "P. Price, W.M. Fisher, J. Bernstein, and D.S. Pallett, \"The DARPA 1000-word resource management database for continuous speech recognition,\" Proceedings of 1998 IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 1, pp. 651-654, 1988.",
+                "links": null
+            },
+            "BIBREF16": {
+                "ref_id": "b16",
+                "title": "Evaluation of spoken language systems: The ATIS domain",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Price",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Proceedings of the Third DARPA Speech and Natural Language Workshop",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "P. Price, \"Evaluation of spoken language systems: The ATIS domain,\" Proceedings of the Third DARPA Speech and Natural Language Workshop, Morgan Kaufmann, 1990.",
+                "links": null
+            },
+            "BIBREF17": {
+                "ref_id": "b17",
+                "title": "A spoken language system for automated call routing",
+                "authors": [
+                    {
+                        "first": "G",
+                        "middle": [],
+                        "last": "Riccardi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Gorin",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Ljolje",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Riley",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "2",
+                "issue": "",
+                "pages": "1143--1146",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "G. Riccardi, A.L. Gorin, A. Ljolje, and M. Riley, \"A spoken language system for automated call routing,\" Proceedings of 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 2, pp. 1143-1146, 1997.",
+                "links": null
+            },
+            "BIBREF18": {
+                "ref_id": "b18",
+                "title": "A phonetically oriented speech database for Mandarin Chinese",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [
+                            "Y"
+                        ],
+                        "last": "Tseng",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Proceedings of 1995 International Congress of Phonetics Sciences",
+                "volume": "3",
+                "issue": "",
+                "pages": "326--329",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C.Y. Tseng, \"A phonetically oriented speech database for Mandarin Chinese,\" Proceedings of 1995 International Congress of Phonetics Sciences, vol. 3, pp. 326- 329, 1995.",
+                "links": null
+            },
+            "BIBREF19": {
+                "ref_id": "b19",
+                "title": "USTC95-A Putonghua corpus",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Wang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Xia",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Ni",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "B",
+                        "middle": [],
+                        "last": "Liu",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Proceedings of the Fourth International Conference on Spoken Language Processing",
+                "volume": "3",
+                "issue": "",
+                "pages": "1894--1897",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R. Wang, D. Xia, J. Ni, and B. Liu, \"USTC95-A Putonghua corpus,\" Proceedings of the Fourth International Conference on Spoken Language Processing, vol. 3, pp. 1894-1897, 1996.",
+                "links": null
+            },
+            "BIBREF20": {
+                "ref_id": "b20",
+                "title": "Speech research infra-structure in Taiwan",
+                "authors": [
+                    {
+                        "first": "H",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Wang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of 1999 Oriental COCOSDA Workshop",
+                "volume": "",
+                "issue": "",
+                "pages": "53--56",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "H.C. Wang, \"Speech research infra-structure in Taiwan,\" Proceedings of 1999 Ori- ental COCOSDA Workshop, pp. 53-56, 1999.",
+                "links": null
+            },
+            "BIBREF21": {
+                "ref_id": "b21",
+                "title": "A common European approach to assessment, corpora and standards",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Winski",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Fourcin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "Advanced Speech Applications: European Research on Speech Technology",
+                "volume": "",
+                "issue": "",
+                "pages": "25--79",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "R. Winski, and A. Fourcin, \"A common European approach to assessment, corpora and standards,\" in Advanced Speech Applications: European Research on Speech Technology, K. Varghese, S. Pfleger, and J.P. Lefvre Eds., pp. 25-79, Spriner- Verlag, 1994.",
+                "links": null
+            },
+            "BIBREF22": {
+                "ref_id": "b22",
+                "title": "Chili Mandarin speech corpus",
+                "authors": [
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Wu",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Newsletter of ISCSLP98 Special Interest Group: Linguistic Database and Tools",
+                "volume": "",
+                "issue": "",
+                "pages": "1--3",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Y. Wu, \"Chili Mandarin speech corpus,\" Newsletter of ISCSLP98 Special Interest Group: Linguistic Database and Tools, pp. 1-3, 1998.",
+                "links": null
+            },
+            "BIBREF23": {
+                "ref_id": "b23",
+                "title": "Notes on speech corpora of standard Chinese in China",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Zhang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Newsletter of ISCSLP98 Special Interest Group: Linguistic Database and Tools",
+                "volume": "",
+                "issue": "",
+                "pages": "4--5",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "J. Zhang, \"Notes on speech corpora of standard Chinese in China,\" Newsletter of ISCSLP98 Special Interest Group: Linguistic Database and Tools, pp. 4-5, 1998.",
+                "links": null
+            },
+            "BIBREF24": {
+                "ref_id": "b24",
+                "title": "HKU96-A Putonghua corpus (CDROM version),\" HKU96 corpus",
+                "authors": [
+                    {
+                        "first": "Y",
+                        "middle": [
+                            "Q"
+                        ],
+                        "last": "Zu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "X"
+                        ],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Ho",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Chan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Y.Q. Zu, W.X. Li, M.C. Ho, and C. Chan, \"HKU96-A Putonghua corpus (CDROM version),\" HKU96 corpus, Department of Computer Science, University of Hong Kong, Hong Kong, 1996.",
+                "links": null
+            },
+            "BIBREF25": {
+                "ref_id": "b25",
+                "title": "JUPlTER: a telephone-based conversational interface for weather information",
+                "authors": [
+                    {
+                        "first": "V",
+                        "middle": [],
+                        "last": "Zue",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Seneff",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Glass",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Polifroni",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Pao",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Hazen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "L",
+                        "middle": [],
+                        "last": "Hetherington",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "IEEE Transactions on Speech and Audio Processing",
+                "volume": "8",
+                "issue": "1",
+                "pages": "86--96",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "V. Zue, S. Seneff, J.R. Glass, J. Polifroni, C. Pao, T.J. Hazen, and L. Hetherington, \"JUPlTER: a telephone-based conversational interface for weather information,\" IEEE Transactions on Speech and Audio Processing, vol. 8, is. 1, pp. 86-96, 2000.",
+                "links": null
+            },
+            "BIBREF26": {
+                "ref_id": "b26",
+                "title": "Speech database development at MIT: TIMIT and beyond",
+                "authors": [
+                    {
+                        "first": "V",
+                        "middle": [],
+                        "last": "Zue",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Seneff",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Glass",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Speech Communication",
+                "volume": "9",
+                "issue": "",
+                "pages": "351--356",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "V. Zue, S. Seneff, and J. Glass, \"Speech database development at MIT: TIMIT and beyond,\" Speech Communication, vol. 9, pp. 351-356, Elsevier Science, 1990.",
+                "links": null
+            },
+            "BIBREF27": {
+                "ref_id": "b27",
+                "title": "Automatic Call Center Through Speech Understanding System",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "http://www.wcl.ee.upatras.gr/access/access.htm Automatic Call Center Through Speech Understanding System.",
+                "links": null
+            },
+            "BIBREF28": {
+                "ref_id": "b28",
+                "title": "Automatic Railway Information Systems for Europe",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "http://www.compuleer.nl/arise.htm Automatic Railway Information Systems for Europe.",
+                "links": null
+            },
+            "BIBREF29": {
+                "ref_id": "b29",
+                "title": "European Language Resources Association",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "http://www.icp.grenet.fr/ELRA/home.html, European Language Resources As- sociation.",
+                "links": null
+            },
+            "BIBREF30": {
+                "ref_id": "b30",
+                "title": "Linguistic Data Consortium",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "http://www.ldc.upenn.edu, Linguistic Data Consortium.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "Phonetically oriented continuous speech data that focus on: (a) coverage through carefully designed corpora materials; and (b) different speaking styles from short paragraphs and free form spontaneous conversation style.2. Application-oriented short phrases and digit strings."
+            },
+            "FIGREF1": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "shows an overview of the organization of the CUCall telephone spoken language corpora.2-1 Phonetically-oriented data 2-1-1 Phonetic coverage orientedThe phonetically oriented data in the CUCall is based on the design of the CUCorpora with some variations. This part of the data set is made up from sentences and short paragraphs. The materials for the sentences are based on the test and training materials of the CUSENT corpus in CUCorpora and the short paragraphs are excerpted from local newspapers."
+            },
+            "FIGREF2": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "The Cantonese prompts (with English translation) for spontaneous spoken response collection."
+            },
+            "FIGREF3": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "Return of filled questionnaires from speakers.Preparation of reading materialsThe reading materials are mixtures of phrases and sentences described in Section 2. Each part is randomly shuffled and printed out on paper. Every 10 to 30 successful calls will gives a complete set for that part of the corpora.In order to differentiate against different gender and different kinds of telephone networks, the reading materials are prepared and distributed in four parallel streams: male mobile, male fixed-line, female mobile and female fixed-line. At the end of each of the prompt sheets, there is a short questionnaire to enable the collection of information about the speaker's age group, telephone network operator (for mobile phone) or type of telephone (whether they are using extension line or direct line).Distribution of prompt sheetThe prepared prompt sheets are distributed through recruited agents. They pass the reading materials to candidate speakers. After recording, the speakers then return the prompt sheet with questionnaire duly completed to the agent and then the agent pass them back to us for processing. The adoption of an agent based distribution network allows an efficient collection process while we could indirectly control the speaker community by choosing appropriate agents."
+            },
+            "FIGREF4": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "Figure 5illustrates the general flow of the post-processing procedures.Validation of the calls Among the large number of calls received, there is a small percentage of useless data. It may be due to the reason that the speakers give up reading after a short while, the recording environment is too noisy that the silence detector failed totally, or even the system broke down. Based on the serial numbers, we validate all of the calls by checking if there is reasonable amount of data being recorded. If the call is finished properly, the information of the speaker provided on the questionnaire is entered into our speaker database anonymously."
+            },
+            "FIGREF5": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "speech data in the form of sentences of length ranging from 23 to 120 characters. These could give us a number of important and unique characteristics in long utterances."
+            },
+            "FIGREF6": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "gives another way to look at the properties of the designed reading materials in the sentences and paragraphs parts of the corpora. These are the frequency-of-frequency (FOF) scattered plots for the base and tonal syllables in these parts of the corpus. The FOF plots show the distributions of the occurrences of the syllables. From these figures,"
+            },
+            "FIGREF7": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "to the time of writing, there have been over 1,000 successful calls received. These calls give a total of around 200 hours of data covering all sorts of acoustic events (speech, silence, noise, background etc). Among this volume of recording, the sentence, speaking style, short phrases and digit parts roughly contains 84, 40, 28 and 59 hours of recording."
+            },
+            "FIGREF8": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "This is an overview of the organization of the CUCall telephone spoken language corpora for Cantonese."
+            },
+            "FIGREF9": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "Calling End : From any l oc a t i on , u s i n g any t e l e p h on e , b y al l w al k s o f l i f e Calling End : From any l oc a t i on , u s i n g any t e l e p h on e , b y r ding End : t e l e p h o ne o u t l e t , t e l e p h o ny h ar d w ar e , r e c o r d i ng s ys t e m , d at a s t o r ag e s ys t e m R e c o r ding End : t e l e p h o ne o u t l e t , t e l e p h o ny h ar d w ar e , r e c o r d i ng s ys t e m , d at a s t o r ag e s ys t e m System Set-up for Speec h D a ta C ol l ec ti on ov er T el eph on e N etw ork System Set-up for System Set-up for System Set-up for Speec h D a ta C ol l ec ti on ov er T el eph on e N etw ork Speec h D a ta C ol l ec ti on ov er T el eph on e N etw ork Speec h D a ta C ol l ec ti on ov er T el eph on e N etw ork"
+            },
+            "FIGREF11": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "The data collection process for the CUCall corpora over the telephone networks."
+            },
+            "FIGREF12": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "The telephone server setup for corpora data collection. C ol l ection ov er T el ep h one N etw ork Post-Processing for Post-Processing for Post-Processing for S p eech D a ta C ol l ection ov er T el ep h one N etw ork S p eech D a ta C ol l ection ov er T el ep h one N etw ork S p eech D a ta C ol l ection ov er T el ep h one N etw ork Pre-processing: speech / n o n -speech sepa r a t i o n et c. Pre-processing: Pre-processing: speech / speech / n o n -speech sepa r a t i o n et c. n o n -speech sepa r a t i o n et c. T ra nscript ion: a ccu r a t e v er b a t i m t r a n scr i pt i o n f o r t he speech d a t a T ra nscript ion: T ra nscript ion: a ccu r a t e a ccu r a t e v er b a t i m t r a n scr i pt i o n v er b a t i m t r a n scr i pt i o n f o r t he speech d a t a f o r t he speech d a t a D a t a S t ora ge: co l l ect ed t el epho n e speech d a t a D a t a S t ora ge: D a t a S t ora ge: co l l ect ed t el epho n e co l l ect ed t el epho n e speech d a t a speech d a t a O rga niz a t ion: o r g a n i u t ion: pr i n t i n g C D R O M f o r d i st r i b u t i o nD ist rib u t ion: D ist rib u t ion: pr i n t i n g pr i n t i n g"
+            },
+            "FIGREF13": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "Data post-processing for the CUCall corpora. Scatter plots showing the frequency-of-frequency statistics for syllables in (a) the sentence and (b) the paragraph reading materials."
+            },
+            "TABREF0": {
+                "num": null,
+                "content": "<table><tr><td>material</td><td>number</td></tr><tr><td>sentences</td><td>5719</td></tr><tr><td>short paragraphs</td><td>90</td></tr><tr><td>2-1-2 Speaking style oriented</td><td/></tr></table>",
+                "text": "The number of reading materials for each type of the phonetically oriented data.",
+                "html": null,
+                "type_str": "table"
+            },
+            "TABREF1": {
+                "num": null,
+                "content": "<table><tr><td>or street?</td></tr><tr><td>5. Besides Cantonese, what other languages do you speak?</td></tr><tr><td>6. What kinds of transportation do you take the most frequently and where do</td></tr><tr><td>you go?</td></tr><tr><td>2-2 Application-specific data</td></tr></table>",
+                "text": "1. Would you please describe the environment of your recording, such as where are you, anybody nearby and anything happening?2. Which schools have you been studying at? Such as primary and secondary school. Did you study other short courses of any kinds?3. Are you using a mobile phone? (this is intentional for a short yes-no answer)4. In which district of the city do you live? And what is the name of the estateThe CUCall corpora also contain digit strings as well as application-specific short phrases in some specific domains. The design of the digit corpus is similar to that of the CUD-IGIT[9] corpus. In CUCall, the reading materials include all of the single digits together with some random generated long digit strings. This makes up a small-scale digit string corpus collected over telephone network from a large number of speakers.The short phrase materials are designed with reference to CUCorpora. Phrases are chosen from various reading materials including names of listed companies and their abbreviations, name of foreign currencies, district names and major housing estates in Hong",
+                "html": null,
+                "type_str": "table"
+            },
+            "TABREF2": {
+                "num": null,
+                "content": "<table><tr><td>material</td><td>amount</td></tr><tr><td>name of places (districts &amp; housing estates)</td><td>228</td></tr><tr><td>listed companies</td><td>1085</td></tr><tr><td>foreign currencies</td><td>37</td></tr><tr><td>navigation commands</td><td>90</td></tr><tr><td>Total</td><td>1440</td></tr></table>",
+                "text": "The amount of different types of phrases for the application-specific data.",
+                "html": null,
+                "type_str": "table"
+            },
+            "TABREF3": {
+                "num": null,
+                "content": "<table/>",
+                "text": "shows the basic information for different parts of the corpora. From this table, it can be observed that out of the 1600 common tonal syllables in Cantonese, the sentence materials have covered over 85% of the syllables. In the short paragraphs corpus, even though the tonal syllable coverage is not as high as that of the sentence recording, we are",
+                "html": null,
+                "type_str": "table"
+            },
+            "TABREF4": {
+                "num": null,
+                "content": "<table><tr><td/><td colspan=\"2\">Phonetically oriented corpora</td><td/><td/></tr><tr><td>sentences</td><td>50 (out of 5719)</td><td>2251</td><td>1030</td><td>4 to 31</td></tr><tr><td>short paragraphs</td><td>3 (out of 90)</td><td>768</td><td>418</td><td>23 to 120</td></tr><tr><td/><td colspan=\"2\">Application-specific corpora</td><td/><td/></tr><tr><td>1-digit string</td><td>10</td><td>N.A.</td><td>N.A.</td><td>N.A.</td></tr></table>",
+                "text": "Statistical information of the reading materials for the phonetically oriented and application-specific parts of the corpora.Part# per speaker # tonal syl. # base syl. syllable count",
+                "html": null,
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1011.json

@@ -0,0 +1,1585 @@
+{
+    "paper_id": "O01-1011",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:33.691332Z"
+    },
+    "title": "An Empirical Study of Zero Anaphora Resolution in Chinese Based on Centering Model",
+    "authors": [
+        {
+            "first": "Ching-Long",
+            "middle": [],
+            "last": "Yeh",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Tatung University",
+                "location": {
+                    "addrLine": "40 Chungshan N. Rd. 3 rd . Section Taipei 104 R.O.C"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Yi-Jun",
+            "middle": [],
+            "last": "Chen",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Tatung University",
+                "location": {
+                    "addrLine": "40 Chungshan N. Rd. 3 rd . Section Taipei 104 R.O.C"
+                }
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "In this paper, we describe the creation of Chinese zero anaphora resolution rules by performing experiments. The rules were constructed based on the centering model. In the experiments, we selected several texts as testing examples. We compared the referents of zero anaphors in the testing texts identified by hand with the ones resolved by using an algorithm employing a resolution rule. Three rules were used to carry out the experiment. The results show that the rule considering grammatical role criteria and domain knowledge obtained the best result: 85% of zero anaphors in the test texts were correctly resolved. We investigate problems of miss-resolution of zero anaphors in the test text and propose solution to deal with them.",
+    "pdf_parse": {
+        "paper_id": "O01-1011",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "In this paper, we describe the creation of Chinese zero anaphora resolution rules by performing experiments. The rules were constructed based on the centering model. In the experiments, we selected several texts as testing examples. We compared the referents of zero anaphors in the testing texts identified by hand with the ones resolved by using an algorithm employing a resolution rule. Three rules were used to carry out the experiment. The results show that the rule considering grammatical role criteria and domain knowledge obtained the best result: 85% of zero anaphors in the test texts were correctly resolved. We investigate problems of miss-resolution of zero anaphors in the test text and propose solution to deal with them.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "In Chinese text, anaphors are frequently eliminated, termed zero anaphor (ZA) hereafter, due to their prominence in discourse [LT81] . For example in (1), the topic of the utterance (1a) is 'Electronics stocks,' which is eliminated in the second utterance and the topic of utterance (1c), 'Securities stocks,' is eliminated in the utterance (1d).",
+                "cite_spans": [
+                    {
+                        "start": 73,
+                        "end": 77,
+                        "text": "(ZA)",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 126,
+                        "end": 132,
+                        "text": "[LT81]",
+                        "ref_id": "BIBREF14"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "(1) a. A simple rule, Rule 1, can be formulated by observing the phenomenon of topic chain in Chinese text. This rule can be used to correctly resolve the referent of the ZA in (1a), for example.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Rule 1: If a ZA occurs in the topic position of utterance i, then its antecedent is the topic of utterance i-1.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In general, zero anaphors in Chinese can occur in any grammatical slot with an antecedent that may occur in any grammatical slot, regardless of their distance [LT79] .",
+                "cite_spans": [
+                    {
+                        "start": 159,
+                        "end": 165,
+                        "text": "[LT79]",
+                        "ref_id": "BIBREF13"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Thus Rule 1 is obviously insufficient to account for the resolution of ZAs.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Within the theories of discourse, Centering is a computational model, which has been developed as a methodology for the explanation of the local coherence and its relationship to attentional state at the local level and focuses on pronominal and nominal anaphora [GJW83, GJW95] . It is formalized as a system of constraints and rules, which can, as part of a computational discourse model, act to control inference [JW81] . In the centering model, each utterance in a discourse segment has two structures associated with it, called Forward-Looking and Backward-Looking centers,",
+                "cite_spans": [
+                    {
+                        "start": 263,
+                        "end": 270,
+                        "text": "[GJW83,",
+                        "ref_id": "BIBREF2"
+                    },
+                    {
+                        "start": 271,
+                        "end": 277,
+                        "text": "GJW95]",
+                        "ref_id": "BIBREF4"
+                    },
+                    {
+                        "start": 415,
+                        "end": 421,
+                        "text": "[JW81]",
+                        "ref_id": "BIBREF11"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "which correspond approximately to Sidner's potential foci and discourse focus [Sid79] .",
+                "cite_spans": [
+                    {
+                        "start": 78,
+                        "end": 85,
+                        "text": "[Sid79]",
+                        "ref_id": "BIBREF18"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Forward-Looking Centers, C f , is a set of discourse entities in an utterance, and",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Backward-Looking Center, C b , is a special member of this set, which is the discourse entity that the utterance most centrally concerns. Our analysis is based on this computational model to resolve the intersentential ZAs.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we aim at formulating rules for the resolution of zero anaphors in Chinese. We start with a rule, Rule 2, formulated by employing the centering model.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "For each utterance U i in a discourse segment U 1 , \u2026 , U m : If C b (U i )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rule 2:",
+                "sec_num": null
+            },
+            {
+                "text": "is realized by a zero anaphor in U i+1 then the C b (U i+1 ) must be realized by",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rule 2:",
+                "sec_num": null
+            },
+            {
+                "text": "C b (U i ).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rule 2:",
+                "sec_num": null
+            },
+            {
+                "text": "We performed an experiment by using an algorithm employing this rule to see how the ZAs in news text are resolved. The initial result showed that about half of the ZAs could not be correctly resolved. Consequently we considered adding other constraints, such as grammatical role criteria and semantic knowledge, to enhance the rule and get better results. We repeated the experiment and the result showed that about 85% can be correctly resolved by using the new rule. The remaining 15% errors of the ZAs resolution occur because of the lack of sufficient semantic knowledge and the character of locality of centering model. We further investigate these situations and",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rule 2:",
+                "sec_num": null
+            },
+            {
+                "text": "propose an approach to solve the problem.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rule 2:",
+                "sec_num": null
+            },
+            {
+                "text": "In the next section we describe t his nature of zero anaphora in Chinese. In Section 3, w e describe the centering model, and we illustrate the result of the empirical study we made by observing the industry news we collected in Section 4.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rule 2:",
+                "sec_num": null
+            },
+            {
+                "text": "The discussion and implementation are in Section 5 and 6, respectively, and finally conclusions and future works are made.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rule 2:",
+                "sec_num": null
+            },
+            {
+                "text": "In Chinese, anaphors can be classified as zero, pronominal and nominal forms, as exemplified in (2) by i , and , respectively [Chen87] 1 . Zero anaphors are generally noun phrases that are understood from the context and do not need to be specified.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Zero Anaphora in Chinese",
+                "sec_num": "2."
+            },
+            {
+                "text": "(2) a.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Zero Anaphora in Chinese",
+                "sec_num": "2."
+            },
+            {
+                "text": "i Zhangsan frightened and ran outside. (3) The house, (someone) has finished building it.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Zero Anaphora in Chinese",
+                "sec_num": "2."
+            },
+            {
+                "text": "In the intersentential case, antecedent and anaphors are located in different sentences. Depending upon the distance between the sentences containing antecedent 1 We use a b a \u03c6 to denote a zero anaphor, where the subscript a is the index of the zero anaphor itself and the superscript b is the index of the referent. A single without any script represents an intrasentential zero anaphor. Also note that a superscript attached to an NP is used to represent the index of the referent. and anaphor, it can further be divided into two types: immediate and long distance.",
+                "cite_spans": [
+                    {
+                        "start": 161,
+                        "end": 162,
+                        "text": "1",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Zero Anaphora in Chinese",
+                "sec_num": "2."
+            },
+            {
+                "text": "The former is where the sentence containing the antecedent is immediately followed by the one containing the anaphor, such as j Since Chinese has no inflection, conjugation, or case markers, the pronominal system is relatively simple, as shown in Table 1 [LT81] . A third-person pronoun can be used to replace an intersentential zero anaphor, except for first-and second-person pronouns, without changing the meaning of the sentence. Though the resulting meaning of each sentence is unchanged, the whole discourse becomes less coherent. ",
+                "cite_spans": [
+                    {
+                        "start": 255,
+                        "end": 261,
+                        "text": "[LT81]",
+                        "ref_id": "BIBREF14"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 247,
+                        "end": 254,
+                        "text": "Table 1",
+                        "ref_id": "TABREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Zero Anaphora in Chinese",
+                "sec_num": "2."
+            },
+            {
+                "text": "Centering has its computational foundations established by Grosz and Sidner [Gro77, Sid79] and were further developed by Groze, Joshi and Weinstein [GJW83, GJW95] .",
+                "cite_spans": [
+                    {
+                        "start": 76,
+                        "end": 83,
+                        "text": "[Gro77,",
+                        "ref_id": "BIBREF7"
+                    },
+                    {
+                        "start": 84,
+                        "end": 90,
+                        "text": "Sid79]",
+                        "ref_id": "BIBREF18"
+                    },
+                    {
+                        "start": 148,
+                        "end": 155,
+                        "text": "[GJW83,",
+                        "ref_id": "BIBREF2"
+                    },
+                    {
+                        "start": 156,
+                        "end": 162,
+                        "text": "GJW95]",
+                        "ref_id": "BIBREF4"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Centering Model",
+                "sec_num": "3."
+            },
+            {
+                "text": "Within the framework of the centering model, each utterance U in a discourse segment has two structures associated with it, called forward-looking centers, C f (U), and backward-looking center, C b (U). The forward-looking centers of U n , C f (U n ), depend only on the expressions that constitute that utterance. They are not constrained by features of any previous utterance in the discourse segment (DS), and the elements of C f (U n ) are partially ordered to reflect relative prominence in U n . The more highly ranked an element of C f (U n ), the more likely it is to be C b (U n+1 ). The highest ranked",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Centering Model",
+                "sec_num": "3."
+            },
+            {
+                "text": "element of C f (U n ) that is realized 2 in U n+1 is the C b (U n+1 ).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Centering Model",
+                "sec_num": "3."
+            },
+            {
+                "text": "The set of forward-looking centers, C f , is ranked according to discourse salience.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Centering Model",
+                "sec_num": "3."
+            },
+            {
+                "text": "The highest ranked member of the set of forward-looking centers is referred to as the preferred center, C p . 3 The preferred center of the utterance U n represents a prediction about the C b of the following utterance U n+1 and is the most preferred antecedent of an anaphoric or elliptical expression in U n+1 . Hence, the most important single construct of the centering model is the ordering of the list of forward-looking centers [WIC94, SH96] .",
+                "cite_spans": [
+                    {
+                        "start": 110,
+                        "end": 111,
+                        "text": "3",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 435,
+                        "end": 442,
+                        "text": "[WIC94,",
+                        "ref_id": "BIBREF21"
+                    },
+                    {
+                        "start": 443,
+                        "end": 448,
+                        "text": "SH96]",
+                        "ref_id": "BIBREF17"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Centering Model",
+                "sec_num": "3."
+            },
+            {
+                "text": "In addition to the structures for centers, C b , and C f , the theory of centering specifies a set of constraints and rules [WIC94, GJW95] .",
+                "cite_spans": [
+                    {
+                        "start": 124,
+                        "end": 131,
+                        "text": "[WIC94,",
+                        "ref_id": "BIBREF21"
+                    },
+                    {
+                        "start": 132,
+                        "end": 138,
+                        "text": "GJW95]",
+                        "ref_id": "BIBREF4"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Constraints and rules",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "For each utterance U i in a discourse segment U 1 , \u2026, U m :",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Constraints",
+                "sec_num": null
+            },
+            {
+                "text": "1. U i has exactly one C b . 2. Every element of C f (U i ) must be realized in U i . 3. Ranking of elements in C f (U i ) guides determination of C b (U i+1 ).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Constraints",
+                "sec_num": null
+            },
+            {
+                "text": "C b (U i ) is from C f (U i-1 )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The choice of",
+                "sec_num": "4."
+            },
+            {
+                "text": ", and can not be from C f (U i-2 ) or other prior sets of C f .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The choice of",
+                "sec_num": "4."
+            },
+            {
+                "text": "Backward-looking centers, C b s, are often omitted or pronominalized and discourses that continue centering the same entity are more coherent than those that shift from one center to another. This means that some transitions are preferred over others.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The choice of",
+                "sec_num": "4."
+            },
+            {
+                "text": "These observations are encapsulated in two rules [WIC90, WIC94, GJW95]:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The choice of",
+                "sec_num": "4."
+            },
+            {
+                "text": "For each utterance U i in a discourse segment U 1 , \u2026 , U m : Rule II reflect the intuition that continuation of the center and the use of retentions when possible to produce smooth trans itions to a new center provide a basis for local coherence. The transition states are further described in the next section.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rules",
+                "sec_num": null
+            },
+            {
+                "text": "I. I. If any element of C f (U i ) is",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Rules",
+                "sec_num": null
+            },
+            {
+                "text": "The typology of transitions from U i-1 to U i is based on two factors: whether the C b (U i )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "is the same as C b (U i-1 ), and whether this discourse entity, C b (U i ), is the same as the ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "C p (U i ): 1. C b (U i ) = C b (U i-1 ), or C b (U i-1 ) is undefined. 2. C b (U i ) = C p (U i ) If",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": ") = C b (U i-1 ), or C b (U i-1 )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "is undefined\" and \" C b",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "(U i ) = C p (U i ), or C b (U i ) is undefined.\" In (1c), the transition state is RETAIN because of \"C b (U 1c ) C p (U 1c )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": ".\". SMOOTH-SHIFT is the last transition state of example (1) while \"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "C b (U 1d ) = C p (U 1d )\" and \" C b (U 1d ) C b (U 1c )\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "hold. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "C b : undefined C f : [ , ] (1a) C p : CONTINUE C b : C f : [ZA ( )] (1b) C p ZA ( ) CONTINUE C b : C f : [ ] (1c) C p : RETAIN C b : C f : [ZA ( ), ] (1d) C p : ZA ( ) SMOOTH-SHIFT",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Transition states",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "This paper is concerned with resolving the problem of zero anaphora in Chinese using the centering model. In this section, we first describe the methodology of zero anaphora resolution we adopted based on centering. Second, we explain how to apply our rules and represent the results of applying the different rules to the test texts.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiment and Result",
+                "sec_num": "4."
+            },
+            {
+                "text": "C b (U i ) = C b (U i-1 ) or C b (U i-1 ) is undefined C b (U i ) C b (U i-1 ) C b (U i ) = C p (U i ) CONTINUE SMOOTH-SHIFT C b (U i ) C p (U i ) RETAIN ROUGH-SHIFT",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiment for zero anaphora resolution",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "The task of zero and nominal anaphora resolution is performed after the semantic interpretation phase that converts the syntactic structure of a sentence into a semantic representation form such as the logic form [JA94] . After semantic interpretation, an anaphor becomes a parameter in a logic form. For example, the logic form of the (5b) is ( ). The task of anaphora resolution is to find out the referent of the omission in the logic forms. as the object of a prepositional phrase [LT81] . Therefore to apply the centering model for zero anaphora resolution, the essential task is to rank the elements in the set. The task of ranking elements is determined according to certain rules, for example Rule 2 described previously in Section 1. In this paper, our goal is to develop effective rules to obtain better result.",
+                "cite_spans": [
+                    {
+                        "start": 213,
+                        "end": 219,
+                        "text": "[JA94]",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 485,
+                        "end": 491,
+                        "text": "[LT81]",
+                        "ref_id": "BIBREF14"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiment for zero anaphora resolution",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "We performed an experiment to examine the effectiveness of using a rule for the resolution of zero anaphors. In the experiment, we selected a number of industry news as the test text s. Table 4 summarizes the total news, paragraphs, utterances, zero anaphors and words in the test texts. In the experiment, we first of all identify by hand the referent of each zero anaphor occurring in the texts. Then we compute the referents of zero anaphors identified by using an algorithm employing a resolution rule. The computed result is then compared with the one by hand to see the correction rate of the resolution rule.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 186,
+                        "end": 193,
+                        "text": "Table 4",
+                        "ref_id": "TABREF5"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Experiment for zero anaphora resolution",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "The correction rate of a resolution rule is defined as below.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiment for zero anaphora resolution",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "Correction rate: Assume that m ZAs occur in n utterances. The correction rate of a resolution rule is the number of referents of ZAs resolved by an algorithm employing the resolution rule that are identical to the ones identified by hand.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiment for zero anaphora resolution",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "The experiment is performed repeatedly by replacing new rules and it is stopped until promising result is obtained. The initial result of using Rule 2 shows that only 55% of the ZAs are correctly resolved, which is obviously not effective enough. The errors occurs in the initial result may be that Rule 2 does contain enough semantic knowledge. In the following, we propose other rules to replace Rule 2 and compare the results.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiment for zero anaphora resolution",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "Grosz et al., in their paper [GJW95] , assume that grammatical roles are the major determinant for ranking the forward-looking centers, with the order \"Subject > Object(s) > Others\". In Chinese, the concept of subject seems to be less significant while the topic in a sentence appears to be crucial in explaining the structure of ordinary sentences in the language [LT81] . ) is realized by both specific nouns in the lexicon and a ZA having the highest priority of grammatical role criteria in U i+1 then the C b (U i+1 ) must be realized by C b (U i ).",
+                "cite_spans": [
+                    {
+                        "start": 29,
+                        "end": 36,
+                        "text": "[GJW95]",
+                        "ref_id": "BIBREF4"
+                    },
+                    {
+                        "start": 365,
+                        "end": 371,
+                        "text": "[LT81]",
+                        "ref_id": "BIBREF14"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 374,
+                        "end": 375,
+                        "text": ")",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Results of using other rules",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "In Rule 4, in addition to grammatical role criteria, we further add the lexical semantic knowledge to the nouns specified in the lexicon. The experiment results of using these rules are investigated as follows.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Results of using other rules",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "The experiment is performed three times by using Rule2, 3 and 4, respectively. The first experiment employs the simplest rule, Rule 2, as described in Section 1. Since",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiment results using three rules",
+                "sec_num": "4.3"
+            },
+            {
+                "text": "Rule 2 does not have constraint to order elements in C f , here we take the surface order of entities from left to right in the utterance. After performing the experiment, the correction rate is 55%, which is obviously not satisfied. In the second experiment, we employed an enhanced rule, Rule 3, and the correction rate is 62%. The result is better but it is still not significant. In the third experiment, we used a further enhanced rule, Rule 4, the correction rate becomes 85%, which is more promising. The results are summarized in Table 5 . having their antecedents outside this scope, the rules would be ineffective. Worse yet, the miss-resolution of a long distance zero anaphora would fail to resolve the following zero anaphors, or error chaining [SH96] . To solve this problem, we extend the referent set of C b (U i ) to be the collection of entities occurring in utterances previously in the discourse, that is, U 1 \u2026 U i-1 . The referent of a long distance zero anaphor is then determined by examining the elements in the extended referent set.",
+                "cite_spans": [
+                    {
+                        "start": 758,
+                        "end": 764,
+                        "text": "[SH96]",
+                        "ref_id": "BIBREF17"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 538,
+                        "end": 545,
+                        "text": "Table 5",
+                        "ref_id": "TABREF7"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Experiment results using three rules",
+                "sec_num": "4.3"
+            },
+            {
+                "text": "The algorithm for resolving long distance zero anaphors is described as below.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiment results using three rules",
+                "sec_num": "4.3"
+            },
+            {
+                "text": "Description: A long distance zero anaphor z is found in the current utterance U i and then it enters the following procedure. Assume that the extended referent set is E. A temporary set, temp_set, is used to record the elements in E that satisfy the semantic constraints of z.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiment results using three rules",
+                "sec_num": "4.3"
+            },
+            {
+                "text": "For each element e in E do If e satisfies the semantic constraints of z, then add e to temp_set. end for; If there is one element in temp_set then return the element as the result; else return the element in temp_set having longest distance from z as the result.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Procedure:",
+                "sec_num": null
+            },
+            {
+                "text": "The s emantic constraints we used in the above procedure come from the selectional restrictions of the main verb in utterance U i [JA94] . This kind of restrictions can be used to select the referents of zero anaphors in the topic position.",
+                "cite_spans": [
+                    {
+                        "start": 130,
+                        "end": 136,
+                        "text": "[JA94]",
+                        "ref_id": "BIBREF10"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Procedure:",
+                "sec_num": null
+            },
+            {
+                "text": "On the one hand, in the sentence which the topic and subject are identical, the zero anaphor in the topic position is restricted by the semantics of the main verb. On the other hand, for sentences with both topic and subject, the topic is frequently moved from the object positio n of the sentence. Thus zero anaphors of this sort are restricted by the main verb as well. We ignore the selectional restrictions of other syntactic constructs such as coverb and adjective phrases because the objects or heads of these kinds of phrases can not be zeroed according to syntactic constraints in Chinese [LT81] . Consider, for example, the long distance zero anaphor i 2 \u03c6 in (6d). Before entering the above procedure, assume that the extended referent set, {",
+                "cite_spans": [
+                    {
+                        "start": 597,
+                        "end": 603,
+                        "text": "[LT81]",
+                        "ref_id": "BIBREF14"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Procedure:",
+                "sec_num": null
+            },
+            {
+                "text": "i , j ,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Procedure:",
+                "sec_num": null
+            },
+            {
+                "text": "k , l }, was obtained, where the first two elements satisfy the selectional restrictions of the main verb of (6d), . Here the first one is selected because it is in a more prominent position. c.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Procedure:",
+                "sec_num": null
+            },
+            {
+                "text": "The NTD's exchange rate stops to slowly fall down.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "l k",
+                "sec_num": null
+            },
+            {
+                "text": "d. i 2 \u03c6 j j",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "l k",
+                "sec_num": null
+            },
+            {
+                "text": "(They) expect that Central Bank of China will not let NTD be depreciated.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "l k",
+                "sec_num": null
+            },
+            {
+                "text": "The goal of this paper is to resolve zero anaphors occurring in discourses based on the centering model. A discourse is a sequence of utterances exhibiting coherence [GJW95] . The resolution of zero anaphors in a discourse is therefore divided into two parts. First, we process each utterance in turn and identify zero anaphors occurring in the utterance. Then we apply a zero anaphor resolution algorithm to resolve the referents of the zero anaphors.",
+                "cite_spans": [
+                    {
+                        "start": 166,
+                        "end": 173,
+                        "text": "[GJW95]",
+                        "ref_id": "BIBREF4"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Implementation",
+                "sec_num": "6."
+            },
+            {
+                "text": "The first part consists of tasks of word segmentation, parsing and semantic",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Implementation",
+                "sec_num": "6."
+            },
+            {
+                "text": "interpretation. An input utterance is fragmented into word sequence, and after parsing and semantic interpretation, the semantic form is obtained. Therefore, in this part, the input is a sequence of utterances and the output is the corresponding sequence of semantic forms. Zero anaphors with the information of either immediate or long distance are represented as arguments in the semantic forms. Basically, a zero anaphor is considered an immediate one. But if there are linguistic cues accompanied with the utterance, such as the utterance is the beginning of a new full sentence, and it has initial adverbial connectives, etc., then the zero anaphor is considered a long distance case. In the second part, the resolution procedure examines each zero anaphor in turn.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Implementation",
+                "sec_num": "6."
+            },
+            {
+                "text": "If an immediate zero anaphor is found, then apply the resolution rules described in Section 4. Otherwise, if it is a long distance zero anaphor, then apply the procedure as described in Section 5. The system architecture is shown in Figure 2 . is build as a sentence-level parser in DCG [GM89] . Each utterance within an input discourse segme nt is converted into a syntactical structure by the parser and the output structure is interpreted to produce the semantic form, which includes the entities in the utterance and is also used to judge whether the utterance contains zero anaphors or not.",
+                "cite_spans": [
+                    {
+                        "start": 287,
+                        "end": 293,
+                        "text": "[GM89]",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 233,
+                        "end": 241,
+                        "text": "Figure 2",
+                        "ref_id": "FIGREF7"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Implementation",
+                "sec_num": "6."
+            },
+            {
+                "text": "ZA resolution by consulting the domain ontology and resolution rules is the second part of our system. If an input utterance contains a zero anaphor, then apply the resolution rules described in Section 4 to obtain the referent of the zero anaphor.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Implementation",
+                "sec_num": "6."
+            },
+            {
+                "text": "Currently, the ZA Resolution only deals with the immediate zero anaphors. We will extend the algorithm to include the resolution of long distance zero anaphors described in Section 5.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Implementation",
+                "sec_num": "6."
+            },
+            {
+                "text": "In this paper, we performed the experiments on zero anaphora resolutio n in Chinese based on centering model. In the experiments, 85% of zero anaphors in the test texts were correctly resolved. The remaining zero anaphors were miss-resolved because of lack of sufficient domain knowledge and occurrence of long distance zero anaphors.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions",
+                "sec_num": "7."
+            },
+            {
+                "text": "Since the centering model only focuses on local coherence in discourse, we therefore propose to extend the referent set of a zero anaphor to include all entities occurring previously in the discourse. Though the experiment results are promising to some extent, we found that there are problems that are worth further study. First we need to build domain ontology to get better resolution. Second, the phenomenon of error chaining is inherent in zero anaphors resolution. Thus an effective method is needed to account for this problem. The method we proposed in Section 5 is a step towards solving this problem. Third, the test texts used in this paper were selected from industry news. We will further extend our experiment to include texts from other domains.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions",
+                "sec_num": "7."
+            },
+            {
+                "text": "An utterance U, realizes c if c is an element of the situation described by U, or c is the semantics interpretation of come subpart of U.3 The notion of preferred center corresponds to Sider's notion of expected focus[Sid83]",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Hanyu lingxin huizhi de huayu fenxi (a discourse approach to zero anaphora in chinese) (in chinese)",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1987,
+                "venue": "Zhongguo Yuwen (Chinese Linguistics)",
+                "volume": "",
+                "issue": "",
+                "pages": "363--378",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "P. Chen. 1987. Hanyu lingxin huizhi de huayu fenxi (a discourse approach to zero anaphora in chinese) (in chinese). Zhongguo Yuwen (Chinese Linguistics), pages 363-378.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Academic Sinica",
+                "authors": [],
+                "year": 1999,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "CKIP. 1999. (Auto tag), Academic Sinica.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Providing a unified account of definite noun phrases in discourse",
+                "authors": [
+                    {
+                        "first": "B",
+                        "middle": [
+                            "J A K"
+                        ],
+                        "last": "Grosz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Joshi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Weinstein",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1983,
+                "venue": "Proc. of 21 st Annual Meeting of the ACL",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "B. J. Grosz. A. K. Joshi and S. Weinstein, 1983. Providing a unified account of definite noun phrases in discourse. Proc. of 21 st Annual Meeting of the ACL",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Attention, intentions, and the structure of discourse",
+                "authors": [
+                    {
+                        "first": "B",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Grosz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Sidner",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1986,
+                "venue": "Computational Linguistics",
+                "volume": "12",
+                "issue": "3",
+                "pages": "175--204",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "B. J. Grosz and C. L. Sidner. 1986. Attention, intentions, and the structure of discourse. Computational Linguistics, No 3 Vol 12, pp. 175-204.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Centering: A Framework for Modeling t he Local Coherence of Discourse",
+                "authors": [
+                    {
+                        "first": "B",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Grosz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Joshi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Weinstein",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Computational Linguistics",
+                "volume": "21",
+                "issue": "2",
+                "pages": "203--225",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "B. J. Grosz, A. K. Joshi and S. Weinstein.1995. Centering: A Framework for Modeling t he Local Coherence of Discourse. Computational Linguistics 21(2), pp. 203-225.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Natural Language Processing in PROLOG -An Introduction to Computational Linguistics",
+                "authors": [
+                    {
+                        "first": "G",
+                        "middle": [],
+                        "last": "Gazdar",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Mellish",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1989,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "G. Gazdar and C. Mellish. 1989. Natural Language Processing in PROLOG -An Introduction to Computational Linguistics, Addison- Wesley.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "The representation and use of focus in dialogue understanding",
+                "authors": [
+                    {
+                        "first": "B",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Grosz",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1977,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "B. J. Grosz. 1977. The representation and use of focus in dialogue understanding Technical Report 151, SRI International.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Segmentation Standard for Chinese Natural Language Processing",
+                "authors": [
+                    {
+                        "first": "Chu-Ren",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Keh-Jia Nn",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Li-Li",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Proceedings of the 1996 International Conference on Computational Linguistics (COLING 96')",
+                "volume": "",
+                "issue": "",
+                "pages": "1045--1048",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chu-Ren Huang, Keh-Jia nn Chen and Li-li Chang. 1996. Segmentation Standard for Chinese Natural Language Processing. Proceedings of the 1996 International Conference on Computational Linguistics (COLING 96'), pp.1045-1048. Copenhagen, Denmark.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Natural Language Understanding",
+                "authors": [
+                    {
+                        "first": "James",
+                        "middle": [],
+                        "last": "Allen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "",
+                "volume": "2",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "James Allen. 1994. Natural Language Understanding 2 nd ed., The Benjamin/Cummings Publishing Company, Inc.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Control of inference: Role of some aspects of discourse structure -centering",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Aravind",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Scott",
+                        "middle": [],
+                        "last": "Joshi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Weinstein",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1981,
+                "venue": "Proc. International Joint Conference on Artificial Intelligence",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Aravind K. Joshi and Scott Weinstein. 1981. Control of inference: Role of some aspects of discourse structure -centering. In Proc. International Joint Conference on Artificial Intelligence.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "From Sentence Processing to Information Access on the World Wide Web",
+                "authors": [
+                    {
+                        "first": "Boris",
+                        "middle": [],
+                        "last": "Katz",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "AAAI Spring Symposium",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Boris Katz. 1997. From Sentence Processing to Information Access on the World Wide Web. 1997 AAAI Spring Symposium.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Third-person pronouns and zero-anaphora in Chinese discourse",
+                "authors": [
+                    {
+                        "first": "Charles",
+                        "middle": [
+                            "N"
+                        ],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Sandra",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Thompson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1979,
+                "venue": "Syntax and Semantics: Discourse and Syntax",
+                "volume": "12",
+                "issue": "",
+                "pages": "311--335",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Charles N. Li and Sandra A. Thompson. 1979. Third-person pronouns and zero-anaphora in Chinese discourse. In T. Givon, editor, Syntax and Semantics: Discourse and Syntax, volume 12, pages 311-335. Academic Press.",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "Chinese Chinese -A Functional Reference Grammar",
+                "authors": [
+                    {
+                        "first": "Charles",
+                        "middle": [
+                            "N"
+                        ],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Sandra",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Thompson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1981,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Charles N. Li and Sandra A. Thompson. 1981. Chinese Chinese -A Functional Reference Grammar, University of California Press.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "Introduction to WordNet: An On-line Lexical Database",
+                "authors": [
+                    {
+                        "first": "George",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Miller",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Richard",
+                        "middle": [],
+                        "last": "Beckwith",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Christiane",
+                        "middle": [],
+                        "last": "Fellbaum",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Derek",
+                        "middle": [],
+                        "last": "Gross",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Katherine",
+                        "middle": [],
+                        "last": "Miller",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "International Journal of Lexicography",
+                "volume": "3",
+                "issue": "4",
+                "pages": "235--244",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "George A. Miller, Richard Beckwith, Christiane Fellbaum, Derek Gross, and Katherine Miller. 1990. Introduction to WordNet: An On-line Lexical Database. International Journal of Lexicography, vol. 3(4), pp. 235--244.",
+                "links": null
+            },
+            "BIBREF16": {
+                "ref_id": "b16",
+                "title": "Sinica Corpus )",
+                "authors": [],
+                "year": 2001,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "(Sinica Corpus ). 2001. Academic Sinica. http://www.sinica.edu.tw/",
+                "links": null
+            },
+            "BIBREF17": {
+                "ref_id": "b17",
+                "title": "Functional Centering. Proc. Of ACL '96",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Strube",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "U",
+                        "middle": [],
+                        "last": "Hahn",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "270--277",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Strube, M. and U. Hahn. 1996. Functional Centering. Proc. Of ACL '96, Santa Cruz, Ca., pp.270-277.",
+                "links": null
+            },
+            "BIBREF18": {
+                "ref_id": "b18",
+                "title": "Toward a Computational Theory of Definite Anaphora Comprehension in English Discourse",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Sider",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1979,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C. L. Sider. 1979. Toward a Computational Theory of Definite Anaphora Comprehension in English Discourse. Ph.D. thesis, MIT.",
+                "links": null
+            },
+            "BIBREF19": {
+                "ref_id": "b19",
+                "title": "Focusing in the comprehension of definite anaphora. Computational Models of Discourse",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Sider",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1983,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "C. L. Sider. 1983. Focusing in the comprehension of definite anaphora. Computational Models of Discourse, MIT Press.",
+                "links": null
+            },
+            "BIBREF20": {
+                "ref_id": "b20",
+                "title": "Centering in Japanese discourse",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Walker",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Iida",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Cote",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Proc. Of COLING-90, Appendix",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Walker, M. A., M. Iida and S. Cote. 1990. Centering in Japanese discourse. Proc. Of COLING-90, Appendix, 6pp.",
+                "links": null
+            },
+            "BIBREF21": {
+                "ref_id": "b21",
+                "title": "Japan Discourse and the Process of Centering",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Walker",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Iida",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Cote",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "Computational Linguistics",
+                "volume": "20",
+                "issue": "2",
+                "pages": "193--233",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Walker, M. A., M. Iida and S. Cote, 1994. Japan Discourse and the Process of Centering. Computational Linguistics, 20(2): 193-233.",
+                "links": null
+            },
+            "BIBREF22": {
+                "ref_id": "b22",
+                "title": "Generation of Anaphors in Chinese",
+                "authors": [
+                    {
+                        "first": "Ching-Long",
+                        "middle": [],
+                        "last": "Yeh",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ching-Long Yeh. 1995. Generation of Anaphors in Chinese, Ph.D. dissertation, University of Edinburgh",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "num": null,
+                "uris": null,
+                "text": "Electronics stocks were affected by high-tech stocks fallen heavily in America.b. i(Electronics stocks) continued falling down today.c.Securities stocks also had respondence.d. j (Securities stocks) fell by close one after another on the market.",
+                "type_str": "figure"
+            },
+            "FIGREF1": {
+                "num": null,
+                "uris": null,
+                "text": "in (4b) and k 1 \u03c6 in (4d). For the long distance type, the sentence containing the antecedent and anaphors, on the other hand, are not in immediately succeeding order, such as i must use the tips of feet on one side to grasp the ground. then uses the feet on the other side to move upwards. pushes (its) body towards one side.",
+                "type_str": "figure"
+            },
+            "FIGREF2": {
+                "num": null,
+                "uris": null,
+                "text": "realized by a pronoun in U i+1 then the C b (U i+1 ) must be realized by a pronoun also. II. Sequences of continuation are preferred over sequence of retaining; and sequences of retaining are to be preferred over sequences of shifting. Rule I represents one function of pronominal reference: the use of a pronoun to realize the C b signals the hearer that the speaker is continuing to talk about the same thing. Psychological research and cross-linguistic research have validated that the C b is preferentially realized by a pronoun in English and by equivalent forms (i.e. zero anaphora) in other languages [GJW95].",
+                "type_str": "figure"
+            },
+            "FIGREF3": {
+                "num": null,
+                "uris": null,
+                "text": "Zhangsan bought an apple.b. i (It) is very fresh. Recall that the centering model, an utterance, U i , is associated with a set of forward-looking centers, C f , with each element an entity in U i . The highest ranked element in the set, C p , becomes the prediction of backward-looking centers, C b , of the following utterance, which is zeroed if it does not violate syntactic constraints, such",
+                "type_str": "figure"
+            },
+            "FIGREF4": {
+                "num": null,
+                "uris": null,
+                "text": "By adopting the concept of grammatical roles and topic-prominence in Chinese, we order the grammatical roles in Chinese with topic having the highest priority as shown in Figure 1. The subject and objects occurring in an embedded clause, that is, Secondary Subject and Secondary Objects, are give lower priority.",
+                "type_str": "figure"
+            },
+            "FIGREF5": {
+                "num": null,
+                "uris": null,
+                "text": "Grammatical role criteriaBy adding the grammatical role criteria to Rule 2, we obtain a new rule, Rule 3:",
+                "type_str": "figure"
+            },
+            "FIGREF6": {
+                "num": null,
+                "uris": null,
+                "text": "People on the market worry that Central Bank will intervene the exchange rate again. b. i 1 \u03c6 (They) are afraid to enter the exchange market.",
+                "type_str": "figure"
+            },
+            "FIGREF7": {
+                "num": null,
+                "uris": null,
+                "text": "System architecture In the system the N LP M odule carries out in order the work of word segmentation, parsing and semantic interpretation by consulting the lexicon and the syntactic and semantic rules. This module corresponds to the first part described previously in this section. According to the segmentation standard proposed by Academia Sinica [HCC96, HC+97], we built a small lexicon for the test texts, and employ a simple algorithm of word segmentation. The algorithm is according to a strategy that prioritizes the longest word first. The syntactic grammar rules we construct are from the utterances of the test texts and refer to Sinica Corpus and Auto tag program [SC01, CKIP99] and then the parser corresponding to the grammar rules",
+                "type_str": "figure"
+            },
+            "TABREF1": {
+                "html": null,
+                "content": "<table><tr><td>Number</td><td>Person</td><td colspan=\"2\">Pronoun</td></tr><tr><td>singular</td><td>first</td><td/><td/></tr><tr><td>singular</td><td>second</td><td>,</td><td/></tr><tr><td>singular</td><td>third</td><td>, ,</td><td/></tr><tr><td>plural</td><td>first</td><td/><td/></tr><tr><td>plural</td><td>second</td><td>,</td><td/></tr><tr><td>plural</td><td>third</td><td>,</td><td>,</td></tr></table>",
+                "text": "Pronominal system in Chinese",
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF2": {
+                "html": null,
+                "content": "<table/>",
+                "text": "both (1) and (2) hold then a pair continuations across U n and across U n+1 . If (1) holds but (2) does not then the utterances are in a retaining transition, which corresponds to a situation where the speaker is intending to shift onto a new entity in the next utterance. If (1) does not hold then the utterances are in one of the shifting transition states depending on whether or not (2) holds. The definition of transition",
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF3": {
+                "html": null,
+                "content": "<table/>",
+                "text": "Transition statesFor illustration purpose, consider the example (1) in Section 1; in theTable 3, the centering structures contain C b , C f and C p where the set of C f are partially ordered to reflect relative prominence in each utterance. The first two transition states of (1a) and",
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF4": {
+                "html": null,
+                "content": "<table/>",
+                "text": "",
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF5": {
+                "html": null,
+                "content": "<table><tr><td/><td>Paragraphs</td><td>Utterances</td><td>Words</td><td>Zero Anaphors</td></tr><tr><td>1</td><td>4</td><td>36</td><td>199</td><td>25</td></tr><tr><td>2</td><td>3</td><td>26</td><td>229</td><td>9</td></tr><tr><td>3</td><td>4</td><td>31</td><td>213</td><td>13</td></tr><tr><td>4</td><td>4</td><td>29</td><td>213</td><td>15</td></tr><tr><td>5</td><td>3</td><td>27</td><td>208</td><td>11</td></tr><tr><td>6</td><td>4</td><td>35</td><td>282</td><td>15</td></tr><tr><td>7</td><td>3</td><td>28</td><td>234</td><td>14</td></tr><tr><td>8</td><td>3</td><td>27</td><td>289</td><td>12</td></tr><tr><td>Total</td><td>28</td><td>239</td><td>1867</td><td>115</td></tr></table>",
+                "text": "Summary of test texts",
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF6": {
+                "html": null,
+                "content": "<table><tr><td>Rule 4:</td></tr></table>",
+                "text": "For each utterance U i in a discourse segment U 1 , \u2026 , U m : If C b (U i ) is realized by a ZA in U i+1 and no other noun phrase having higher priority of grammatical role criteria than the ZA then the C Secondary Objects Rule 3 is used to verify if the order of the elements in grammatical role criteria we assumed is helpful to raise the correction rate of zero anaphora resolution. We further developed another rule, Rule 4 , by considering the domain knowledge corresponding to the test texts. For each utterance U i in a discourse segment U 1 , \u2026 , U",
+                "num": null,
+                "type_str": "table"
+            },
+            "TABREF7": {
+                "html": null,
+                "content": "<table><tr><td/><td>Rule 2</td><td>Rule 3</td><td>Rule 4</td></tr><tr><td>ZAs correctly resolved</td><td>63</td><td>71</td><td>98</td></tr><tr><td>Correction Rate</td><td>55%</td><td>62%</td><td>85%</td></tr><tr><td>5. Discussions</td><td/><td/><td/></tr></table>",
+                "text": "Summary of experiment results using three rulesWe have performed experiments on ZA resolution by using three rules with different complexities. The result is promising to some extent; however, there are still 15% of ZAs in the test texts can not be correctly resolved. In the following, we investigate the problems and propose methods to deal with them. One problem is because of insufficient semantic knowledge, namely domain ontology. In the lexical database, one word may have several word senses and there is a set of synonyms for each sense[MBF+90]. Besides, one word may have hypernyms, hyponyms, coordinate sisters, and other relationship to another word, e.g.,",
+                "num": null,
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1012.json

@@ -0,0 +1,387 @@
+{
+    "paper_id": "O01-1012",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:37.077851Z"
+    },
+    "title": "Automatic Classification of Chinese Unknown Verbs",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "We present a new method for automatic classification of Chinese unknown verbs. The method employs the instance-based categorization using the k-nearest neighbor method for the classification. The accuracy of the classifier is about 70.92%.",
+    "pdf_parse": {
+        "paper_id": "O01-1012",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "We present a new method for automatic classification of Chinese unknown verbs. The method employs the instance-based categorization using the k-nearest neighbor method for the classification. The accuracy of the classifier is about 70.92%.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "y,1 x,1 y,1 x,1 y,1 x,1 \u2229 \u2212 = \u2229 = \u2229 \u800c\u7b2c\u3193\u90e8\u5206\u7684\u76f8\u4f3c\u5ea6\u7684\u5b9a\u7fa9\u70ba\uff1a ( )( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) \uf8f7 \uf8f7 \uf8f8 \uf8f6 \uf8ec \uf8ec \uf8ed \uf8eb \u2212 \uf8f7 \uf8f7 \uf8f8 \uf8f6 \uf8ec \uf8ec \uf8ed \uf8eb \u2229 \u2212 = \uf8f7 \uf8f7 \uf8f8 \uf8f6 \uf8ec \uf8ec \uf8ed \uf8eb \uf8f7 \uf8f7 \uf8f8 \uf8f6 \uf8ec \uf8ec \uf8ed \uf8eb \u2212 \u2229 = \u2211 \u2211 = = = = 1 n / System /Entropy sem Sem Entropy System Entropy 1 n / System /Entropy Sem Sem nContent Informatio ...Sem Sem , ...Sem Sem core SecondaryS n 2 i j y, i x, {1...m} j j y, i x, n 2 i {1...m} j m y, y,2 n x, x,2 Max Max \u6211\u5011\u4ee4(n>=m)\uff0c\u4e5f\u5c31\u662f\u7b2c\u3192\u500b\u8a5e\u689d\u7684\u5b9a\u7fa9\u7684\u7fa9\u539f\u591a\u65bc\u6216\u7b49\u65bc\u7b2c\u3193\u500b\u8a5e\u689d\u7684\u7fa9 \u539f\uff0c\u5f9e\u7b2c\u3192\u500b\u8a5e\u689d\u3197\u7b2c\u3193\u500b\u7fa9\u539f\u958b\u59cb\uff0c\u6bcf\u500b\u7fa9\u539f\u8207\u7b2c\u3193\u500b\u8a5e\u689d\u3197\u7684\u6bcf\u500b\u7fa9\u539f\u8a08\u7b97 \u76f8\u4f3c\u5ea6\uff0c\u7b2c\u3192\u500b\u8a5e\u689d\u3197\u6bcf\u500b\u7fa9\u539f\u7559\u3198\u8207\u7b2c\u3193\u500b\u8a5e\u689d\u7fa9\u539f\u76f8\u4f3c\u5206\u6578\u6700\u9ad8\u7684\u7d44\u5408\uff0c\u5c07 \u7b2c\u3192\u500b\u8a5e\u689d\u3197\u6bcf\u500b\u7fa9\u539f\u5f97\u5230\u7684\u5206\u6578\u5e73\u5747\uff0c\u5c31\u662f\u6211\u5011\u6240\u5b9a\u7fa9\u7684\u7b2c\u3193\u90e8\u5206\u7684\u76f8\u4f3c\u5ea6\u3002 \u4ee5\u3196\u5169\u5f0f\u3197\u5404\u9805\u7686\u9664\u4ee5 Entropy \u662f\u70ba\u7dad\u6301\u76f8\u4f3c\u503c\u4ecb\u65bc 0,1 \u4e4b\u9593\u3002 (System) 2-2-2 \u8a5e\u985e\u76f8\u4f3c\u5ea6\u6e2c\u91cf \u6211\u5011\u5c07 1.0 \u7248\u3197\u7684\u53e5\u7d50\u69cb\u6a39\u3197\u6b78\u7d0d\u51fa\u898f\u5247\uff0c\u4e26\u7d71\u8a08\u6bcf\u689d\u898f\u5247\u51fa\u73fe\u7684\u983b\u7387\uff0c\u5982 \u5716 1",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "3-1 \u8a9e\u610f\u6bd4\u91cd\u76f8\u4f3c\u5ea6\u8a55\u91cf \u6211\u5011\u9996\u5148\u8981\u56fa\u5b9a\u5169\u500b\u8b8a\u6578\uff0c\u8a9e\u610f\u8207\u8a5e\u985e\u7684\u6bd4\u91cd\u8207 K \u503c\uff0c\u624d\u80fd\u89c0\u5bdf\u51fa\u76f8\u4f3c\u5ea6\u6bd4 \u91cd\u7684\u8b8a\u5316\u5c0d\u6b63\u78ba\u7387\u7684\u5f71\u97ff\u3002\u56e0\u6b64\u5148\u7d66\u4e88 K=1\uff0c\u8a9e\u610f\u8207\u8a5e\u985e\u6bd4\u91cd\u70ba 1 \u8207 0\uff0c\u4f9d\u7167\u76f8 \u4f3c\u5ea6\u6bd4\u91cd\u7684\u8b8a\u5316\u5c0d\u6b63\u78ba\u7387\u7684\u5f71\u97ff\u88fd\u6210\u3198\u8868\u3002 \u8868\u683c 2 \u76f8\u4f3c\u5ea6\u6bd4\u91cd\u8207\u6b63\u78ba\u7387\u8b8a\u5316\u8868 \u8a9e\u610f\u8207\u8a5e\u985e\u6bd4\u91cd(\u8a9e\u610f,\u8a5e\u985e) \u8a9e\u610f\u76f8\u4f3c\u5ea6\u6bd4\u91cd(w 1 ,w 2 ) \u6b63\u78ba\u7387 (1,0) (1,0) 54.04%",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(1,0) (0.9,0.1) 57.58% (1,0) (0.8,0.2) 57.70%",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(1,0) (0.7,0.3) 57.58%",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(1,0) (0.6,0.4) 56.97%",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(1,0) (0.5,0.5) 56.85%",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(1,0) (0.4,0.6) 56.23%",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(1,0) (0.3,0.7) 55.87%",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "( ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "1,0) (0.2,0.8) 56.11% (1,0) (0.1,0.9) 56.11% (1,0) (0,1) 56.09% \u7531\u3196\u8868\u3197\u53ef\u770b\u51fa\u4e3b\u8981\u7fa9\u539f\u7684\u6bd4\u91cd\u70ba 0.8 \u8207\u6b21\u8981\u7fa9\u539f\u7684\u6bd4\u91cd\u70ba 0.2 \u6642\u53ef\u4ee5\u5f97\u5230\u6700\u9ad8 \u7684\u6b63\u78ba\u7387\uff0c\u56e0\u6b64\u5728\u672c\u5be6\u9a57\u3197\u6211\u5011\u4f7f\u7528 0.8 \u8207 0.2 \u4f5c\u70ba\u4e3b\u8981\u7fa9\u539f\u8207\u6b21\u8981\u7fa9\u539f\u7684\u6bd4\u91cd\u3002 3-2 \u8a9e\u610f\u8207\u8a5e\u985e\u6bd4\u91cd\u8a55\u91cf \u6211\u5011\u5c07\u76f8\u4f3c\u5ea6\u6bd4\u91cd\u8a2d\u5b9a w 1 \u70ba 0.8 \u8207 w 2 \u70ba 0.2\uff0cK=1\uff0c\u89c0\u5bdf\u8a9e\u610f\u8207\u8a5e\u985e\u6bd4\u91cd\u7684 \u8b8a\u5316\u5c0d\u6b63\u78ba\u7387\u7684\u5f71\u97ff\u3002 \u8868\u683c 3 \u8a9e\u610f\u8207\u8a5e\u985e\u6bd4\u91cd\u8207\u6b63\u78ba\u7387\u8b8a\u5316\u8868 \u8a9e\u610f\u8207\u8a5e\u985e\u6bd4\u91cd(\u8a9e\u610f,\u8a5e\u985e) \u8a9e\u610f\u76f8\u4f3c\u5ea6\u6bd4\u91cd(w 1 ,w 2 ) \u6b63\u78ba\u7387 (1,0) (0.8,0.2) 57.70% (0.9,0.1) (0.8,0.2)",
+                        "eq_num": "58."
+                    }
+                ],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Category Guessing for Chinese Unknown Words",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ming-Hung",
+                        "middle": [],
+                        "last": "Chao-Jan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Keh-Jiann",
+                        "middle": [],
+                        "last": "Bai",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the Natural Language Processing Pacific Rim Symposium",
+                "volume": "",
+                "issue": "",
+                "pages": "35--40",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, Chao-Jan, Ming-Hung Bai and Keh-Jiann Chen. \"Category Guessing for Chinese Unknown Words,\" Proceedings of the Natural Language Processing Pacific Rim Symposium, 1997, pp. 35-40.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Unknown Word Detection for Chinese by a Corpus-based Learning Method",
+                "authors": [
+                    {
+                        "first": "Keh-Jiann",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ming-Hong",
+                        "middle": [],
+                        "last": "Bai",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Computational Linguistics and Chinese Language Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "27--44",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, Keh-Jiann and Ming-Hong Bai. \"Unknown Word Detection for Chinese by a Corpus-based Learning Method,\" Computational Linguistics and Chinese Language Processing vol3 no. 1, 1998, pp. 27-44.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Knowledge Extraction for Identification of Chinese Organization Names",
+                "authors": [],
+                "year": 2000,
+                "venue": "Proceedings of the second Chinese Language Processing Workshop",
+                "volume": "",
+                "issue": "",
+                "pages": "15--21",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "---. \"Knowledge Extraction for Identification of Chinese Organization Names,\" Proceedings of the second Chinese Language Processing Workshop, 2000, pp. 15-21.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Mandarin Chinese: A Functional Reference Grammar",
+                "authors": [
+                    {
+                        "first": "Charles",
+                        "middle": [],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Sandra",
+                        "middle": [],
+                        "last": "Thompson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1981,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Li, Charles and Sandra Thompson. \"Mandarin Chinese: A Functional Reference Grammar\". Berkeley: University of California Press, 1981.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Using Information Content to Evaluate Semantic Similarity in a Taxonomy",
+                "authors": [
+                    {
+                        "first": "Philip",
+                        "middle": [],
+                        "last": "Resnik",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Proceedings of the 14th International Joint Conference on Artificial Intelligence",
+                "volume": "",
+                "issue": "",
+                "pages": "448--453",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Resnik, Philip. \"Using Information Content to Evaluate Semantic Similarity in a Taxonomy,\" Proceedings of the 14th International Joint Conference on Artificial Intelligence, 1995, pp. 448-453.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Semantic Similarity in a Taxonomy: An Information-Based Measure and its Application to Problems of Ambiguity in Natural Language",
+                "authors": [],
+                "year": 1998,
+                "venue": "Journal of Artificial Intelligence Research XI",
+                "volume": "",
+                "issue": "",
+                "pages": "95--130",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "---. \"Semantic Similarity in a Taxonomy: An Information-Based Measure and its Application to Problems of Ambiguity in Natural Language,\" Journal of Artificial Intelligence Research XI, 1998, pp. 95-130.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Measuring Verbal Similarity",
+                "authors": [
+                    {
+                        "first": "Philip",
+                        "middle": [],
+                        "last": "Resnik",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Mona",
+                        "middle": [],
+                        "last": "Diab",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Resnik, Philip and Mona Diab. Measuring Verbal Similarity. Technical Report: LAMP-TR-047//UMIACS-TR-2000-40/CS-TR-4149/MDA-9049-6C-1250. University of Maryland, College Park, 2000.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "A Corpus-Based Analysis of Mandarin Nominal Root Compound",
+                "authors": [
+                    {
+                        "first": "Sproat",
+                        "middle": [],
+                        "last": "Richard",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Shilin",
+                        "middle": [],
+                        "last": "Shih",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Journal of East Asian Linguistics",
+                "volume": "5",
+                "issue": "",
+                "pages": "49--71",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Sproat Richard and Shilin Shih. \"A Corpus-Based Analysis of Mandarin Nominal Root Compound,\" Journal of East Asian Linguistics 5, 1996, pp. 49-71.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Coping with Ambiguity and Unknown Words through Probalistic Model",
+                "authors": [
+                    {
+                        "first": "Ralph",
+                        "middle": [],
+                        "last": "Weischedel",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Marie",
+                        "middle": [],
+                        "last": "Meteer",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Richard",
+                        "middle": [],
+                        "last": "Schwartz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Lance",
+                        "middle": [],
+                        "last": "Ramshaw",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Jeff",
+                        "middle": [],
+                        "last": "Palmucci",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Computational Linguistics",
+                "volume": "19",
+                "issue": "",
+                "pages": "359--382",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Weischedel, Ralph, Marie Meteer, Richard Schwartz, Lance Ramshaw and Jeff Palmucci. \"Coping with Ambiguity and Unknown Words through Probalistic Model,\" Computational Linguistics 19, 1993, pp. 359-382.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF0": {
+                "type_str": "table",
+                "content": "<table><tr><td>\u4eca\u7121\u6cd5\u63d0\u9ad8\u6b63\u78ba\u7387\u7684\u4e3b\u56e0\u70ba\u52d5\u8a5e\u7e41\u8907\u7684\u5167\u90e8\u7d50\u69cb\u3002 \u6211\u5011\u7684\u76ee\u6a19\u70ba\u5c07\u52d5\u8a5e\u81ea\u52d5\u5206\u985e\u5230\u3197\u7814\u9662\u8a5e\u5eab\u5c0f\u7d44(1993)\u7684\u8a5e\u985e\u67b6\u69cb\u3196\uff0c\u52d5\u8a5e \u5011\u7684\u672a\u77e5\u52d5\u8a5e\u3002\u672a\u77e5\u52d5\u8a5e\u7684\u7b2c\u3193\u500b\u7d44\u6210\u6210\u5206\u8207\u8a13\u7df4\u8a9e\u6599\u3197\u7684\u4f8b\u5b50\u76f8\u540c\u90fd\u70ba\u300c\u5b8c\u300d \uff0c \u4e8b\u300d \u3001 \u300c\u5b78\u300d\u8207\u300c\u6559\u80b2\u300d\u3194\u500b\u7fa9\u539f\u5b9a\u7fa9\u800c\u6210\uff0c\u5728\u77e5\u7db2\u6a19\u8a18\u7fa9\u539f\u7684\u898f\u5247\u3197\uff0c\u5728\u8a5e\u689d\u7684 1-3 \u8a9e\u6599\u5206\u6790\u8207\u8655\u7406 \u56e0\u6b64\u6211\u5011\u50c5\u9700\u8981\u5f97\u77e5\u300c\u8b1b\u300d\u8207\u300c\u5531\u300d\u7684\u76f8\u4f3c\u5ea6\uff0c\u82e5\u300c\u8b1b\u300d\u8207\u300c\u5531\u300d\u5206\u5c6c\u7684\u8a5e\u985e\u76f8 \u6240\u6709\u5b9a\u7fa9\u7fa9\u539f\u3197\uff0c\u7b2c\u3192\u500b\u7fa9\u539f\u3192\u5b9a\u662f\u4e3b\u8981\u610f\u7fa9\u5206\u985e\uff0c\u5f62\u6210\u6982\u5ff5\u9593\u7684\u3196\u3198\u4f4d\u95dc\u4fc2</td><td/></tr><tr><td>\u7684\u8a5e\u985e\u5206\u985e\u5171\u6709 15 \u985e\uff0c\u4f46\u4e26\u975e\u6bcf\u3192\u985e\u90fd\u5177\u6709\u5b73\u751f\u6027\u3002\u6709\u4e9b\u985e\u5225\u5982\u529f\u80fd\u8a5e\u3192\u822c\uff0c \u6211\u5011\u5728\u6b64\u4ecb\u7d39\u672a\u77e5\u52d5\u8a5e\u7684\u7279\u6027\u8207\u53ef\u731c\u6e2c\u672a\u77e5\u52d5\u8a5e\u8a5e\u985e\u7684\u53ef\u80fd\u56e0\u7d20\u3002\u9996\u5148\uff0c\u8a0e \u4f3c\u5ea6\u9ad8\uff0c\u5247\u8868\u793a\u300c\u8b1b\u300d\u8207\u300c\u5531\u300d\u7684\u7d50\u69cb\u985e\u4f3c\uff1b\u82e5\u300c\u8b1b\u300d\u8207\u300c\u5531\u300d\u7684\u8a9e\u610f\u76f8\u4f3c\u7a0b\u5ea6 (is-a relation)\uff0c\u7b2c\u3193\u500b\u4ee5\u5f8c\u7684\u7fa9\u539f\u70ba\u6b21\u8981\u5340\u5206\u8207\u8a5e\u5f59\u4e4b\u9593\u7684\u95dc\u4fc2\u5c31\u4e0d\u78ba\u5b9a\uff0c\u4f9d\u7167</td><td/></tr><tr><td>\u5c6c\u65bc\u5c01\u9589\u6027\u8a5e\u985e\uff0c\u5c01\u9589\u6027\u8a5e\u985e\u70ba\u8a72\u5206\u985e\u3197\u7684\u8a5e\u5f59\u4e0d\u6703\u589e\u52a0\uff0c\u800c\u5728\u3197\u7814\u9662\u8a5e\u5eab\u5c0f\u7d44 \u8ad6\u672a\u77e5\u52d5\u8a5e\u7684\u7279\u6027\u3002\u672a\u77e5\u52d5\u8a5e\u70ba\u8907\u5408\u8a5e\uff0c\u901a\u5e38\u7531\u6578\u500b\u5177\u6709\u5b73\u751f\u6027\u7684\u8a5e\u57fa\u6240\u7d44\u6210\uff0c \u9ad8\u7684\u8a71\uff0c\u5247\u300c\u5531\u5b8c\u300d\u7684\u52d5\u8a5e\u5206\u985e\u5247\u5f88\u53ef\u80fd\u8207\u300c\u8b1b\u5b8c\u300d\u76f8\u540c\u3002 \u77e5\u7db2\u6a19\u8a18\u6c7a\u5b9a\u3002\u8a08\u7b97\u5169\u500b\u8a5e\u689d\u9593\u76f8\u4f3c\u5ea6\u6642\u4e3b\u8981\u7fa9\u539f\u8207\u6574\u500b\u8a5e\u5f59\u4e4b\u9593\u7684\u95dc\u4fc2\u5341\u5206\u91cd</td><td/></tr><tr><td>\u7684\u5206\u985e\u3197 15 \u985e\u3197\u6709 9 \u985e\u662f\u5177\u6709\u5b73\u751f\u6027\u7684\u5206\u985e\uff1b\u9019 9 \u985e\u5206\u985e\u3197\u7684\u52d5\u8a5e\u8a5e\u5f59\uff0c\u6703\u96a8 \u672c\u8eab\u8a9e\u8a00\u5177\u6709\u9ad8\u900f\u660e\u6027\u3002\u4f8b\u5982\uff0c\u672a\u77e5\u52d5\u8a5e\u300c\u6c42\u65b0\u300d\u8207\u300c\u8b1b\u932f\u300d\u76f8\u5c0d\u65bc\u5217\u5165\u8fad\u5178\u3197 \u4f7f\u7528\u76f8\u4f3c\u6cd5\u7684\u597d\u8655\u5728\u65bc\u76f8\u4f3c\u6cd5\u6240\u5c0b\u627e\u7684\u7684\u76f8\u4f3c\u8a5e\uff0c\u82e5\u76f8\u4f3c\u5ea6\u9ad8\u7684\u8a71\uff0c\u4e0d\u50c5\u53ef \u8981\uff0c\u5fc5\u9808\u8207\u5176\u4ed6\u7684\u6b21\u8981\u7fa9\u539f\u5206\u958b\u8a08\u7b97\u3002\u56e0\u6b64</td><td/></tr><tr><td>\u8457\u8a9e\u6599\u5eab\u7684\u589e\u9577\u800c\u589e\u591a\uff0c\u6211\u5011\u5e0c\u671b\u5c07\u672a\u77e5\u52d5\u8a5e\u81ea\u52d5\u5206\u985e\u5230\u9019 9 \u985e\u52d5\u8a5e\u5206\u985e\u3197\uff0c\u9019 \u4e5d\u985e\u70ba\u52d5\u4f5c\u4e0d\u53ca\u7269\u52d5\u8a5e(VA)\u3001\u52d5\u4f5c\u53ca\u7269\u52d5\u8a5e(VC)\u3001\u52d5\u4f5c\u53ca\u7269\u52d5\u8a5e\uff0b\u319e\u65b9\u8cd3\u8a9e (VCL)\u3001\u52d5\u4f5c\u96d9\u8cd3\u52d5\u8a5e(VD)\u3001\u52d5\u4f5c\u53e5\u8cd3\u52d5\u8a5e(VE)\u3001\u5206\u985e\u52d5\u8a5e(VG)\u3001\u72c0\u614b\u4e0d\u53ca\u7269\u52d5 \u8a5e(VH)\u3001\u72c0\u614b\u4f7f\u52d5\u52d5\u8a5e(VHC)\u3001\u72c0\u614b\u53ca\u7269\u52d5\u8a5e(VJ)\u3002 \u7684\u300c\u5fd0\u5fd1\u300d \u3001 \u300c\u4fb7\u4fc3\u300d\u9019\u3192\u985e\u7684\u8a5e\u5f59\u591a\u5177\u6709\u8a9e\u610f\u900f\u660e\u6027\uff0c\u4e26\u4e14\u53ef\u4ee5\u5f9e\u5176\u7d44\u6210\u6210\u5206\u9810 \u6e2c\u51fa\u8a72\u8a5e\u7684\u8a9e\u610f\u3002 \u5176\u6b21\uff0c\u6211\u5011\u8a8d\u70ba\u6709\u5169\u500b\u56e0\u7d20\u53ef\u9810\u6e2c\u672a\u77e5\u52d5\u8a5e\u7684\u5206\u985e\u3002\u3192\u3001\u8a9e\u610f\u3002\u8a9e\u610f\u76f8\u8fd1\u7684 \u8a5e\u5f59\uff0c\u6240\u5c6c\u7684\u8a5e\u985e\u61c9\u985e\u4f3c\u3002\u6211\u5011\u5c07\u540c\u7fa9\u8a5e\u8a5e\u6797\u3197\u7684\u8a9e\u610f\u985e\u8207\u3197\u7814\u9662\u8a5e\u5eab\u5c0f\u7d44(1993) \u4ee5\u9810\u6e2c\u8a5e\u985e\u5206\u985e\uff0c\u540c\u6642\u4e5f\u53ef\u4ee5\u9810\u6e2c\u8a9e\u610f\u8207\u7d50\u69cb\u5206\u985e\u3002\u7576\u5169\u500b\u8a5e\u5f59\u76f8\u4f3c\u5ea6\u9ad8\u6642\uff0c\u8868 \u793a\u9019\u5169\u500b\u8a5e\u5f59\u7684\u8a5e\u985e\u3001\u8a9e\u610f\u985e\u8207\u7d50\u69cb\u5fc5\u5b9a\u76f8\u4f3c\u3002 \u6211\u5011\u5728\u672c\u7bc0\u3197\u9996\u5148\u4ecb\u7d39\u8a9e\u610f\u8207\u8a5e\u985e\u76f8\u4f3c\u5ea6\u7684\u6e2c\u91cf\u65b9\u6cd5\uff0c\u63a5\u3198\u4f86\u8aaa\u660e\u76f8\u4f3c\u8a5e\u7684 \u9078\u53d6\u8207\u672a\u77e5\u52d5\u8a5e\u8a5e\u985e\u7684\u9810\u6e2c\u3002 ( ) ( y 2 x 1 Entry Word , Entry Word core HowNetSimS \u2212 \u2212 ) ( )( ( ) m y, y,2 n x, x,2 2 y,1 x,1 1 ...Sem Sem , ...Sem Sem core SecondaryS * w Sem Sem re PrimarySco * w + \u2229 = )</td><td/></tr><tr><td>\u8a5e\u985e\u4f5c\u5c0d\u61c9\uff0c\u3197\u7814\u9662\u8a5e\u5eab\u5c0f\u7d44\u8a5e\u985e\u6709 45 \u985e\u3002\u5e73\u5747\u4f86\u8aaa\uff0c\u540c\u7fa9\u8a5e\u8a5e\u6797\u3192\u500b\u8a9e\u610f\u985e</td><td/></tr><tr><td>1-2 \u7814\u7a76\u65b9\u6cd5 \u50c5\u5c0d\u61c9\u5230\u8a5e\u5eab\u5c0f\u7d44 1.97 \u7a2e\u8a5e\u985e\uff0c\u5373\u3192\u500b\u8a9e\u610f\u985e\u3197\u7684\u8a5e\u5f59\u5171\u6709\u7684\u8a5e\u985e\u6578\u91cf\u3002\u56e0\u6b64 \u6211\u5011\u8a8d\u70ba\u8a9e\u610f\u56e0\u7d20\u53ef\u5de6\u53f3\u8a5e\u5f59\u7684\u8a5e\u985e\u3002\u3193\u3001\u7d50\u69cb\u3002\u7d50\u69cb\u901a\u5e38\u6703\u9650\u5b9a\u7d44\u6210\u7684\u8a5e\u985e\uff0c \u77e5\u7db2\u3197\u6709\u63cf\u8ff0\u7fa9\u539f\u8207\u7fa9\u539f\u4e4b\u9593\u7684\u968e\u5c64\u95dc\u4fc2\u7684\u5206\u985e\u6a39\uff0c\u6211\u5011\u5728\u9019\u908a\u5229\u7528\u63cf\u8ff0 2-2 \u76f8\u4f3c\u5ea6\u6e2c\u91cf \u7fa9\u539f\u4e4b\u9593\u95dc\u4fc2\u7684\u5206\u985e\u6a39\u4f86\u5e6b\u52a9\u6211\u5011\u8a08\u7b97\u7fa9\u539f\u9593\u7684\u76f8\u4f3c\u5ea6\u3002\u9673\u514b\u5065\u3001\u9673\u8d85\u7136</td><td/></tr><tr><td>\u672c\u8ad6\u6587\u3197\u672a\u77e5\u8a5e\u7684\u5b9a\u7fa9\u70ba\u4e0d\u5b58\u5728\u8fad\u5178\u3197\u7684\u8a5e\u5f59\u3002\u9673\u514b\u5065\u3001\u9673\u8d85\u7136(1997)\u5206\u6790 \u82e5\u7d50\u69cb\u70ba\"VC+Na\"\u7684\u672a\u77e5\u52d5\u8a5e\uff0c\u901a\u5e38\u6703\u7d44\u6210 VA \u8a5e\u985e\uff0c\u56e0\u70ba\u5728\u9019\u500b\u672a\u77e5\u52d5\u8a5e\u7684\u5167 \u5728\u672c\u8ad6\u6587\u3197\u6211\u5011\u4f7f\u7528\u77e5\u7db2\u4f5c\u70ba\u8a9e\u610f\u6e2c\u91cf\u7684\u5de5\u5177\uff0c\u3197\u592e\u7814\u7a76\u9662\u3197\u6587\u53e5\u7d50\u69cb\u6a39\u6e2c (1997:270)\u8a8d\u70ba\u5169\u500b\u8a9e\u610f\u985e\u7684\u76f8\u4f3c\u5ea6\u5728\u65bc\u5169\u500b\u8a9e\u610f\u985e\u5728\u5206\u985e\u6a39\u4ea4\u96c6\u7bc0\u9ede\u7684\u8a9e\u610f\u8a0a</td><td/></tr><tr><td>\u672a\u77e5\u8a5e\u7684\u7a2e\u985e\u70ba\u5169\u7a2e\uff0c\u7b2c\u3192\u7a2e\u70ba\u5c01\u9589\u6027\uff0c\u9019\u3192\u985e\u578b\u96d6\u7136\u5728\u6578\u91cf\u3196\u53ef\u80fd\u70ba\u7121\u6578\u500b\uff0c \u90e8\u7d50\u69cb\u3197\u5df2\u7d93\u51fa\u73fe\u4e86\u3192\u500b\u666e\u901a\u540d\u8a5e(Na)\u4f86\u6eff\u8db3\u524d\u9762\u7684\u52d5\u4f5c\u53ca\u7269\u52d5\u8a5e(VC)\u6240\u8981\u6c42 \u91cf\u8a5e\u985e\u76f8\u4f3c\u5ea6\uff0c\u4ecb\u7d39\u5982\u3198\u3002 \u606f\u91cf(Information Content)\uff0c\u5c07\u6574\u500b\u8a5e\u5206\u985e\u67b6\u69cb\u770b\u6210\u3192\u500b\u8a0a\u606f\u7cfb\u7d71\uff0c\u3192\u500b\u8a9e\u610f\u985e</td><td/></tr><tr><td>\u4f46\u662f\u53ef\u7528\u898f\u5247\u8a9e\u6cd5(Regular Expression)\u4f86\u7522\u751f\u8207\u8fa8\u8b58\uff0c\u5982\uff1a\u897f\u5143\u3192\u4e5d\u4e5d\u4e5d\u5e74(\u6642 \u7684\u8ad6\u5143\uff0c\u5728\u9019\u7a2e\u60c5\u5f62\u3198\u901a\u5e38\u6703\u5f62\u6210\u4e0d\u53ca\u7269\u52d5\u8a5e\uff0c\u56e0\u6b64\u6211\u5011\u8a8d\u70ba\u7d50\u69cb\u6703\u5f71\u97ff\u5230\u52d5\u8a5e \u3192\u3001\u77e5\u7db2\u70ba\u3192\u96d9\u8a9e(\u3197\u6587\u3001\u82f1\u6587)\u7684\u77e5\u8b58\u6027\u8fad\u5178\uff0c\u7531\u8463\u632f\u6771\u8207\u8463\u5f37\u7de8\u64b0\u5b8c\u6210\u6536 Sem (\u76f8\u7576\u65bc\u77e5\u7db2\u3197\u7684\u7fa9\u539f)\u7684\u8a0a\u606f\u91cf\u5b9a\u7fa9\u70ba Entropy(System)-Entropy(Sem)\u3002\u6211\u5011</td><td/></tr><tr><td>\u9593)\u3001\u3192\u5343\u5169\u767e\u4e03\u5341\u3193(\u6578\u5b57)\u3001\u3193\u4e03\u516b\u516b\u3194\u4e03\u4e5d\u4e5d(\u96fb\u8a71)\u7b49\u3002\u7b2c\u3193\u985e\u5247\u70ba\u958b\u653e\u6027\uff0c \u7684\u8a5e\u985e\u3002 \u9304\u7d04\u5341\u3192\u842c\u689d\u8a5e\u689d\uff0c\u77e5\u7db2\u7cfb\u7d71\u3197\u5305\u542b\u6709\u3197\u82f1\u96d9\u8a9e\u77e5\u8b58\u8fad\u5178\u3001\u3197\u6587\u7c21\u9ad4\u77e5\u8b58\u8fad\u5178\u3001 \u5728\u9019\u908a\u4f7f\u7528\u9673\u514b\u5065\u3001\u9673\u8d85\u7136(1997)\u8a08\u7b97\u8a9e\u610f\u8a0a\u606f\u91cf\u7684\u65b9\u6cd5\u4f86\u8a08\u7b97\u77e5\u7db2\u3197\u5404\u7fa9\u539f\u7684</td><td/></tr><tr><td>\u9019\u3192\u985e\u7684\u672a\u77e5\u8a5e\u5f88\u96e3\u7528\u898f\u5247\u8a9e\u6cd5\u4f86\u8868\u9054\uff0c\u8907\u5408\u8a5e\u5373\u5c6c\u9019\u3192\u985e\u3002\u767d\u660e\u5b8f\u3001\u9673\u8d85\u7136\u8207 \u5728\u672c\u7bc7\u8ad6\u6587\u3197\u6211\u5011\u5229\u7528\u9019\u4e9b\u7dda\u7d22\u5c0b\u627e\u8207\u672a\u77e5\u52d5\u8a5e\u76f8\u4f3c\u7684\u8a5e\u5f59\uff0c\u4f86\u9810\u6e2c\u672a\u77e5\u52d5 \u3197\u6587\u7e41\u9ad4\u77e5\u8b58\u8fad\u5178\u3001\u6982\u5ff5\u7279\u5fb5\u3001\u52d5\u614b\u89d2\u8272\u8207\u5c6c\u6027\u3001\u8a5e\u985e\u8868\u3001\u53cd\u7fa9\u95dc\u4fc2\u8868\u3001\u5c0d\u7fa9\u95dc \u8a0a\u606f\u91cf\u3002 1. \u7dd2\u8ad6 \u9673\u514b\u5065(1998)\u5728\u5206\u6790\u3197\u7814\u9662\u5e73\u8861\u8a9e\u6599\u5eab\u5f8c\u6b78\u7d0d\u51fa\u672a\u77e5\u8a5e\u4e3b\u8981\u7684\u5206\u985e\u70ba\u7565\u8a9e\u3001\u5c08\u6709 \u8a5e\u6240\u5c6c\u7684\u8a5e\u985e\u3002 \u4fc2\u8868\u3001\u6a19\u793a\u7b26\u865f\u8207\u8aaa\u660e\u3001\u77e5\u7db2\u7ba1\u7406\u7a0b\u5e8f\u7b49\u3002\u6211\u5011\u5728\u672c\u7bc0\u7576\u3197\u5c07\u4ecb\u7d39\u5982\u4f55\u4f7f\u7528\u77e5\u7db2 \u77e5\u7db2\u3197\u5169\u500b\u7fa9\u539f\u7684\u76f8\u4f3c\u5ea6\u70ba\u9019\u5169\u500b\u7fa9\u539f\u6240\u4ea4\u96c6\u7bc0\u9ede\u7684\u8a9e\u610f\u8a0a\u606f\u91cf\uff0c\u6240\u5f97\u5230\u8a9e</td><td/></tr><tr><td>\u81ea\u7136\u8a9e\u8a00\u8655\u7406\u3197\u91cd\u8981\u7684\u6b65\u9a5f\u662f\u5c07\u3197\u6587\u6587\u4ef6\u65b7\u8a5e\u4e26\u9644\u52a0\u8a5e\u985e\u6a19\u8a18\uff1b\u5728\u65b7\u8a5e\u6a19\u8a18 \u540d\u8a5e\u3001\u884d\u751f\u8a5e\u3001\u8907\u5408\u8a5e\u8207\u6578\u5b57\u578b\u8907\u5408\u8a5e\u3002 \u8a08\u7b97\u8a9e\u610f\u76f8\u4f3c\u5ea6\u8207\u8a55\u91cf\u65b9\u6cd5\u3002 \u610f\u8a0a\u606f\u91cf\u8d8a\u9ad8\u8868\u793a\u9019\u5169\u500b\u7fa9\u539f\u8d8a\u76f8\u4f3c\uff0c\u56e0\u6b64\u7b2c\u3192\u90e8\u4efd\u7684\u76f8\u4f3c\u5ea6\u5b9a\u7fa9\u5982\u3198\uff1a</td><td/></tr><tr><td>\u7684\u904e\u7a0b\u3197\u6703\u9047\u5230\u7684\u3192\u500b\u554f\u984c\u70ba\u672a\u77e5\u8a5e\u7684\u5b58\u5728\u3002\u73fe\u884c\u7684\u65b7\u8a5e\u6a19\u8a18\u7cfb\u7d71\u4ee5\u8fad\u5178\u70ba\u57fa\u790e \u8f14\u4ee5\u69cb\u8a5e\u7684\u898f\u5247\u8a0a\u606f\u9032\u884c\u65b7\u8a5e\u6a19\u8a18\uff0c\u4f46\u56e0\u70ba\u8a9e\u8a00\u7684\u7279\u6027\u4e4b\u3192\u300c\u7121\u7aae\u76e1\u7684\u5275\u9020\u529b\u300d \uff0c \u7121\u6cd5\u7aae\u8209\u51fa\u6240\u6709\u7684\u8a5e\u5f59\uff1b\u3192\u672c\u597d\u7684\u8fad\u5178\u4e5f\u4e0d\u61c9\u8a72\u7121\u6b62\u76e1\u7684\u64f4\u5927\u6240\u6536\u9304\u7684\u8a5e\u5f59\uff0c\u56e0 \u6b64\u5982\u4f55\u8fa8\u8b58\u8655\u7406\u8fad\u5178\u3197\u4e0d\u5b58\u5728\u7684\u8a5e\u5f59\u5c31\u6210\u4e86\u3192\u500b\u91cd\u8981\u7684\u8ab2\u984c\u3002 \u672a\u77e5\u52d5\u8a5e\u901a\u5e38\u70ba\u8907\u5408\u8a5e\uff0c\u7531\u5169\u500b\u4ee5\u3196\u7684\u7d44\u6210\u6210\u5206\u7d44\u5408\u800c\u6210\uff0c\u9019\u7a2e\u7d44\u6210\u6210\u5206\u6211 \u5011\u7a31\u70ba\u8a5e\u57fa(base) 2 \u3002\u8d99\u5143\u4efb(1968)\u3001Li \u8207 Thompson (1981)\u8207\u6e6f\u5ef7\u6c60(1988)\u63d0\u53ca\u6f22 \u8a9e\u7684\u8907\u5408\u8a5e\u5177\u6709\u7279\u5b9a\u7684\u5167\u90e8\u53e5\u6cd5\u7d50\u69cb\uff1b\u5982\uff1a \u300c\u6b3a\u6575\u300d \uff0c\u7531\u300c\u6b3a\u300d\u8207\u300c\u6575\u300d\u9019\u5169\u500b\u8a5e \u57fa\u7d44\u6210\uff0c\u5169\u500b\u8a5e\u57fa\u4e4b\u9593\u7684\u95dc\u4fc2\u70ba\u52d5\u8cd3\u7d50\u69cb\u3002\u96d6\u7136\u8a5e\u57fa\u662f\u6709\u9650\u7684\uff0c\u4f46\u662f\u8a5e\u57fa\u8207\u8a5e\u57fa 2. \u5be6\u9a57\u65b9\u6cd5 \u6211\u5011\u5229\u7528\u76f8\u4f3c\u6cd5\u4f86\u5224\u65b7\u52d5\u8a5e\u7684\u5206\u985e\uff0c\u5c0b\u627e\u672a\u77e5\u52d5\u8a5e\u7684\u76f8\u4f3c\u8a5e\uff0c\u8a08\u7b97\u672a\u77e5\u52d5\u8a5e \u8207\u76f8\u4f3c\u8a5e\u4e4b\u9593\u7684\u76f8\u4f3c\u5ea6\uff0c\u518d\u5c07\u9019\u4e9b\u76f8\u4f3c\u8a5e\u4f9d\u7167\u8a5e\u985e\u5206\u7d44\u3002\u5f9e\u6bcf\u500b\u8a5e\u985e\u7576\u3197\u53d6\u51fa K \u3193\u3001\u3197\u592e\u7814\u7a76\u9662\u3197\u6587\u53e5\u7d50\u69cb\u6a39\u8cc7\u6599\u5eab 1.0 \u3197\u5305\u542b\u4e86\u5341\u500b\u6a94\u6848\uff0c\u3194\u842c\u516b\u5343\u4e03\u767e \u3193\u5341\u4e94\u68f5\u3197\u6587\u7d50\u69cb\u6a39\uff0c\u542b\u6709\u3193\u5341\u3194\u842c\u4e5d\u5343\u4e94\u767e\u3194\u5341\u3193\u500b\u8a5e\u8a5e\u5f59\uff0c\u6bcf\u3192\u53e5\u7d50\u69cb\u6a39\uff0c \u6a19\u793a\u6f22\u8a9e\u53e5\u6cd5\u8207\u8a9e\u610f\u8a0a\u606f\uff0c\u8a5e\u985e\u6a19\u8a18\u8207\u65b7\u8a5e\u6a19\u8a18\u7cfb\u7d71\u3195\u5341\u4e94\u500b\u6a19\u8a18\u4e0d\u540c\uff0c\u7d50\u69cb\u6a39 \u3197\u7684\u6a19\u8a18\u662f\u7531\u3195\u5341\u4e94\u500b\u6a19\u8a18\u7d30\u5206\u800c\u6210\u3002\u5728\u672c\u7bc0\u3197\u6211\u5011\u5229\u7528\u3197\u7814\u9662\u3197\u6587\u53e5\u7d50\u69cb\u6a39\u6e2c ( ) ( ) ( ) ( ( ) ( ) ( System /Entropy Sem Sem Entropy System Entropy ) System /Entropy Sem Sem nContent Informatio ) Sem Sem re PrimarySco</td><td/></tr><tr><td>\u7684\u7d44\u5408\u6578\u91cf\u9f90\u5927\uff0c\u4e14\u7d44\u5408\u6210\u5206\u9593\u7684\u8a9e\u610f\u95dc\u4fc2\u8907\u96dc\uff0c\u56e0\u6b64\u9020\u6210\u4e86\u6211\u5011\u7121\u6cd5\u5c07\u6240\u6709\u7684 \u500b\u76f8\u4f3c\u8a5e\u51fa\u4f86\uff0c\u5c07\u9019\u4e9b\u76f8\u4f3c\u8a5e\u7684\u5206\u6578\u4e88\u4ee5\u5e73\u5747\uff0c\u5f97\u5230\u672a\u77e5\u52d5\u8a5e\u5230\u6bcf\u500b\u8a5e\u985e\u7684\u5e73\u5747 \u91cf\u8a5e\u985e\u7684\u76f8\u4f3c\u5ea6\u3002</td><td/></tr><tr><td>\u672a\u77e5\u52d5\u8a5e\u6536\u9304\u9032\u5b57\u5178\u3197\u3002 \u8ddd\u96e2\uff0c\u672a\u77e5\u52d5\u8a5e\u7684\u8a5e\u985e\u5373\u8207\u5176\u8ddd\u96e2\u6700\u76f8\u8fd1\u7684\u8a5e\u985e\u3002 1-1 \u7814\u7a76\u52d5\u6a5f\u8207\u76ee\u6a19 \u524d\u319f\u5c0d\u65bc\u672a\u77e5\u8a5e\u7684\u63a2\u8a0e\u91cd\u9ede\u96c6\u3197\u5728\u540d\u8a5e\u7d30\u76ee\u7684\u8fa8\u8a8d\u3196\uff0c\u5982\u7d44\u7e54\u540d\u3001\u319f\u540d\u3001\u319e \u540d\u8fa8\u8b58\u7b49(\u674e\u632f\u660c(1993)\uff0c\u674e\u632f\u660c\u3001\u674e\u5fa1\u74bd\u8207\u9673\u4fe1\u5e0c(1994)\u319f\u540d\u8fa8\u8b58\u7b49\u7b49)\u3002\u50c5\u6709 \u5728\u672c\u8ad6\u6587\u3197\u6211\u5011\u5229\u7528\u76f8\u4f3c\u6cd5\u4f86\u5224\u65b7\u52d5\u8a5e\u7684\u5206\u985e\uff0c\u5c0b\u627e\u672a\u77e5\u52d5\u8a5e\u7684\u76f8\u4f3c\u8a5e\uff0c\u8a08 2-2-1 \u8a9e\u610f\u76f8\u4f3c\u5ea6\u6e2c\u91cf \u7b97\u672a\u77e5\u52d5\u8a5e\u8207\u76f8\u4f3c\u8a5e\u4e4b\u9593\u7684\u76f8\u4f3c\u5ea6\uff0c\u518d\u5c07\u9019\u4e9b\u76f8\u4f3c\u8a5e\u4f9d\u7167\u8a5e\u985e\u5206\u7d44\u3002\u5f9e\u6bcf\u500b\u8a5e\u985e \u7576\u3197\u53d6\u51fa K \u500b\u76f8\u4f3c\u8a5e\u51fa\u4f86\uff0c\u5c07\u9019\u4e9b\u76f8\u4f3c\u8a5e\u7684\u5206\u6578\u4e88\u4ee5\u5e73\u5747\uff0c\u5f97\u5230\u672a\u77e5\u52d5\u8a5e\u5230\u6bcf \u500b\u8a5e\u985e\u7684\u5e73\u5747\u8ddd\u96e2\uff0c\u672a\u77e5\u52d5\u8a5e\u7684\u8a5e\u985e\u5373\u8207\u5176\u8ddd\u96e2\u6700\u76f8\u8fd1\u7684\u8a5e\u985e\u3002 \u77e5\u7db2\u7d04\u9078\u7528\u4e86\u3192\u5343\u4e03\u767e\u591a\u500b\u7fa9\u539f\u4f86\u5b9a\u7fa9\u3197\u82f1\u96d9\u8a9e\u77e5\u8b58\u8fad\u5178\u3197\u7684\u6bcf\u500b\u8a5e\uff0c\u4e26\u4e14 2-\u6211\u5011\u5728\u9019\u7bc0\u8aaa\u660e\u5982\u4f55\u4f7f\u7528\u76f8\u4f3c\u6cd5\u4f86\u9810\u6e2c\u52d5\u8a5e\u7684\u5206\u985e\u3002\u672a\u77e5\u52d5\u8a5e\u7684\u7279\u6027\u4e4b\u3192\u70ba \u7d44\u6210\u6210\u5206\u5c6c\u65bc\u5e38\u7528\u8a5e\u4e14\u8a9e\u610f\u660e\u78ba\uff0c\u4f8b\u5982\uff1a\u8a66\u5370\u3001\u8b1b\u5b8c\u3002\u9019\u5169\u500b\u8a5e\u5f59\u90fd\u7121\u6cd5\u5728\u8fad\u5178 \u3197\u67e5\u8a62\u5230\uff0c\u4f46\u6211\u5011\u537b\u5f88\u6e05\u695a\u7684\u53ef\u4ee5\u5f9e\u5b57\u9762\u3196\u5f97\u77e5\u9019\u5169\u500b\u52d5\u8a5e\u7684\u8a9e\u610f\uff0c\u800c\u4e14\u9019\u6a23\u7684 \u5efa\u6709\u63cf\u8ff0\u5404\u500b\u7fa9\u539f\u4e4b\u9593\u7684\u95dc\u4fc2\u7684\u5206\u985e\u6a39\u3002\u4f8b\u5982\uff1a \u300c\u8b80\u66f8\u300d\u3192\u8a5e\u7531\u300c\u5f9e\u4e8b\u300d \u3001 \u300c\u5b78\u300d \u8207\u300c\u6559\u80b2\u300d\u3194\u500b\u7fa9\u539f\u5b9a\u7fa9\u800c\u6210\uff0c\u77e5\u7db2\u3197\u4e26\u6709\u5206\u985e\u6a39\u8868\u793a\u300c\u5f9e\u4e8b\u300d \u3001 \u300c\u5b78\u300d\u8207\u300c\u6559\u80b2\u300d \u3194\u500b\u7fa9\u539f\u4e4b\u9593\u7684\u95dc\u4fc2\u3002 () Chen\u3001Bai \u8207 \u4e4b\u8655\uff0c\u5c07\u6b63\u78ba\u7387\u63d0\u9ad8\u81f3 83.83%\u3002\u5728\u52d5\u8a5e\u8fa8\u8b58\u6b63\u78ba\u7d50\u679c\u4e0d\u9ad8\u7684\u60c5\u6cc1\u3198\uff0c\u672c\u8ad6\u6587\u5c07 \u8655\u7406\u91cd\u5fc3\u653e\u5728\u672a\u77e5\u52d5\u8a5e\u7684\u8fa8\u8b58\u8655\u7406\u3196\uff0c\u4e26\u4e14\u5e0c\u671b\u5c07\u9019\u7a2e\u8655\u7406\u672a\u77e5\u52d5\u8a5e\u7684\u65b9\u6cd5\u5728\u672a \u6839\u64da\u6211\u5011\u5c0d\u672a\u77e5\u52d5\u8a5e\u8a9e\u6599\u7684\u89c0\u5bdf\uff0c\u672a\u77e5\u52d5\u8a5e\u7684\u7d44\u6210\u96d6\u7136\u6709\u3192\u5b9a\u7684\u6a21\u5f0f\uff0c\u4f46\u56e0 \u6b64\u6211\u5011\u5728\u9019\u908a\u5b9a\u7fa9\u5169\u500b\u8a5e Word 1 ,Word 2 \u9593\u7684\u76f8\u4f3c\u5ea6\u76f8\u7b49\u65bc\u5169\u500b\u8a5e\u5404\u5c6c\u7684\u8a5e\u689d\u9593\u6700 2 Sproat \u8207 Shih (1996) \u7a31\u5167\u90e8\u7684\u8655\u7406\u55ae\u4f4d\u70ba\u8a5e\u6839(root)\uff0cChen\u3001Bai \u8207 Chen (1997)\u7a31\u8655\u7406\u7684\u55ae\u4f4d\u70ba \u7d44\u5408\u65b9\u5f0f\u662f\u975e\u5e38\u5177\u6709\u5b73\u751f\u6027\u7684\uff0c\u53ef\u4ee5\u7e7c\u7e8c\u5b73\u751f\u300c\u5531\u5b8c\u300d \u3001 \u300c\u8aaa\u5b8c\u300d\u7b49\u7b49\u5404\u6a23\u7684\u8a5e\u5f59\u3002 \u3192\u822c\u4f86\u8aaa\uff0c\u3192\u500b\u8a5e\u5728\u77e5\u7db2\u3197\u53ef\u80fd\u64c1\u6709\u591a\u500b\u8a5e\u689d\uff0c\u539f\u56e0\u5728\u65bc\u8a5e\u5f59\u7684\u591a\u7fa9\u6027\uff0c\u56e0 ( ) ()</td><td/></tr><tr><td>\u4f86\u53ef\u4ee5\u8f49\u79fb\u8655\u7406\u540d\u8a5e\u8207\u5f62\u5bb9\u8a5e\u3002 \u524d\u7db4(prefix)\u8207\u5f8c\u7db4(suffix)\u3002\u6211\u5011\u5247\u7a31\u8655\u7406\u55ae\u4f4d\u70ba\u8a5e\u57fa(base)\uff0c\u4e26\u63a1\u7528 Katamba (1993:45) \u5c0d\u8a5e\u57fa \u70ba\u8a9e\u8a00\u7684\u8907\u96dc\u5ea6\uff0c\u7121\u6cd5\u5c07\u6240\u6709\u7684\u898f\u5247\u689d\u5217\u51fa\u4f86\u3002\u56e0\u6b64\u6211\u5011\u5728\u9019\u908a\u4f7f\u7528\u76f8\u4f3c\u6cd5\uff0c\u5c07 \u5927\u76f8\u4f3c\u5ea6\u3002</td><td/></tr><tr><td>\u52d5\u8a5e\u4e0d\u7ba1\u5728\u4efb\u4f55\u6587\u6cd5\u7406\u8ad6\u3197\uff0c\u5728\u5256\u6790\u53e5\u5b50\u6642\u90fd\u662f\u4f4d\u65bc\u6700\u3197\u5fc3\u7684\u90e8\u5206\uff0c\u82e5\u52d5\u8a5e \u70ba\u672a\u77e5\u8a5e\uff0c\u52e2\u5fc5\u5c07\u5f71\u97ff\u53e5\u5b50\u5256\u6790\u7684\u6b63\u78ba\u6027\u3002\u73fe\u4ee3\u6f22\u8a9e\u7684\u52d5\u8a5e\u7d50\u69cb\u7e41\u8907\uff0c\u5167\u90e8\u898f\u5247 \u8907\u96dc\uff0c\u82e5\u7121\u8db3\u5920\u7684\u8a9e\u8a00\u8a0a\u606f\u5b8c\u5168\u7121\u6cd5\u5224\u65b7\u5176\u5206\u985e\uff0c\u6211\u5011\u8a8d\u70ba\u52d5\u8a5e\u81ea\u52d5\u5206\u985e\u7814\u7a76\u81f3 \u77e5\u52d5\u8a5e\u8207\u8a13\u7df4\u8a9e\u6599\u3197\u7684\u52d5\u8a5e\u8d8a\u76f8\u4f3c\u6642\uff0c\u65b0\u7684\u672a\u77e5\u52d5\u8a5e\u8d8a\u6709\u53ef\u80fd\u5c6c\u65bc\u8207\u5176\u76f8\u4f3c\u52d5\u8a5e (base)\u5728\u6b64\u8655\u6c7a\u5b9a\u4f7f\u7528\u8a5e\u57fa\u70ba\u6211\u5011\u5207\u5272\u7684\u55ae\u4f4d\u7684\u539f\u56e0\u5728\u65bc\u8a5e\u57fa\u7684\u5b9a\u7fa9\u8f03\u8a5e\u6839 (root) \u3001\u8a5e\u5e79 (stem) \u5bec \u6bcf\u500b\u8a13\u7df4\u8a9e\u6599\u3197\u7684\u672a\u77e5\u52d5\u8a5e\u90fd\u7576\u4f5c\u662f\u3192\u689d\u898f\u5247\uff0c\u7576\u6709\u65b0\u7684\u672a\u77e5\u52d5\u8a5e\u51fa\u73fe\u6642\uff0c\u5c07\u5176 \u8207\u6240\u6709\u7684\u52d5\u8a5e\u505a\u6bd4\u8f03\uff0c\u6e2c\u91cf\u65b0\u7684\u672a\u77e5\u52d5\u8a5e\u8207\u8a13\u7df4\u8a9e\u6599\u3197\u7684\u52d5\u8a5e\u7684\u76f8\u4f3c\u5ea6\uff0c\u65b0\u7684\u672a ( ) ( 2 x 1 2 1 Entry y Word , Entry Word imScore maxHowNetS Word , Word core HowNetSimS \u2212 \u2212 =</td><td>)</td></tr><tr><td>\u9b06\u3002\u672a\u77e5\u52d5\u8a5e\u88ab\u6211\u5011\u65b7\u8a5e\u7cfb\u7d71\u5207\u5206\u51fa\u4f86\u5f88\u591a\u55ae\u4f4d\uff0c\u6211\u5011\u4e26\u4e0d\u78ba\u5b9a\u9019\u4e9b\u55ae\u4f4d\u771f\u6b63\u7684\u610f\u7fa9\uff0c\u56e0\u6b64\u6211\u5011 \u7684\u8a5e\u985e\u3002\u4f8b\u5982\uff1a\u8b1b\u5b8c\u8207\u5531\u5b8c\u3002\u82e5\u300c\u8b1b\u5b8c\u300d\u6211\u5011\u8a13\u7df4\u8a9e\u6599\u3197\u7684\u52d5\u8a5e\uff0c \u300c\u5531\u5b8c\u300d\u70ba\u6211 \u5176\u6b21\uff0c\u6bcf\u3192\u500b\u8a5e\u689d\u53ef\u80fd\u7531\u3192\u5230\u516b\u500b\u7fa9\u539f\u5b9a\u7fa9\u800c\u6210\uff0c\u5982\u300c\u8b80\u66f8\u300d\u3192\u8a5e\u7531\u300c\u5f9e</td><td/></tr><tr><td>\u5e0c\u671b\u9078\u7528\u3192\u500b\u6700\u5bec\u9b06\u7684\u5b9a\u7fa9\u53ef\u4ee5\u6db5\u84cb\u6240\u6709\u88ab\u65b7\u8a5e\u7cfb\u7d71\u6240\u5207\u5206\u7684\u55ae\u4f4d\u3002</td><td/></tr></table>",
+                "text": "Chen(1997)\u5229\u7528\u524d\u7db4(prefix)\u3001\u5f8c\u7db4(suffix)\u7684\u8a0a\u606f\u8655\u7406\u5168\u90e8\u7684\u672a\u77e5\u8a5e\uff0c \u6b63\u78ba\u7387\u7d04\u70ba 76%\uff0c\u800c\u767d\u660e\u5b8f\u3001\u9673\u8d85\u7136\u8207\u9673\u514b\u5065(1998)\u4f7f\u7528 Chen\u3001Bai \u8207Chen (1997) \u6240\u63d0\u51fa\u7684\u65b9\u6cd5\uff0c\u518d\u5229\u7528\u524d\u5f8c\u6587\u7684\u8a0a\u606f\u4f86\u88dc\u5f37 Chen\u3001Bai \u8207 Chen (1997)\u65b9\u6cd5\u4e0d\u8db3 \u6240\u3198\u5b9a\u7fa9\uff1a\"\u2026a base is any unit whatsoever to which affixes of any kind can be added\u2026.In other words, all roots are bases. Bases are called stems only in the context of inflectional morphology.\" \u6211\u5011",
+                "num": null,
+                "html": null
+            },
+            "TABREF1": {
+                "type_str": "table",
+                "content": "<table><tr><td colspan=\"3\">( Category \u82e5\u63a1\u7528\u9019\u7a2e\u65b9\u6cd5\u5fc5\u9808\u8a08\u7b97\u8a13\u7df4\u8a9e\u6599\u3197\u7684\u6bcf\u3192\u500b\u8a5e\u5f59\u8207\u6211\u5011\u672a\u77e5\u52d5\u8a5e\u7684\u76f8\u4f3c ) j i j i j i Category Category Category Category Category ore CategorySc , \u2022 = * \u5ea6\uff0c\u5c07\u6703\u6d6a\u8cbb\u8a31\u591a\u4e0d\u5fc5\u8981\u7684\u8a08\u7b97\u6642\u9593\uff0c\u56e0\u6b64\u50c5\u5c31\u8a13\u7df4\u8a9e\u6599\u3197\u8207\u65b0\u7684\u672a\u77e5\u52d5\u8a5e\u524d\u8a5e</td></tr><tr><td colspan=\"3\">\u57fa\u76f8\u540c\u8207\u5f8c\u8a5e\u57fa\u76f8\u540c\u7684\u76f8\u4f3c\u8a5e\u70ba\u8a08\u7b97\u6a19\u7684\u3002\u5c0b\u627e\u5230\u524d\u8a5e\u57fa\u76f8\u540c\u8207\u5f8c\u8a5e\u57fa\u76f8\u540c\u7684\u76f8</td></tr><tr><td colspan=\"3\">\u4f3c\u8a5e\u5f8c\uff0c\u7b2c\u3193\u6b65\u9700\u8a08\u7b97\u9019\u4e9b\u9078\u53d6\u51fa\u4f86\u7684\u76f8\u4f3c\u8a5e\u3197\u8207\u65b0\u7684\u672a\u77e5\u52d5\u8a5e\u8a5e\u57fa\u76f8\u7570\u7684\u90e8\u5206</td></tr><tr><td colspan=\"3\">\u53ef\u6b78\u7d0d\u51fa\u53f3\u908a\u7684\u3194\u689d\u898f\u5247\uff0c\u898f\u5247\u4e4b\u524d\u7684\u6578\u91cf\u8868\u793a\u898f\u5247\u51fa\u73fe\u7684\u6b21\u6578\u3002\u3198\u5716\u70ba\u3197 \u7684\u76f8\u4f3c\u5ea6\u3002\u8a08\u7b97\u5169\u500b\u8a5e\u5f59\u76f8\u4f3c\u5ea6\u7684\u65b9\u6cd5\uff0c\u5982\u3198\uff1b</td></tr><tr><td>\u7814\u9662\u3197\u6587\u53e5\u7d50\u69cb\u6a39\u7684\u7bc4\u4f8b\uff1a</td><td/><td/></tr><tr><td>Sim(Word unknown ,Word known )</td><td/><td/></tr><tr><td>=w 1 *Score 1 +w 2 *Score 2</td><td/><td/></tr><tr><td colspan=\"3\">S VH \u985e\u8207 VA \u985e\u540c\u5c6c\u4e0d\u53ca\u7269\u52d5\u8a5e\uff0c\u4ed6\u5011\u7684\u5dee\u5225\u50c5\u5728\u65bc\u52d5\u4f5c\u8207\u72c0\u614b\u7684\u5340\u5206\u3002 =w 1 *HowNetSimScore(Base i ,Base j )</td></tr><tr><td colspan=\"3\">\u8868\u683c 1 \u8a5e\u985e\u76f8\u4f3c\u5ea6(\u90e8\u5206) +w 2 *CategoryScore(category(Base i ),category(Base j ))</td></tr><tr><td colspan=\"3\">\u8a5e\u985e 1 Word known \u70ba\u76f8\u4f3c\u8a5e theme NP quantity NP VH Base i \u70ba\u672a\u77e5\u52d5\u8a5e\u8207\u76f8\u4f3c\u8a5e\u76f8\u7570\u7684\u8a5e\u57fa \u8a5e\u985e 2 VA 1 quantity NP --&gt; Head_Neqa \u76f8\u4f3c\u5ea6 0.674 1 theme NP --&gt; head_Nad VH VC 0.611 Base j \u70ba\u76f8\u4f3c\u8a5e\u8207\u672a\u77e5\u52d5\u8a5e\u76f8\u7570\u7684\u8a5e\u57fa</td></tr><tr><td colspan=\"3\">VH VH \u6700\u5f8c\u3192\u500b\u6b65\u9a5f\u662f\u6c7a\u5b9a\u672a\u77e5\u52d5\u8a5e\u7684\u8a5e\u985e\u3002\u6211\u5011\u5df2\u6709\u4e86\u3192\u7fa4\u76f8\u4f3c\u8a5e\uff0c\u540c\u6642\u6bcf\u500b\u76f8\u4f3c\u8a5e VD 0.643 1 S --&gt; quantity_NP theme_NP Head_VH11 Head VH11 Head VE 0.540 Head Neqa Nad VH VG 0.591 \u4e5f\u6709\u8207\u672a\u77e5\u52d5\u8a5e\u7684\u76f8\u4f3c\u5206\u6578\u3002\u5148\u5c07\u9019\u4e9b\u76f8\u4f3c\u8a5e\u4f9d\u7167\u8a5e\u985e\u5206\u7d44\uff0c\u5f9e\u6bcf\u500b\u8a5e\u985e\u7576\u3197\u53d6</td></tr><tr><td colspan=\"3\">VH \u51fa K \u500b\u76f8\u4f3c\u8a5e\u51fa\u4f86\uff0c\u5c07\u9019\u4e9b\u76f8\u4f3c\u8a5e\u7684\u5206\u6578\u4e88\u4ee5\u5e73\u5747\uff0c\u5f97\u5230\u672a\u77e5\u52d5\u8a5e\u5230\u6bcf\u500b\u8a5e\u985e VH 1.000</td></tr><tr><td colspan=\"3\">VH \u7684\u5e73\u5747\u8ddd\u96e2\uff0c\u672a\u77e5\u52d5\u8a5e\u7684\u8a5e\u985e\u5373\u8207\u5176\u8ddd\u96e2\u6700\u76f8\u8fd1\u7684\u8a5e\u985e\u3002\u6211\u5011\u5c07\u5728\u3198\u3192\u7bc0\u6e2c\u8a66\u8a9e VI 0.736</td></tr><tr><td colspan=\"3\">Word1 \u610f\u76f8\u4f3c\u5ea6\u3197\u7684\u6bd4\u91cd\u3001\u8a9e\u610f\u8207\u8a5e\u985e\u7684\u6bd4\u91cd\u4ee5\u53ca K \u503c\u7684\u5927\u5c0f\u5c0d\u6b63\u78ba\u7387\u7684\u5f71\u97ff\u3002 Word2 VH VHC 0.852 Word3 VH VJ 0.655</td></tr><tr><td>\u5716 1 \u3197\u6587\u53e5\u7d50\u69cb\u6a39\u6a39\u72c0\u5716\u8207\u6b78\u7d0d\u898f\u5247 3. \u5be6\u9a57\u7d50\u679c</td><td/><td/></tr><tr><td>2-3 \u76f8\u4f3c\u8a5e\u7684\u9078\u53d6</td><td/><td/></tr><tr><td colspan=\"3\">\u6bcf\u3192\u500b\u8a5e\u985e\u7684\u5411\u91cf\u7531\u5404\u7236\u7bc0\u9ede\u8207\u5144\u7bc0\u9ede\u51fa\u73fe\u7684\u983b\u7387\u7d44\u6210\uff0c\u5148\u70ba\u653e\u5165\u5404\u7236\u7bc0\u9ede\u7684\u983b \u5728\u4f7f\u7528\u76f8\u4f3c\u6cd5\u4f86\u9810\u6e2c\u52d5\u8a5e\u5206\u985e\u7684\u904e\u7a0b\u3197\uff0c\u3194\u500b\u4e3b\u8981\u7684\u6b65\u9a5f\u3002\u3192\u70ba\u672a\u77e5\u52d5\u8a5e\u7684 \u7387\uff0c\u518d\u4f9d\u6b21\u653e\u5165\u5144\u7bc0\u9ede\u7684\u983b\u7387\uff0c\u82e5\u8a72\u500b\u7bc0\u9ede\u6c92\u51fa\u73fe\u5728\u8a72\u8a5e\u985e\u3197\uff0c\u5247\u653e\u5165\u70ba 0\u3002\u5b9a \u76f8\u4f3c\u8a5e\u7684\u9078\u53d6\uff0c\u3193\u70ba\u6e2c\u91cf\u672a\u77e5\u52d5\u8a5e\u8207\u76f8\u4f3c\u8a5e\u7684\u76f8\u4f3c\u5ea6\uff0c\u3194\u70ba\u6c7a\u5b9a\u672a\u77e5\u52d5\u8a5e\u7684\u8a5e\u985e\u3002 \u7fa9\u5982\u3198\uff1a \u9996\u5148\uff0c\u7576\u3192\u500b\u65b0\u7684\u672a\u77e5\u52d5\u8a5e\u51fa\u73fe\u6642\uff0c\u6211\u5011\u4e26\u4e0d\u77e5\u9053\u54ea\u4e9b\u8a13\u7df4\u8a9e\u6599\u7684\u52d5\u8a5e\u8207\u65b0</td></tr><tr><td colspan=\"3\">Set)\uff0c\u6e2c\u8a66\u8a9e\u6599\u7684\u6b63\u78ba\u7b54\u6848\u70ba\u319f\u5de5\u6a19\u8a18\u7684\u8a5e\u985e\u3002 \u7684\u672a\u77e5\u52d5\u8a5e\u8f03\u76f8\u4f3c\uff0c\u56e0\u6b64\u7406\u8ad6\u3196\u6211\u5011\u5fc5\u9808\u8a08\u7b97\u6bcf\u500b\u8a13\u7df4\u8a9e\u6599\u3197\u7684\u52d5\u8a5e\u8207\u65b0\u7684\u672a\u77e5 i={VA, VAC, VB, VC, VCL,\u2026Na, Nb\u2026.A,\u2026P,\u2026} \u52d5\u8a5e\u7684\u76f8\u4f3c\u5ea6\uff0c\u5c0b\u627e\u51fa\u76f8\u4f3c\u5ea6\u8f03\u9ad8\u7684\u76f8\u4f3c\u8a5e\u4f5c\u70ba\u65b0\u7684\u672a\u77e5\u52d5\u8a5e\u9810\u6e2c\u8a5e\u985e\u7684\u4f9d\u64da\uff0c \u5728\u76f8\u4f3c\u6cd5\u3197\u9700\u8981\u8a0e\u8ad6\u3198\u5217\u3194\u9ede\u3002\u3192\u3001\u8abf\u6574\u8a9e\u610f\u76f8\u4f3c\u5ea6\u3197\u7684\u4e3b\u8981\u7fa9\u539f\u8207\u6b21\u8981\u7fa9</td></tr><tr><td colspan=\"3\">=&lt; \u8a08\u7b97\u65b0\u7684\u672a\u77e5\u52d5\u8a5e\u8207\u8a13\u7df4\u8a9e\u6599\u3197\u52d5\u8a5e\u7684\u5b9a\u7fa9\u5982\u3198\uff1a ),...freq( node t freq(paren ), node t freq(paren Category 2 1 i \u539f\u9593\u7684\u6bd4\u91cd\uff0c\u3193\u3001\u8abf\u6574\u8a9e\u610f\u8207\u8a5e\u985e\u5169\u7a2e\u76f8\u4f3c\u5ea6\u7684\u6bd4\u91cd\uff0c\u3194\u3001\u8abf\u6574 K \u503c\u7684\u5927\u5c0f\uff0c node (sibling freq ), node parent n</td><td>1</td><td>),</td></tr><tr><td>node \u4f7f\u6574\u500b\u7cfb\u7d71\u7684\u6b63\u78ba\u7387\u9054\u5230\u6700\u4f73\u72c0\u614b\u3002 ng freq(sibli ),..., node ng freq(sibli m 2</td><td>)</td><td>&gt;</td></tr><tr><td colspan=\"3\">If Word= wordbase 1 +wordbase 2 +wordbase 3 ...+wordbase n</td></tr><tr><td colspan=\"3\">\u6b63\u78ba\u7387\u7684\u5b9a\u7fa9\u70ba\uff1a Sim(Word unknown ,Word known ) \u5f97\u5230\u5404\u500b\u8a5e\u985e\u7684\u5411\u91cf\u5f8c\uff0c\u6211\u5011\u5229\u7528\u3198\u5217\u516c\u5f0f\u8a08\u7b97\u8a5e\u985e\u8207\u8a5e\u985e\u4e4b\u9593\u7684\u76f8\u4f3c\u7a0b\u5ea6\uff0c\u6240 \u5f97\u7684\u5206\u6578\u4ecb\u65bc 0~1 \u4e4b\u9593\uff0c1 \u8868\u793a\u5b8c\u5168\u76f8\u540c\uff0c0 \u8868\u793a\u5b8c\u5168\u4e0d\u76f8\u540c\u3002 =weight 1 *Sim(wordbase 1-unknown ,wordbase 1-known ) \u6b63\u78ba\u7387\uff1d\u731c\u6e2c\u6b63\u78ba\u7684\u672a\u77e5\u52d5\u8a5e/(1000-\u7121\u6cd5\u731c\u6e2c\u7684\u672a\u77e5\u52d5\u8a5e)</td></tr><tr><td colspan=\"3\">+weight 2 *Sim(wordbase 2-unknown ,wordbase 2-known )</td></tr><tr><td>+...</td><td/><td/></tr><tr><td colspan=\"3\">+weight n *Sim(wordbase n-unknown ,wordbase n-known )</td></tr></table>",
+                "text": "\u6211\u5011\u5217\u51fa\u90e8\u5206 VH \u985e\u7684\u52d5\u8a5e\u8207\u5404\u985e\u52d5\u8a5e\u7684\u76f8\u4f3c\u5ea6\u65bc\u8868\u683c 1\u3002\u9664\u4e86 VH \u985e\u3198\u7684\u5206\u985e VHC \u985e\u5916\uff0cVH \u985e\u52d5\u8a5e\u8207 VI \u985e\u76f8\u4f3c\u7a0b\u5ea6\u6700\u9ad8\uff0cVH \u985e\u8207 VI \u985e\u5169\u8005\u7686\u70ba\u72c0\u614b\u52d5\u8a5e\uff0c \u4ed6\u5011\u7684\u5dee\u5225\u50c5\u5728\u65bc\u53ef\u63a5\u7684\u8ad6\u5143\u6578\u91cf\u3002VI \u985e\u70ba\u985e\u55ae\u8cd3\u52d5\u8a5e\uff0c\u57fa\u672c\u3196\u4e5f\u662f\u4e0d\u53ca\u7269\u52d5 \u8a5e\uff0c\u4f46\u662f VI \u985e\u7684\u52d5\u8a5e\u5728\u8a9e\u610f\u3196\u53ef\u63a5\u53d7\u3192\u500b\u8ad6\u5143\uff0c\u4f46\u8a72\u8ad6\u5143\u7684\u4f4d\u7f6e\u4e0d\u51fa\u73fe\u5728\u52d5\u8a5e \u4e4b\u5f8c\uff0c\u901a\u5e38\u4f7f\u7528\u3192\u500b\u4ecb\u8a5e\u5c07\u8ad6\u5143\u5f15\u4ecb\u51fa\u4f86\u3002\u800c VH \u985e\u8207 VA \u985e\u7684\u76f8\u4f3c\u7a0b\u5ea6\u70ba\u6b21\u9ad8\uff0c",
+                "num": null,
+                "html": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1013.json

@@ -0,0 +1,1062 @@
+{
+    "paper_id": "O01-1013",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:35.597615Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "",
+    "pdf_parse": {
+        "paper_id": "O01-1013",
+        "_pdf_hash": "",
+        "abstract": [],
+        "body_text": [
+            {
+                "text": "B \u01ee & \u02c0 \u0317 \u202b\u0698\u202c \\ \u02aa \u00d2 Oard 1999, Kwok 2001\u00d3 C u \u02fc \u0352 \u02de \u0739 \u0317 \u02da > \u01ed ( \u202b\u06ba\u202c \u0317 \u00a4 \u00b1 & H \u03c3 \u00b1 \u0317 [ \u0204 \\ \u023b \u02de \u0739 Kwok E \u0366 \u01b8 B \u01ee 2 \u00a4 \u00b1 \u00b8 & \u02c0 \u202b\u0698\u202c \\ \u02aa \u02b1 Michael Jordan ( \u0317 \u01e5 \u03c3 \u00b1 \u00e4 \u01c3 0 \u00e5 \u0739 \u0372 \u0424 \u0282 \u202b\u05f0\u202c \u0278 \u0317 & > \u202b\u069c\u202c \u0187 \u00a4 \u00b1 ( \u0317 \u202b\u06ba\u202c \u02c0 Chen \u00d2 1999\u00d3\\\\ \u0584 9 \u00d2 occurrence statistics\u00d3\u018b\u00a4\u00b1\\ \u0317 \u0217 \u0204 \u202b\u071b\u202c H \u03c3 \u00b1 [ \u0204 \\ \u0424 \u0282 \u202b\u05f0\u202c \u0317 \u03bf ] \u02c0 \u0204 \u0278 \u01b8 \u0187 & \u0298 \u03ac \u02c4 \u01f8 \u0317 \u02fc \\ \u03c3 \u00b1 \u00a6 2 c [ \u0204 \\ \u00d2 Knight and Graehl 1997, Chang et al. 2001\u00d3 \u0298 \u02fc H \u0424 \u0282 \u202b\u05f0\u202c \u0204 & \u0317 \u02e4 \u0230 * ) \u042c _ \u02c0 \u0218 \u202b\u06d0\u202c \u202b\u067f\u202c \u0317 \u054b X p 3 \u04ff \u02b1 \u0317 O a ] \u0424 \u0282 \u202b\u05f0\u202c \u0317 h \u202b\u06ba\u202c \\ \u0317 \u00a4 \u00b1 \u01de \u0204 & _ \u0218 \u202b\u06d0\u202c \u0317 \u019b \u02fc \u01ee \u0317 \u0584 \u00a4 \u00b1 \u0334 \u0584 \u00a4 \u00b1 \u0317 \u02fc \u0518 B \u01e1 \u02fc \u00d2 Example-based Approach\u00d3 \u03ac \u01f8 \u0584 \u00a4 \u00b1 \u0372 ) \u042c \u0421 \u00d2 data-driven\u00d3\u0317 \u02fc ) \u042c > \u01ed \u0421 \u202b\u05f0\u202c ( \u202b\u06ba\u202c \u0317 \u00a4 \u00b1 \u02a9 ' \u01ae IBM Watson \u0278 \u0198 \u0317 Brown \u00d2 1988, 1990, 1993\u00d3 \u01b8 \u0317 \u03ac \u01f8 \u0584 \u00a4 \u00b1 \u02fc \u054b \u042c \u202b\u06d0\u202c \u202b\u067f\u202c X p \u02fc \u042c \u02de \u01c3 \u0225 3 a \u0436 & u A \u0317 \u03ac \u01f8 \u202b\u05e4\u202c \u0584 \u00a4 \u00b1 \u02fc \u00d2 Statistical Alignment and Machine Translation\u00d3\u029e \u0225 h \u202b\u06ba\u202c \\ \u0317 \u00a4 \u00b1 \u0424 \u0282 \u202b\u05f0\u202c \u0317 h \\ \u00a4 \u00b1 B & \u03d5 ) \u042c \u0204 \u0218 g \u202b\u06d0\u202c \u202b\u067f\u202c \u0317 \u02fc \u051b \u054b ( a \u01b8 & u A \u0317 \u00a4 \u00b1 \u202b\u05e4\u202c \u0584 9 \u00d2 Alignment Probability\u00d3 \u0317 \u02fc \u0294 \u0225 \u00b8 \u00b8 \u0317 0 \u02e4 A \u0317 \u0334 \u0317 \u01a6 \u202b\u05e4\u202c \u00a4 \u00b1 \u0317 9 2. \u03ac \u01f8 \u0584 \u00a4 \u00b1 \u0334 \u0584 \u00a4 \u00b1 \u01fe \u0372 \u01ae \u00a4 \u00b1 \u01bb \u0317 6 \u0549 \u0317 \u031a \u02fc \u00d2 Direct Approach\u00d3 \u0357 \u029e r \u00a6 2 * ] \u0372 \u01ae \u01cd \u02fc J M \u0663 \u0317 \u00a6 \u01f8 \u0317 \u02fc \u00d2 Transfer Approach\u00d3 1980 \u01f7 \u01b4 \u01e3 \u0278 \u0317 ) \u042c \u01f8 \u0317 \u02fc \u00d2 Empirical Approach\u00d3 \u01ae \u00a4 \u00b1 \u0317 \u0518 B \u02c4 \u01d9 \u0225 \u01e1 \u0584 \u00a4 \u00b1 \u0279 Brown \u01b8 \u0317 \u01e1 \u03ac \u01f8 \u02fc \u054b \u0317 X p Brown \u0317 \u03ac \u01f8 \u0584 \u00a4 \u0334 c ( S \u00b1 ( T \u0317 \u00a4 \u00b1 \u0584 9 \u00d2 Translation Probability\u00d3 Pr( T | S ) \u01c3 \u01ae g 2 \u01ae \u0317 ' \u03d5 \u0584 9 \u02ac (a) \\ \u00a4 \u00b1 \u0584 9 \u00d2 Lexical Translation Probability\u00d3 Pr( S i | T j ) (b) = \u01ed \u0584 9 \u00d2 Fertility Probability\u00d3 Pr( a | b ) (c) 6 \u0202 \u0584 9 \u00d2 Distortion Probability\u00d3 Pr(i | j, k, m ) \u02a9 S i S \u0317 S i \u03d5 T j T \u0317 S j \u03d5 a S i \u0317 T b T j \u0317 T k S \u0317 T m T \u0317 T Brown B \u01ee \u01bb \u00b1 \u0297 \u0317 \\ \u02fc \u01d9 \u0225 \u0436 \u00a5 \u0317 \u00e4 \u03d0 \u022d \u03ac \u00e5 \u00e4 B \u00e5 q \u04a5 \u02fc \u00d2 Expectation and Maximization Algorithm\u00d3 \u01c3 \u01ae \u02b1 ' \u03d5 \u0669 * \u0317 \u0584 9 \u02ac \u0317 \u03ac \u022d \u03ac \u03d0 \u02a9 \u00e4 B \u00e5 \u0317 \u0259 \u00ba E \u0372 1 \u0317 \u0584 9 \u02ac \u022d \u03ac \u03d0 \u025b \u02b9 \u01c3 \u0317 \u00a4 \u00b1 \u202b\u05e4\u202c \u0218 \u00e4 \u03d0 \u022d \u03ac \u00e5 \u0317 \u0259 \u00ba E \u0372 \u01ae F \u0204 \u0317 \u202b\u0698\u202c \u01e1 \u0317 B \u0317 \u00a4 \u00b1 \u202b\u05e4\u202c \u022d \u03ac ' \u03d5 \u0584 9 \u02ac \u03d0 \u0436 EM q \u04a5 \u02fc \u03ac \u01f8 \u0584 \u00a4 \u00b1 \u0334 \u0317 \u00a4 \u00b1 \u0584 9 \u02ac \u0317 \u022d \u03ac \u03d0 \u01c3 H 2 \u202b\u05ef\u202c \u202b\u0697\u202c \u042e \u0334 \u00d2 Noisy Channel Model\u00d3 0 \u00a4 \u00b1 \u0584 9 \u02ac 1 \u0317 N-gram \u0282 \u0334 \u00d2 Language Model\u00d3 \u01c3 \u01ae \u01ee D q \u04a5 \u02fc _ D \u02fc \u00d2 Beam Search\u00d3\u025bB\u05849\u03d0\u0317\u019b\u01f8 > \u01ed \u00a4 \u00b1 3. \u01ee H \u01a6 \u202b\u05e4\u202c \u00a4 \u00b1 \u0317 \u03ac \u01f8 \u0334 Brown c \u0334 \u0317 6 \u0202 \u0584 9 \u0372 M H \u025a & \u0317 \u00a4 \u00b1 1 6 \u02a9 \u01b2 \u202b\u06ba\u202c \u0317 % \u0597 \u0297 \u0317 % X & \u03d5 \u00a4 \u00b1 \u202b\u05e4\u202c \u00d2 alignment\u00d3 \u019b \u01f8 \u0317 \u0584 9 6 \u019b _ \u0218 \u0282 \u0372 F \u0204 \u0317 \u202b\u05e4\u202c \u0317 6 2 \u0317 6 \u0202 \u0584 9 \u03d0 \u0317 ` \u05a8 y \u025a & \u0317 \u00a4 \u00b1 1 6 \u02a9 \u01b2 \u0317 \u00a4 \u00b1 6 \u0204 T \u0317 \u202b\u06ba\u202c \u02c0 \u02e4 S i' , i' \u2260 i \u202b\u05e4\u202c \u02b1 T j S i \u202b\u05e4\u202c \u02b1 1 6 j \u0317 \u0584 9 I * 1 Pr(j | i, k, m) \u2245 1 \u03a1 Pr(j | i', k, m) = 0, i' \u2260 i 3 \u0597 % \u0317 \u0584 9 I * \u0317 \u0301 \u0372 \u0436 \u0235 \u0317 \u022d \u03ac 9 \u0372 \u0352 \u01c3 \u0317 \u00a4 \u00b1 \u202b\u05e4\u202c \u019b \u01f8 \u02a9 \u0584 9 \u03d0 \u0372 & 6 6 \u0202 \u0584 9 \u0317 ` \u05a8 3 H \u0317 \u0317 \u03d0 B \u202b\u05f0\u202c ' \\ ( ( \u0317 \u01a6 \u01e1 \u01c3 \u00a4 \u00b1 \u202b\u05e4\u202c A* \u0317 ' \u03d5 S 1 S 2 S 3 \u00a4 \u00b1 1 6 \u023b \u0372 S 1 \u2192 {T 1 , T 2 } S 2 \u2192 {T 3 , T 4 } S 3 \u2192 {T 5 } E \u0372 A* = (0, 12, 34, 5)\u00d2S&\u03d5 0 \u01b4 F \u0204 \u0317 ( \u0204 \u202b\u05e4\u202c \u0264 \u0204 ( \u02fc \u202b\u05e4\u202c \u02b1 \\ ( \u0317 \u0301 \u00d3 k = 3 m = 5 \u0317 \u01a6 \u01e1 \u00a4 \u00b1 \u202b\u05e4\u202c A*\u0317 \u0301 Z 7 35% \u031a \u022d \u03ac A*\u0317 \u01c3 \u022d \u03ac \u03d0 \u00d2 Maximum Likelihood Estimation\u00d3 \u02b1 6 Pr MLE ( A* ) = 0.35 \u0218 \u0584 9 \u0597 \u0317 %",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Pr( A* ) = P(1|1,3,5) P(2|1,3,5) P(3|2,3,5) P(4|2,3,5) P(5|3,3,5)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "9 B \u01ae \u042c \u0317 6 \u0202 \u0584 9 \u03d0 \u00d2 0.6\u00d3\u022d\u03ac P(j | i,3,5) \u02a9 ` \u05a8 \u0436 \u0235 \u0235 H \u0317 \u022d \u03ac \u03d0",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Pr( A* ) < (0.6) 5 = 0.046656 << 0.35 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Pr( A* ) << Pr MLE ( A* ) \u0187 \u04a8 \u0317 \u022d \u03ac \u00a4 \u00b1 1 6 \u0317 \u0584 9 a \u01b8 \u0187 \u031a \u022d \u03ac \u00b8 \u00a4 \u00b1 6 \u0317 \u02fc 3 \u02fc = \u01ed \u0584 9 6 \u0202 \u0584 9 \u02a1 2 \u00a4 \u00b1 \u202b\u05e4\u202c \u0584 9 \u00d2 Alignment Probability\u00d3 3 \u0597 \u0217 \u03d5 \u023b \u0317 \u0317 6 \u00a4 \u00b1 1 6 = \u01ed \u0317 \u0218 \u0372 \u01ae \u00a4 \u00b1 \u202b\u05e4\u202c \u029e & \u02a1 \u0217 \u0317 + \u02fc a \\ c ( S \u00b1 ( T \u0317 \u00a4 \u00b1 \u0584 9 Pr\u00d2T | S\u00d3 g 2 \u01ae \u0317 C \u03d5 \u0584 9 \u02ac (a) \\ \u00a4 \u00b1 \u0584 9 \u00d2 Lexical Translation Probability\u00d3 Pr( T(A i ) | S i ) (b) \u00a4 \u00b1 \u202b\u05e4\u202c \u0584 9 \u00d2 Alignment Probability\u00d3 Pr ( A | k, m ) = Pr ( A 0 , A 1 , A 2 , \u2026, A k | k, m ) \u02a9 S i S \u0317 S i \u03d5 T(A i ) T \u202b\u05e4\u202c \u02b1 S i \u0317 A 0 T \u0264 \u0204 \u202b\u05e4\u202c \u02b1 S \u0317 \u0317 \u0402 A i T \u202b\u05e4\u202c \u02b1 S i \u0317 \u0317 \u0402 , i >0 k S \u0317 T m T \u0317 T \u02e4 \u03d5 \u023b S i \u0317 \u202b\u05e4\u202c T a \u01c3 \u01ae \u01ee \u202b\u05e4\u202c 1 6 7 \u0317 B(A, i) ` E(A, i) \u029e \u0669 \u202b\u05e4\u202c \u202b\u06ba\u202c \u0317 \u0430 E \u0372 S i \u2192 T(A i ) = {T B(A,i) , T B(A,i)+1 , T E(A,i) } A i = { B(A,i), B(A,i)+1, \u2026 , E(A,i) } 4. \u00b8 a \u0225 \u0187 & \u0279 \u0317 \u00b8 \u01ae \u00b8 a \u01b8 \u0317 A \u0317 \u01a6 \u00a4 \u00b1 \u0334 \u0317 \u02e4 \u01c3 \u0225 \u02c0 \u0436 \u00b8 a + \u0187 g A \u0334 \u0204 \u202b\u06ba\u202c \u0317 I \u03d5 > \u202b\u069c\u202c 1. \u01ae \u00a4 \u00b1 \u202b\u05e4\u202c \u0584 9 \u01b4 = \u01ed \u0584 9 6 \u0202 \u0584 9 \u0372 \u0245 \u01c3 \u01ae \u02b1 \u042c \u01e5 \u0317 \u202b\u05e4\u202c J 2. \u00a4 \u00b1 \u202b\u05e4\u202c \u0372 \u0245 I u \u01f8 \u0218 \u0372 \u03d5 \u023b \u202b\u05e4\u202c 1 6 \u0317 ^ \u00a4 \u00b1 \u202b\u05e4\u202c \u0584 9 \u0317 8 \u0436 \u0398 \u022d \u03ac \u0317 T \u0436 n 3. \u00a4 \u00b1 \u202b\u05e4\u202c \u0584 9 \u0317 8 \u01e1 \u042c \u02a9 \u0584 9 \u03d0 \u0317 \u03ac \u01c3 T \u0436 \u0235 4. \u0357 \u0317 \u0584 \u00a4 \u00b1 \u0334 \u202b\u05e4\u202c \u01ee \u02b1 \u0424 \u0282 \u202b\u05f0\u202c \u0317 \u01c3 \u0225 \u02c0 \u02c4 \u0235 4.1 \u00b8 \u0317 \u03ac \u0584 9 \u03d0 \u0317 \u01f1 H \u02da \u02b9 \u202b\u0698\u202c \u01a6 \u0317 a \u01ee BDC \u047b \\ \u02aa \u00d2 BDC 1992\u00d3 \u0317 \u01a6 1 \u0232 \u00b8 \u0317 c \u0187 \u00b8 \u0317 1 \u0294 \u0669 > \u202b\u069c\u202c a \u03c8 \u01c2 \\ ( H 3 \u03d5 \\ \u0317 1 \u0230 ( T _ - \u01d2 a \u01c2 ( \u0317 \u01d0 2 1 \u0298 1 \u0317 \u00a4 \u00b1 \u0372 _ \u00a4 \u00b1 \u01c2 \u0357 \u01c3 \u01ae _ \u0235 \u0421 \u0317 \u202b\u0697\u202c c \u0421 \u0436 \u0357 a \u02b1 96,156 \u01c3 \u01ee \u0317 \\ \u01a6 \u00a4 \u00b1 \u0317 a \u01ae \u00d2 P n , Q n \u00d3 , n = 1, N \u029e \u01b4 ^ \u0412 \u00b8 a \u01ae EM q \u04a5 \u02fc \u029e \u02b1 S ' F \u01b8 \u0317 \u00aa \u00a4 \u00b1 \u0584 9 \u00a4 \u00b1 \u202b\u05e4\u202c \u0584 9 a \u02b9 \u0187 & \u0230 \u00ac 7 Och \u0189 \u00d2 2000\u00d3 8 H IBM \u0584 9 \u0334 \u0317 \u00b8 \u0317 \u02fc \u02a9 1 \u0317 \u0372 \u01bb \u0584 9 \u0317 \u022d \u03ac 1. \u0317 \u03d6 a \u02b9 Brown \u0334 c \u0204 \u0317 6 \u0202 \u0584 9 EM q \u04a5 \u02fc \u0317 S \u0188 \u0357 B \u01ee A \u0334 \u0317 \u00a4 \u00b1 \u202b\u05e4\u202c \u0584 9 2. a % \\ \u01a6 \u00a4 \u00b1 \\ ( ( \u0317 & \u0398 \u0317 \u0584 \u042c F \u01ae S & j \u04a5 \u0584 9 \u0334 \u0317 6 \u0202 \u0584 9 \u01ee & \u01ee \u0317 \u01d9 \u01d8 Pr(j | i, k, m ) = 1/m \u0218 \u01ee \u01e6 \u0317 \u03ac \u02fc \u01b5 6 \u0202 \u0584 9 \u0317 \u03d0 k i m j m k j i 5 . 0 5 . 0 1 ) , , | ( \u2212 \u2212 \u2212 \u2212 = 3 I [1] \u02a9 i = \\ ( 6 k = \\ ( \u202b\u0626\u202c j = ( 6 m = ( \u202b\u0626\u202c S T i k j M Pr( j | i,k,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "6 \u0202 \u0584 9 \u0317 \u01e6 \u03ac H \u025a & \u202b\u0698\u202c \u01a6 a % \u025a \u03d5 \\ ( \u01c3 \u01ae \u00a4 \u00b1 2 \u02a9 \u022c & \u03d5 ( \u0230 \u0372 \u02a9 \u0584 9 6 \u0218 B B X & \u0372 2 \u03d5 \\ ( \u00a4 \u00b1 2 4 \u03d5 ( a \u01c3 \u01ae \u02b1 8 \u03d5 \\ ( \u0317 ( \u025a & \u03d5 \u0317 6 \u0202 \u0584 9 \u01f8 1 \u0317 Pr( j | i,k,m)\u03d02\u01e5) B \u202b\u0698\u202c \u01a6 \u00d2 flight eight, \u03c6 \u0225 \u00d3 a \u01ee \u01f8 1 \u01c3 \u01ae \u03ac \u04a5 \u02b1 1 \u0317 ( \\ \u0317 6 \u0202 \u0584 9 \u0204 \u0187 ( \u0317 6 \u0202 \u0584 9 \u0357 a E \u01c3 3 \u022d \u03ac \u01a6 \u0317 \u022c \\ ( E ( C \u0317 \u00a4 \u00b1 \u0584 9",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Pr ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "i P E m k i j j Q C i P E E C 1 1 1 1 1 1 ) , , | Pr( )) ( , ( ) , , | Pr( )) ( , ( )) ( , ( ) | Pr( \u03b4 \u03b4 \u03b4 [2] \u02a9 P n (i) P n S i Q n (j) Q n S j k = |P n | m = |Q n | \u03b4(x, y) = 1 \u03a1 x = y, \u03b4(x, y) = 0 \u03a1 x \u2260 y S i T j i k j m Pr(j|i,k,m) Pr(T j | S i ) Pr(T j | S i ) Pr(j|i,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "6 \u0202 \u0584 9 \\ \u00a4 \u00b1 \u0584 9 \u0317 \u022d \u03ac \u03d0 \u01f8 2 \u0317 \u01ee ( H \u01bb \u202b\u0626\u202c E C \u0317 F \u0204 \u01a6 \u0317 \u0584 9 \u03d0 \u0294 \u01ae E F \u0204 ( \u0317 \u0584 9 \u03d0 \u0317 \u202b\u0626\u202c B Pr (C | E)\u0317\u05849\u03d0 H 0 1 \u029d \u01f8 2 F \u02b1 \u0317 \u0584 9 \u03d0 a \u01c3 \u01ae \u022d \u03ac \u022c \u01a6 ( \u0317 \u0317 \u0584 9 \u03d0 2 \u01b8 1 \u0317 ( \u0317 \\ \u00a4 \u00b1 \u0584 9 10 4.2 EM q \u04a5 \u02fc \u0317 S & \u03ac \u04a5 S & 3 \u0317 \u202b\u05e4\u202c B \u0204 \u0187 \u0317 \u0584 9 \u02ac \u022d \u03ac \u03d0 a E \u01c3 \u01ae \u0225 EM q \u04a5 \u02fc \u0317 B \u0259 \u00ba a \u02b9 \u0669 * \u0317 R \u02fc \u00d2 Greedy Method\u00d3 \u029e \u025b \u02b9 \u025a & ^ \u202b\u0698\u202c \u01a6 \u00d2 P n , Q n \u00d3 \u0317 B \u202b\u05e4\u202c a % \u0669 * \u0317 = \u01ed \u0334 & \u03d5 \\ ( \u01c3 \u01ae \u202b\u05e4\u202c \u02b1 0 \u02b1 \u03d5 ( \u0218 \u025a \u03d5 ( - \u202b\u05e4\u202c \u02b1 & \u03d5 \\ ( \u0204 \u0187 \u01a6 \u0317 \\ \u00a4 \u00b1 6 \u0202 \u0584 9 \u0317 \u022d \u03ac \u03d0 \u02a9 ` \u05a8 \u00d2 2\u00d3 a E \u01c3 \u01ae 3 \u02b9 \u0584 9 \u03d0 > \u01ed \\ ( ( \u0317 \u0294 % \u0317 = \u01ed \u0334 \u02a9 \u01b2 \u0317 \\ ( 3 ( \u0317 \u00a5 \u0317 \u0225 ^ \u0259 \u00ba \u031a \u02b1 \u0264 \u0204 \u0317 ( \u02c4 \u0584 9 \u03d0 \u0235 H X & \u03d5 \u0650 \u03d0 (threshold) \u01a1 \u03a1 \u0204 \u0317 ( E \u0264 \u0204 \u202b\u05e4\u202c \u02b1 \\ ( \u0235 \u202b\u05e4\u202c \u0317 \u0584 9 \u0650 \u03d0 \u01c3 \u01ae 8 \u202b\u05da\u202c T \u0235 \u0317 \u202b\u05e4\u202c \u0204 \u0241 H \u018b 0 1 0 \u0317 = \u01ed \u01f8 \u0436 y G \u202b\u0758\u202c \u0357 \u01ae 0.008 \u0650 \u03d0 \u01c3 \u01c2 \u0235 \u202b\u05da\u202c T \u0317 \u02b1 \"flight eight\" \u0317 B \u01f1 2 \u0317 \u0584 9 \u03d0 a \u01c3 \u02b1 3 \u0317 \u202b\u05e4\u202c \u019b \u01f8 \u00d2 0, 34, 12\u00d3 S i T i i j k m Pr(j|i,k,m) Pr(T j | S i ) Pr(T j | S i ) Pr(j|i,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u01a6 \u202b\u05e4\u202c \u0584 9 \u03d0 5 \u0317 y B S T T(A 0 ) S 1 T(A 1 ) S 2 T(A 2 ) T-shaped antenna \u0136 \u0523 T-shaped \u0136 antenna \u0523 X-ray",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "S T A 0 A 1 A 2 T(A 0 ) T(A 1 ) T(",
+                        "eq_num": "A"
+                    }
+                ],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u0584 9 \u03d0 5 S \u0188 j \u04a5 a \u025a & \u202b\u0698\u202c \u01a6 \u00d2 S, T\u00d3 \u029d \u02a9 \\ ( ( \u0217 T \u0317 F \u0204 \u0317 \u202b\u05e4\u202c \u019b \u01f8 A \u03ac \u04a5 \u02a9 \u00a4 \u00b1 \u0584 9 Pr(T | S, A) H X & \u202b\u05e4\u202c \u019b \u01f8 A Pr(T | S, A) A \u0317 \u0584 9 \u01f1 A F < \u0317 \\ (S i , T(A i ))\u0317\u05849`\u05a8 \u220f = = = k i i i A A S A T m k A A S T S T 1 ) | ) ( Pr( ) , | Pr( max ) , | Pr( max ) | Pr( 3 \u01c3 \u0317 \u202b\u05e4\u202c A* \u01c3 \u01f1 \u01f8 < \u220f = = = k i i i A A S A T m k A A S T A 1 * ) | ) ( Pr( ) , |",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u03c6 \u0225 \u00d3 \u0317 I \u03d5 \u00a4 \u00b1 \u0584 9 \u03d0 \u0317 \u202b\u05e4\u202c \u019b \u01f8 7 \u0317 \u03d0 \u0739 S \u0188 \u0317 \u03ac \u022d \u03ac \u03d0 \u0317 2 \u202b\u05ef\u202c \u01c3 \u01ae \u01b8 \u01e5 \u0317 \u202b\u05e4\u202c J A* = (0, 34, 12) 5. \u00b8 0 \u02e4 [ \u022d a \u0225 \u0317 \u00b8 \u02de \u0187 A \u0317 \u03ac \u01f8 \u01a6 \u00a4 \u00b1 \u0334 \u01c3 \u0225 > \u01ed \u01e5 \u0317 \u202b\u05e4\u202c J A \u0334 \u018b \u0317 \u00a4 \u00b1 \u202b\u05e4\u202c \u0584 9 \u0317 8 \u0436 T \u0317 \u05bd 3 10 \u0317 \u0421 E \u01c3 \u01ae \u022d \u03ac \u01b8 \u01c3 \u0317 \u0584 9 \u03d0 \u01f1 H A \u0317 \u0334 8 \u0187 \u0584 9 \u03d0 \u0317 ` \u05a8 EM q \u04a5 \u02fc \u0317 \u0317 \u042c \u0584 9 \u02ac \u0317 2 \u202b\u05ef\u202c T ) \u042c S & 0 \u02e4 S \u0188 0 \u02e4 S T T(A 0 ) T(A 1 ) T(A 2 ) T(A 0 ) T(A 1 ) T(A 2 ) association football \u01f8 \u0289 \u01f8 \u0289 \u01f8 \u0289 delay flip-flop \u012c \u0334 \u01e5 \u012c \u0334 \u01e5 \u012c \u0334 \u01e5 I demodulator \u0131 \u0402 g \u0549 \u0131 \u0402 g \u0549 \u0131 \u0402 g \u0549 Disgraceful",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "= = = k i i i A T A T T T S A T m k A T A S T T S T T 1 * ) Pr( ) | ) ( Pr( ) , | Pr( max max arg ) Pr( ) , | Pr( max max arg ) Pr( ) | Pr( max arg \u02a9 Pr(T) = \u00a4 \u00b1 T \u0317 \u0282 \u0334 \u0584 9 k = S \u0317 m = T \u0317 Pr(A | k, m)\u0317\u0334\u0669 \u0187 Pr(T | S)\u0317\u03ac\u04a5 B a \u02da \u01ae _ D q \u04a5 \u02fc \u00d2 Branch and Bound Algorithm\u00d3 D \u0584 9 \u0317 T*\u03d0 a \u01c3 \u01ae \u01f1 \u0584 9 \u03d0 \u0317 Pr(A | k, *) Pr(* | S i )^ \u0317 g \u00d2 solution\u00d3 D T D \u0317 _ \u00d2 upper bound\u00d3 \u023d \u01ee Pr(* | S i ) Pr(T)\u0317 N-gram \u0334 \u0317 \u0584 9 \u03d0 \u0430 \u02b1 _ D D \u0518 \u0317 \u02e4 \u0479 B g \u0317 \u0301 D \u0317",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "BDC 1992 The BDC Chinese-English electronic dictionary (version 2.0), Behavior Design Corporation",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "BDC 1992 The BDC Chinese-English electronic dictionary (version 2.0), Behavior Design Corporation, Taiwan.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "A Statistical Approach to Language Translation",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Brown",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Cocke",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Della Pietra",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Della",
+                        "middle": [],
+                        "last": "Pietra",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "V",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Jelinek",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Mercer",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Roosin",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1988,
+                "venue": "Proceedings of the 12th International Conference on Computational Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "71--76",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Brown, P. F., Cocke J., Della Pietra S. A., Della Pietra V. J., Jelinek F., Mercer R. L., and Roosin P. S. 1988 A Statistical Approach to Language Translation, In Proceedings of the 12th International Conference on Computational Linguistics, Budapest, Hungary, pp. 71-76.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "A Statistical Approach to Machine Translation",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Brown",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Cocke",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Della Pietra",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Della",
+                        "middle": [],
+                        "last": "Pietra",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "V",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Jelinek",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Lafferty",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Mercer",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Roosin",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Computational Linguistics",
+                "volume": "16",
+                "issue": "2",
+                "pages": "79--85",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Brown, P. F., Cocke J., Della Pietra S. A., Della Pietra V. J., Jelinek F., Lafferty J. D., Mercer R. L., and Roosin P. S. 1990 A Statistical Approach to Machine Translation, Computational Linguistics, 16/2, pp. 79-85.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "The Mathematics of Statistical Machine Translation: Parameter Estimation, Computational Linguistics",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Brown",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Della Pietra",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Della",
+                        "middle": [],
+                        "last": "Pietra",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "V",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Mercer",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "",
+                "volume": "19",
+                "issue": "",
+                "pages": "263--311",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Brown, P. F., Della Pietra S. A., Della Pietra V. J., and Mercer R. L. 1993 The Mathematics of Statistical Machine Translation: Parameter Estimation, Computational Linguistics, 19/2, pp. 263-311.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Nathu IR System at NTCIR-2",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2001,
+                "venue": "Proceedings of the Second NTCIR Workshop Meeting on Evaluation of Chinese and Japanese Text Retrieval and Text Summarization",
+                "volume": "",
+                "issue": "",
+                "pages": "49--52",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chang, J. S. et al. 2001. Nathu IR System at NTCIR-2. In Proceedings of the Second NTCIR Workshop Meeting on Evaluation of Chinese and Japanese Text Retrieval and Text Summarization, pp. (5) 49-52, National Institute of Informatics, Japan.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Taxonomy and Lexical Semantics -From the Perspective of Machine Readable Dictionary",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Chang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Ker",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Chen",
+                        "middle": [
+                            "M H"
+                        ],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the third Conference of the Association for Machine Translation in the Americas (AMTA)",
+                "volume": "",
+                "issue": "",
+                "pages": "199--212",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chang, J. S., Ker S. J., and Chen M. H. 1998 Taxonomy and Lexical Semantics -From the Perspective of Machine Readable Dictionary, In Proceedings of the third Conference of the Association for Machine Translation in the Americas (AMTA), pp. 199-212.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Resolving Translation Ambiguity and Target Polysemy in Cross-Language Information Retrieval",
+                "authors": [
+                    {
+                        "first": "H",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Bian",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Lin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of the 37 th Annual Meeting of the Association for Computation Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "215--222",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, H.H., G.W. Bian and W.C. Lin. 1999. Resolving Translation Ambiguity and Target Polysemy in Cross-Language Information Retrieval. In Proceedings of the 37 th Annual Meeting of the Association for Computation Linguistics, pp 215-222.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Robust Bilingual Word Alignment or Machine Aided Translation",
+                "authors": [
+                    {
+                        "first": "I",
+                        "middle": [],
+                        "last": "Dagan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Church",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Gale",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Proceedings of the Workshop on Very Large Corpora Academic and Industrial Perspectives",
+                "volume": "",
+                "issue": "",
+                "pages": "1--8",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Dagan, I., Church K. W. and Gale W. A. 1993 Robust Bilingual Word Alignment or Machine Aided Translation, In Proceedings of the Workshop on Very Large Corpora Academic and Industrial Perspectives, pp. 1-8.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Aligning Noisy Parallel Corpora across Language Groups: Word Pair Feature Matching by Dynamic Time Warping",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Fung",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Mckeown",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "Proceedings of the First Conference of the Association for Machine Translation in the Americas (AMTA)",
+                "volume": "",
+                "issue": "",
+                "pages": "81--88",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Fung, P. and McKeown K. 1994 Aligning Noisy Parallel Corpora across Language Groups: Word Pair Feature Matching by Dynamic Time Warping, In Proceedings of the First Conference of the Association for Machine Translation in the Americas (AMTA), pp. 81-88, Columbia, Maryland, USA.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Identifying Word Correspondences in Parallel Texts",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Gale",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Church",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "Proceedings of the Fourth DARPA Speech and Natural Language Workshop",
+                "volume": "",
+                "issue": "",
+                "pages": "152--157",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Gale, W. A. and Church K. W. 1991 Identifying Word Correspondences in Parallel Texts, In Proceedings of the Fourth DARPA Speech and Natural Language Workshop, pp. 152-157.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Phrase Discovery for English and Cross-Language Retrieval at TREC-6",
+                "authors": [
+                    {
+                        "first": "F C",
+                        "middle": [],
+                        "last": "Gey",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the 6 th Text Retrieval Evaluation Conference",
+                "volume": "",
+                "issue": "",
+                "pages": "637--648",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Gey, F C and A. Chen. 1997. Phrase Discovery for English and Cross-Language Retrieval at TREC-6. In Proceedings of the 6 th Text Retrieval Evaluation Conference, pp 637-648.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "Machine Translation at the TAUM Group",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Isabelle",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1987,
+                "venue": "Machine Translation Today: The State of the Art, Proceedings of the Third Lugano Tutorial",
+                "volume": "",
+                "issue": "",
+                "pages": "247--277",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Isabelle, P. 1987 Machine Translation at the TAUM Group, In M. King, editor, Machine Translation Today: The State of the Art, Proceedings of the Third Lugano Tutorial, pp. 247-277.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Proceedings of the Second NTCIR Workshop Meeting on Evaluation of Chinese and Japanese Text Retrieval and Text Summarization",
+                "authors": [
+                    {
+                        "first": "Noriko",
+                        "middle": [],
+                        "last": "Kando",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Kenro",
+                        "middle": [],
+                        "last": "Aihara",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Koji",
+                        "middle": [],
+                        "last": "Eguchi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Hiroyuki",
+                        "middle": [],
+                        "last": "Kato",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2001,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Kando, Noriko, Kenro Aihara, Koji Eguchi and Hiroyuki Kato. 2001. Proceedings of the Second NTCIR Workshop Meeting on Evaluation of Chinese and Japanese Text Retrieval and Text Summarization, National Institute of Informatics, Japan.",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "Text-Translation Alignment",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Kay",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "R\u00f6scheisen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1988,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Kay, M. and R\u00f6scheisen M. 1988 Text-Translation Alignment, Technical Report P90-00143, Xerox Palo Alto Research Center.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "A Class-base Approach to Word Alignment",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Ker",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Computational Linguistics",
+                "volume": "23",
+                "issue": "2",
+                "pages": "313--343",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ker, S. J. and Chang J. S. 1997 A Class-base Approach to Word Alignment, Computational Linguistics, 23/2, pp. 313-343.",
+                "links": null
+            },
+            "BIBREF16": {
+                "ref_id": "b16",
+                "title": "Machine Transliteration",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Knight",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the 35 th Annual Meeting of the Association for Computational Linguistics and 8 th Conference of ACL European Chapter",
+                "volume": "",
+                "issue": "",
+                "pages": "128--135",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Knight, K. and J Graehl. 1997. Machine Transliteration, In Proceedings of the 35 th Annual Meeting of the Association for Computational Linguistics and 8 th Conference of ACL European Chapter, pp. 128-135.",
+                "links": null
+            },
+            "BIBREF17": {
+                "ref_id": "b17",
+                "title": "An Algorithm for finding noun phrase correspondence in bilingual corpus",
+                "authors": [
+                    {
+                        "first": "Julian",
+                        "middle": [],
+                        "last": "Kupiec",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "ACL 31",
+                "volume": "23",
+                "issue": "",
+                "pages": "17--22",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Kupiec, Julian. 1993 An Algorithm for finding noun phrase correspondence in bilingual corpus, In ACL 31, 23/2, pp. 17-22.",
+                "links": null
+            },
+            "BIBREF18": {
+                "ref_id": "b18",
+                "title": "NTCIR-2 Chinese, Cross-Language Retrieval Experiments Using PIRCS",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Kwok",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2001,
+                "venue": "Proceedings of the Second NTCIR Workshop Meeting on Evaluation of Chinese and Japanese Text Retrieval and Text Summarization",
+                "volume": "",
+                "issue": "",
+                "pages": "14--20",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Kwok, K L. 2001. NTCIR-2 Chinese, Cross-Language Retrieval Experiments Using PIRCS. In Proceedings of the Second NTCIR Workshop Meeting on Evaluation of Chinese and Japanese Text Retrieval and Text Summarization, pp. (5) 14-20, National Institute of Informatics, Japan.",
+                "links": null
+            },
+            "BIBREF19": {
+                "ref_id": "b19",
+                "title": "Should we Translate the Documents or the Queries in Cross-Language Information Retrieval",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Mccarley",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Scott",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of the 37 th Annual Meeting of the Association for Computation Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "208--214",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "McCarley, J. Scott. 1999. Should we Translate the Documents or the Queries in Cross-Language Information Retrieval? In Proceedings of the 37 th Annual Meeting of the Association for Computation Linguistics, pp 208-214.",
+                "links": null
+            },
+            "BIBREF20": {
+                "ref_id": "b20",
+                "title": "Automatic Construction of Clean Broad-Coverage Translation Lexicons",
+                "authors": [
+                    {
+                        "first": "I",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Melamed",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Proceedings of the second Conference of the Association for Machine Translation in the Americas (AMTA)",
+                "volume": "",
+                "issue": "",
+                "pages": "125--134",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Melamed, I. D. 1996 Automatic Construction of Clean Broad-Coverage Translation Lexicons, In Proceedings of the second Conference of the Association for Machine Translation in the Americas (AMTA), pp. 125-134.",
+                "links": null
+            },
+            "BIBREF21": {
+                "ref_id": "b21",
+                "title": "Machine Translation: How Far Can it Go?",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Nagao",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1986,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Nagao, M. 1986 Machine Translation: How Far Can it Go? Oxford University Press, Oxford. 25",
+                "links": null
+            },
+            "BIBREF22": {
+                "ref_id": "b22",
+                "title": "Effect of Term Segmentation on",
+                "authors": [
+                    {
+                        "first": "D W",
+                        "middle": [],
+                        "last": "Oard",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Wang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Oard, D W and J. Wang. 1999. Effect of Term Segmentation on",
+                "links": null
+            },
+            "BIBREF23": {
+                "ref_id": "b23",
+                "title": "Cross-Language Information Retrieval",
+                "authors": [
+                    {
+                        "first": "/",
+                        "middle": [],
+                        "last": "Chinese",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "English",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "Proceedings of the Symposium on String and Processing and Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chinese/English Cross-Language Information Retrieval. In Proceedings of the Symposium on String and Processing and Information Retrieval. http://www.glue.umd.edu/~oard/research.html.",
+                "links": null
+            },
+            "BIBREF24": {
+                "ref_id": "b24",
+                "title": "Improved Statistical Alignment Models",
+                "authors": [
+                    {
+                        "first": "Franz",
+                        "middle": [
+                            "Josef"
+                        ],
+                        "last": "Och",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Hermann",
+                        "middle": [],
+                        "last": "Ney",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Proceedings of the 38 th Annual Meeting of the Association for Computation Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Och, Franz Josef and Hermann Ney. 2000. Improved Statistical Alignment Models. In Proceedings of the 38 th Annual Meeting of the Association for Computation Linguistics.",
+                "links": null
+            },
+            "BIBREF25": {
+                "ref_id": "b25",
+                "title": "The Effect of Query Structure and Dictionary Setups in Dictionary-based Cross-Language Retrieval",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Pirkola",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the 21 st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "55--63",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Pirkola, A. 1998. The Effect of Query Structure and Dictionary Setups in Dictionary-based Cross-Language Retrieval. In Proceedings of the 21 st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, pp 55-63.",
+                "links": null
+            },
+            "BIBREF26": {
+                "ref_id": "b26",
+                "title": "Translating Collocations for Bilingual Lexicons: A Statistical Approach, Computational Linguistics",
+                "authors": [
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Smadja",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Mckeown",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "V",
+                        "middle": [],
+                        "last": "Hatzivassiloglou",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "",
+                "volume": "22",
+                "issue": "",
+                "pages": "1--38",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Smadja, F., McKeown K., and Hatzivassiloglou V. 1996 Translating Collocations for Bilingual Lexicons: A Statistical Approach, Computational Linguistics, 22/1, pp. 1-38.",
+                "links": null
+            },
+            "BIBREF28": {
+                "ref_id": "b28",
+                "title": "Bilingual Text Matching Using Bilingual Dictionary and Statistics",
+                "authors": [],
+                "year": null,
+                "venue": "Proceedings of the 15th International Conference on Computational Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "1076--1082",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Bilingual Text Matching Using Bilingual Dictionary and Statistics, In Proceedings of the 15th International Conference on Computational Linguistics, pp. 1076-1082.",
+                "links": null
+            },
+            "BIBREF29": {
+                "ref_id": "b29",
+                "title": "Learning an English-Chinese Lexicon from a Parallel Corpus",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Wu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "X",
+                        "middle": [],
+                        "last": "Xia",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "Proceedings of the first Conference of the Association for Machine Translation in the Americas (AMTA)",
+                "volume": "",
+                "issue": "",
+                "pages": "206--213",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Wu, D. and Xia X. 1994 Learning an English-Chinese Lexicon from a Parallel Corpus, In Proceedings of the first Conference of the Association for Machine Translation in the Americas (AMTA), pp. 206-213.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF5": {
+                "text": "| S,A) = Pr(0,12,34 | 2,4) Pr( |flight) Pr(\u03c6\u0225|eight) A = (0, 34, 12) Pr(T | S,A) = Pr(0,34,12 | 2,4) Pr(\u03c6\u0225 | flight) Pr( |eight) A = (0, 3, 124) Pr(T | S,A) = Pr(0, 3, 124 | 2,4) Pr(\u03c6| flight) Pr( Pr(T | S,A) = Pr(2, 34, 1 | 2,4) Pr(\u03c6\u0225| flight) Pr( |eight) A = (12, 34, 0) Pr(T | S,A) = Pr(12, 34, 0 | 2,4) Pr(\u03c6\u0225| flight) Pr($empty$|eight)",
+                "uris": null,
+                "type_str": "figure",
+                "num": null
+            },
+            "FIGREF6": {
+                "text": "j+B(A, k), j+E (A, k) P",
+                "uris": null,
+                "type_str": "figure",
+                "num": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-1014.json

@@ -0,0 +1,778 @@
+{
+    "paper_id": "O01-1014",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:42.932704Z"
+    },
+    "title": "Metaphor, Inference, and Conceptualisation * : On the Development of V-diao Construction in Mandarin",
+    "authors": [
+        {
+            "first": "Wei-Lun",
+            "middle": [],
+            "last": "Louis",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Taiwan University",
+                "location": {}
+            },
+            "email": "weilunlu@taiwan.com"
+        },
+        {
+            "first": "",
+            "middle": [],
+            "last": "Lu",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Taiwan University",
+                "location": {}
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "V-diao constructions, according to their semantics, fall into three categories: A) Physical disappearance from its original position, with the V slot filled by physical verbs, such as tao-diao \"escape\", diu-diao \"throw away\", and so on. B) Disappearance from a certain conceptual domain, rather than from the physical space, with the V slot filled by less physically perceivable verbs, such as jie-diao \"quit\", wang-diao \"forget\" and the like. C) The third category of V-diao involves speaker's subjective, always negative, attitude to the result. Examples include: lan-diao \"rot\", ruan-diao \"soften\", huang-diao \"turn yellow\", and so forth. The meaning in Type C constructions cannot be gained by simply putting their component parts together, so in this study, I shall term V-diao as a construction (Goldberg 1995) rather than merely a resultative compound (Li and Thompson 1981). Metaphor, as a mechanism of semantic change (Sweetser 1990, Bybee, Perkins and Pagliuca 1994, Heine, Claudi and Hunnemeyer 1991), is a plausible account of the polysemy between Type A and B. Type A denotes disappearance from physical space, while Type B disappearance from the conceptual space. I thus speculate on the mapping relation between the physical and the abstract, conceptual domain. Other than metaphor, pragmatic inference is claimed to be a major mechanism of semantic change (Hopper and Traugott 1993, Bybee, Perkins and Pagliuca 1994). In such changes, context plays a crucial role. Frequent use of a grammatical or lexical unit in a particular context may lead to the inference that the context is an integral part of its meaning. The development of Type C V-diao may relate to frequent co-occurrence of negative verbs and-diao. (The reason why only negative verbs are allowed in the construction will be further addressed in the next section.) Consequently, negative connotation may spread to the entire construction and give rise to the constructional meaning.",
+    "pdf_parse": {
+        "paper_id": "O01-1014",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "V-diao constructions, according to their semantics, fall into three categories: A) Physical disappearance from its original position, with the V slot filled by physical verbs, such as tao-diao \"escape\", diu-diao \"throw away\", and so on. B) Disappearance from a certain conceptual domain, rather than from the physical space, with the V slot filled by less physically perceivable verbs, such as jie-diao \"quit\", wang-diao \"forget\" and the like. C) The third category of V-diao involves speaker's subjective, always negative, attitude to the result. Examples include: lan-diao \"rot\", ruan-diao \"soften\", huang-diao \"turn yellow\", and so forth. The meaning in Type C constructions cannot be gained by simply putting their component parts together, so in this study, I shall term V-diao as a construction (Goldberg 1995) rather than merely a resultative compound (Li and Thompson 1981). Metaphor, as a mechanism of semantic change (Sweetser 1990, Bybee, Perkins and Pagliuca 1994, Heine, Claudi and Hunnemeyer 1991), is a plausible account of the polysemy between Type A and B. Type A denotes disappearance from physical space, while Type B disappearance from the conceptual space. I thus speculate on the mapping relation between the physical and the abstract, conceptual domain. Other than metaphor, pragmatic inference is claimed to be a major mechanism of semantic change (Hopper and Traugott 1993, Bybee, Perkins and Pagliuca 1994). In such changes, context plays a crucial role. Frequent use of a grammatical or lexical unit in a particular context may lead to the inference that the context is an integral part of its meaning. The development of Type C V-diao may relate to frequent co-occurrence of negative verbs and-diao. (The reason why only negative verbs are allowed in the construction will be further addressed in the next section.) Consequently, negative connotation may spread to the entire construction and give rise to the constructional meaning.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "There also exists a cognitive constraint on its applicability. The construction does not allow verbs with positive connotation in the V slot. This is because, the semantics of the construction cannot contradict the metaphor it is based on (Huang and Chang 1996) . Also, it cannot override, either, the orientational metaphor based on human experiential basis (Lakoff and Johnson 1980) : GOOD IS UP; DOWN IS BAD.",
+                "cite_spans": [
+                    {
+                        "start": 239,
+                        "end": 261,
+                        "text": "(Huang and Chang 1996)",
+                        "ref_id": "BIBREF7"
+                    },
+                    {
+                        "start": 359,
+                        "end": 384,
+                        "text": "(Lakoff and Johnson 1980)",
+                        "ref_id": "BIBREF8"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Hopefully, this study can serve as a valid argument for the interaction of our language use and grammar, and the conceptual basis of human language.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "V-diao is traditionally termed as a resultative compound (Li and Thompson 1981) . However, in this study, I shall avert this conventional terminology and treat it as a construction instead, because it actually denotes something more than what its components literally give. In this paper, I shall adopt the definition of Goldberg (1995) , and Fillmore, Kay, and O'Connor (1988) , to define a \"construction\" as follows: A construction refers to a form-meaning pair, the meaning of which cannot be strictly predictable from its component parts or from other previously established constructions. It may specify not only syntactic, but also lexical, semantic, and pragmatic information.",
+                "cite_spans": [
+                    {
+                        "start": 57,
+                        "end": 79,
+                        "text": "(Li and Thompson 1981)",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 321,
+                        "end": 336,
+                        "text": "Goldberg (1995)",
+                        "ref_id": "BIBREF3"
+                    },
+                    {
+                        "start": 339,
+                        "end": 377,
+                        "text": "and Fillmore, Kay, and O'Connor (1988)",
+                        "ref_id": "BIBREF2"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "V-diao as a Construction",
+                "sec_num": "1"
+            },
+            {
+                "text": "But as a construction, what is its constructional meaning? Also, what are the driving forces of the emergence of constructional meaning? Furthermore, a selectional restriction seems to stand in what properly fits into the V-slot. In this paper, I shall attempt to look into the constructional meaning, its driving forces, and finally, the selectional constraint of the verb. V-diao construction comprises a verb (be it action or stative), and a verbal suffix -diao. It gives the final state of the agent, if used intransitively, and of the receiver of the action, in transitive cases. It may represent: A) Physical disappearance of an entity from its original position, B) Disappearance from a certain conceptual domain, and C) Speaker's subjective evaluation on the result of an event, as in (1)- 3respectively.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "V-diao as a Construction",
+                "sec_num": "1"
+            },
+            {
+                "text": "(1) ta qiaoqiao pao-diao le he quietly run away CRS \"He ran away quietly.\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "V-diao as a Construction",
+                "sec_num": "1"
+            },
+            {
+                "text": "(2) ta jie-diao le nage huai",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "V-diao as a Construction",
+                "sec_num": "1"
+            },
+            {
+                "text": "xiguan he get rid of Perf that bad habit \"He got rid of that bad habit.\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "V-diao as a Construction",
+                "sec_num": "1"
+            },
+            {
+                "text": "(3) diennau zuotien huai-diao le computer yesterday break down CRS \"The computer broke down yesterday.\" I shall begin this paper with a close look at the semantics of the foregoing types of V-diao, especially the last one, since Type C constructions involve an intriguing phenomenon: the interpretation of negative results cannot be gained directly from the compositional parts of a construction, as indicated in Goldberg (1995) . Later I shall further look into how the constructional meaning emerges.",
+                "cite_spans": [
+                    {
+                        "start": 413,
+                        "end": 428,
+                        "text": "Goldberg (1995)",
+                        "ref_id": "BIBREF3"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "V-diao as a Construction",
+                "sec_num": "1"
+            },
+            {
+                "text": "It is reported that a suffix in a resultative verb compound indicates the result of an action (Li and Thompson 1981) . The first kind of -diao gives the final state, i.e., physical absence, of the agent or the patient. Mostly this kind of -diao is affixed to easily perceivable physical action verbs such as pao \"run\", as in (1), diu \"throw\", shao \"burn\", and so on.",
+                "cite_spans": [
+                    {
+                        "start": 94,
+                        "end": 116,
+                        "text": "(Li and Thompson 1981)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Physical Disappearance",
+                "sec_num": "1.1"
+            },
+            {
+                "text": "The second sort of V-diao denotes also the result of an action. However, this differs from type A in the sense that it represents a less \"concrete\" disappearance. It is often attached to low transitive verbs, without obvious physical motion, and accompanies an abstract noun phrase. Consider example (2) again:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Disappearance from a Conceptual Domain",
+                "sec_num": "1.2"
+            },
+            {
+                "text": "( This has everything to do with our conceptual system. We experience many things, through sight and touch, as having distinct physical shapes and boundaries.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Disappearance from a Conceptual Domain",
+                "sec_num": "1.2"
+            },
+            {
+                "text": "We thus tend to project physical shapes and boundaries on them, conceptualising them as entities and imposing on them physical characteristics such as existence and disappearance, even though we can never really feel them with our hands or sense them with our eyes or nose (Lakoff and Johnson 1980) . Therefore, in this case, a habit is conceptualised as a physical entity. It can be done away with, fade out, and finally disappear from our conceptual domain as physical things do from the physical space. Thus, Type B seems to represent the final state of usually a non-physical action, i.e., an abstract entity being done away with and finally disappearing from one's conceptual domain.",
+                "cite_spans": [
+                    {
+                        "start": 273,
+                        "end": 298,
+                        "text": "(Lakoff and Johnson 1980)",
+                        "ref_id": "BIBREF8"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Disappearance from a Conceptual Domain",
+                "sec_num": "1.2"
+            },
+            {
+                "text": "Type C V-diao denotes a somewhat negative evaluation on the result in question.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Negative Evaluation on the Result",
+                "sec_num": "1.3"
+            },
+            {
+                "text": "It often co-occurs with verbs with negative connotation, such as lan-diao \"rotten\", si-diao \"die\", shu-diao \"lose\", and et cetera. However, its negative meaning does not seem to come from the preceding verb in every case. Consider the following instances (4) and 5 \"Vegetables won't be fresh if they turn yellow.\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Negative Evaluation on the Result",
+                "sec_num": "1.3"
+            },
+            {
+                "text": "In (4) and (5), the words huang \"yellow\" and ruan \"soft\" do not themselves carry negative meanings, but the constructions clearly involve one's unfavourable attitude to the final state of vegetables and cookies. The constructional meaning, which carries negative attitude, cannot be gained from the compositional parts (Goldberg 1995) , in this case, -diao and the verb preceding it. In the following sections, I shall examine the semantic change of -diao, and try to account for the emergence of the constructional meaning of V-diao.",
+                "cite_spans": [
+                    {
+                        "start": 319,
+                        "end": 334,
+                        "text": "(Goldberg 1995)",
+                        "ref_id": "BIBREF3"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Negative Evaluation on the Result",
+                "sec_num": "1.3"
+            },
+            {
+                "text": "Two main sources provide the examples of the expressions discussed in this research. The written source mostly comes from the Academia Sinica Corpus. The spoken source comprises the Taida Spoken Corpus, and another eight hours of data from Professor Lily I-wen Su. An approximate total of sixteen hours of conversational Mandarin is adopted to serve the purpose of this study.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Data and Methodology",
+                "sec_num": "1.4"
+            },
+            {
+                "text": "It is argued that, when a grammatical meaning is derived from its source, there often exists a metaphorical relation between the two meanings (Sweetser 1990, Bybee, Perkins and Pagliuca 1994) . Such semantic change takes place to serve certain functional end in grammar and discourse, as indicated by Heine, Claudi and Hunnemeyer (1991:48) :",
+                "cite_spans": [
+                    {
+                        "start": 142,
+                        "end": 176,
+                        "text": "(Sweetser 1990, Bybee, Perkins and",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 177,
+                        "end": 191,
+                        "text": "Pagliuca 1994)",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 301,
+                        "end": 339,
+                        "text": "Heine, Claudi and Hunnemeyer (1991:48)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Relation",
+                "sec_num": "2"
+            },
+            {
+                "text": "We try to demonstrate that metaphorical transfer forms one of the main driving forces in the development of grammatical categories; that is, in order to express more \"abstract\" functions, concrete entities are recruited.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Relation",
+                "sec_num": "2"
+            },
+            {
+                "text": "Similarly, Goldberg considers metaphor a mechanism to develop polysemous construction. Her study on the way construction indicates that metaphor be a plausible cause of semantic change, since it involves a \"metaphorical self-created path\" (1995: 203) . This corresponds to my observation on V-diao: a metaphorical transfer takes place when the construction proceeds from the physical domain to a conceptual domain, denoting metaphorical disappearance.",
+                "cite_spans": [
+                    {
+                        "start": 239,
+                        "end": 250,
+                        "text": "(1995: 203)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Relation",
+                "sec_num": "2"
+            },
+            {
+                "text": "The above claim seems to be the case in the development of -diao. Pao-bu-diao in (6a) denotes the unsuccessful outcome of the agent's escape.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type A to B: Metaphor at Work",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "The agent fails to escape and will not disappear. In (6b), it means that, the landmark \"a hundred thousand\" is certain to be met. However, not every single case of Type B has a counterpart in A. Actually, most Type B constructions do not. Pao-bu-diao is simply a case employed to illustrate the metaphorical relation between the two polysemous constructions. In most cases of type B constructions, the V slot is filled by less physical verbs, such as jie \"get rid of\" in (2), hulue \"ignore\", wang \"forget\", and so on.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type A to B: Metaphor at Work",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "In this section, I have shown that the physical \"resultative compound\" V-diao has undergone a metaphorical transfer, and develops the sense of disappearance from a conceptual domain. Thus, it makes perfect sense to speculate that the polysemy of the construction is at least partly contributed by metaphor, since disappearance is a common feature of Type A and B. This corresponds to the observation of Goldberg Other than metaphor, pragmatic inference is claimed to be a major mechanism of semantic change (Hopper and Traugott 1993, Bybee, Perkins and Pagliuca 1994) . In such changes, context plays a crucial role. Frequent use of a grammatical or lexical unit in a particular context may lead to the inference that the context is an incorporated part of its meaning. Goossens' research on Old English modals (1982) reports that, there were rarely \"real\" epistemic markers in OE, and that possibility markers frequently combined with adverbs to express epistemic functions. That is, speakers could have generalised and have extracted the epistemic meanings from the context and have imposed them on modals. This suggests that frequent co-occurrence with a particular context may \"colour\" the semantics of a gram.",
+                "cite_spans": [
+                    {
+                        "start": 507,
+                        "end": 518,
+                        "text": "(Hopper and",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 519,
+                        "end": 552,
+                        "text": "Traugott 1993, Bybee, Perkins and",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 553,
+                        "end": 567,
+                        "text": "Pagliuca 1994)",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 811,
+                        "end": 817,
+                        "text": "(1982)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "It is highly likely that the final stage of development of V-diao is based on such mechanism. I have argued for the existence of the constructional meaning in 1.3. Now let us see how language use and context collaborate to lead to the constructional meaning in this case.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "In Type C construction, the sense of disappearance retains, but there seems to be something more than the combination of the verbal sense and disappearance. In general, these constructions involve undesirable assessment from the speaker. That is, the speaker obviously does not favour the consequence of the change of state.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type B to C: Emergence of Constructional Meaning",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "It is noteworthy that Type C constructions can be further divided into two subtypes by the verbs in the V slot: 1) Verbs with negative connotation, such as lan \"rot\", si \"die\", po \"break\", shu \"lose\", and so on. 2) Neutral verbs, such as huang \"yellow\", ya \"croak\", ruan \"soft\", and so on. This classification highly pertains to the emergence of constructional meaning. Let us see how.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type B to C: Emergence of Constructional Meaning",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "Initially, only the former constructions are formed. They simply denote a metaphorical disappearance, labeled Type B. As frequency of use increases, speakers tend to associate the construction with the adverse image. Such frequent collocation of verbs and the suffix may invite the inference that the constructions are used to express one's unfavourable appraisal of the situation at issue. The context is thus \"semanticized\" (Hopper and Traugott 1993:75) , and becomes an integral part of the construction. Consequently, the construction may accommodate neutral stative verbs in the V slot and still gain a negative interpretation. See (4) and (5) again for illustration: \"Vegetables won't be fresh if they turn yellow.\"",
+                "cite_spans": [
+                    {
+                        "start": 426,
+                        "end": 455,
+                        "text": "(Hopper and Traugott 1993:75)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type B to C: Emergence of Constructional Meaning",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "\"Yellow\" and \"soft\" themselves do not signal any adversity. The adverse meaning is subtly signalled and triggered by the frequent occurrence of negative verbs in the construction. In other words, the constructional meaning, i.e. speaker's negative attitude, derives neither from the suffix denoting disappearance nor the verb preceding it, but could have been generalised from the constant collocation of negative words and -diao. Now even neutral verbs may fit into the V slot and indicate negative assessments.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type B to C: Emergence of Constructional Meaning",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "Pragmatic inference is one of the driving forces of semantic change, and I have proven that it is at crucial play in the development of V-diao construction as well. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "Previous studies on Mandarin -qilai constructions claim that the development of grammatical units cannot contradict the metaphor that they are based on, and that the collocation of -qilai and verbs are conceptually restricted on a semantic basis (Chang 1994, Huang and Chang 1996) . My following observation on V-diao corresponds to this claim.",
+                "cite_spans": [
+                    {
+                        "start": 246,
+                        "end": 268,
+                        "text": "(Chang 1994, Huang and",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 269,
+                        "end": 280,
+                        "text": "Chang 1996)",
+                        "ref_id": "BIBREF7"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "I have argued for metaphor as the driving force of semantic change from Type A to Type B V-diao. Further, this metaphorical transfer obeys the orientational metaphor GOOD IS UP; BAD IS DOWN (Lakoff and Johnson 1980:16) :",
+                "cite_spans": [
+                    {
+                        "start": 190,
+                        "end": 218,
+                        "text": "(Lakoff and Johnson 1980:16)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "Physical basis for personal well-being: Happiness, health, life, and control-the things that principally characterize what is good for a person-are all UP.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "The physical and experiential basis for DOWN IS BAD is also evident in our language use and conceptual system. Syncronically, the most basic meaning of diao is physical dropping / falling and involves downward movement. It follows that diao can relate to something bad in our conceptual system. Be it grammaticalised or not, diao should never override the conceptual restriction to modify something good.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "In other words, if the metaphor DOWN IS BAD is truly at work in the emergence of the construction, it seems rather natural for the construction not to accommodate a verb with positive connotation. Thus, conceptual / cognitive restriction can fully account for the intrinsic incompatibility of positive verbs in V-diao construction.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "The above semantic restriction is critical in the development from Type B to C V-diao, and without it, the rise of constructional meaning would be impossible. The constructional meaning is language users' generalisation from a previous existing pattern. The constraint must have existed prior to the formation of constructional meaning. Otherwise, without such a selectional restriction, the construction would fail to emerge, since positive verbs would intervene. Therefore, it is justified to say that this constraint metaphorically shapes, or at least helps to shape, the constructional meaning.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "In this section, the incompatibility of positive verbs and -diao is closely examined from a semantic viewpoint. The meaning of diao metaphorically constrains the verb types it co-occurs with, which proves the metaphorical nature of our conceptual system. Also, such selectional restriction results in the emergence of constructional meaning. The metaphorical condition on constructional meaning thus reflects the interaction between grammar and human conceptual system.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "In this study, I have classified V-diao constructions according to their semantics, and have explained the constructional meaning. In the second section, metaphorical transfer is argued to be an important mechanism in the development of the construction. Furthermore, I have discussed how pragmatic inference enables language users to arrive at the constructional meaning. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "5"
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of the World",
+                "authors": [
+                    {
+                        "first": "Joan",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Bybee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Revere",
+                        "middle": [],
+                        "last": "Perkins",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "William",
+                        "middle": [],
+                        "last": "Pagliuca",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Bybee, Joan L., Revere Perkins, and William Pagliuca. 1994. The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of the World. Chicago: The University of Chicago Press.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "V-qi-lai Constructions in Mandarin Chinese: A Study of Their Semantics and Syntax",
+                "authors": [
+                    {
+                        "first": "Shen-Ming",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chang, Shen-ming. 1994. V-qi-lai Constructions in Mandarin Chinese: A Study of Their Semantics and Syntax. M. A. Thesis. National Tsing Hua University.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Regularity and Idiomaticity in Grammatical Constructions: The Case of Let Alone",
+                "authors": [
+                    {
+                        "first": "Charles",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Fillmore",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Paul",
+                        "middle": [],
+                        "last": "Kay",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Mary",
+                        "middle": [],
+                        "last": "Catherine",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "O'",
+                        "middle": [],
+                        "last": "Connor",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1988,
+                "venue": "Language",
+                "volume": "64",
+                "issue": "",
+                "pages": "501--539",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Fillmore, Charles J., Paul Kay, and Mary Catherine O'Connor. 1988. Regularity and Idiomaticity in Grammatical Constructions: The Case of Let Alone. Language 64:501-38",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Constructions: A Construction Grammar Approach to Argument Structure",
+                "authors": [
+                    {
+                        "first": "Adele",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Goldberg",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Goldberg, Adele E. 1995. Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "On the Development of the Modals and of the Epistemic Functions in English",
+                "authors": [
+                    {
+                        "first": "Louis",
+                        "middle": [],
+                        "last": "Goossens",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1982,
+                "venue": "Papers from the Fifth International Conference on Historical Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "74--84",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Goossens, Louis. 1982. On the Development of the Modals and of the Epistemic Functions in English. Papers from the Fifth International Conference on Historical Linguistics, ed. by Anders Ahlqvist, 74-84. Amsterdam: Benjamins.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "From Cognition to Grammar --Evidence from African Languages",
+                "authors": [
+                    {
+                        "first": "Bernd",
+                        "middle": [],
+                        "last": "Heine",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ulrike",
+                        "middle": [],
+                        "last": "Claudi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Friederike",
+                        "middle": [],
+                        "last": "Hunnemeyer",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "",
+                "volume": "1",
+                "issue": "",
+                "pages": "149--87",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Heine, Bernd, Ulrike Claudi, and Friederike Hunnemeyer. 1991. From Cognition to Grammar --Evidence from African Languages. In Traugott and Heine eds., Vol. 1, 149-87.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Metaphor, Metaphorical Extension, and Grammaticalization: A Study of Mandarin Chinese -qilai",
+                "authors": [
+                    {
+                        "first": "Chu",
+                        "middle": [
+                            "-"
+                        ],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Shen-Ming",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Conceptual Structure, Discourse, and Language",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Huang, Chu-ren and Shen-ming Chang. 1996. Metaphor, Metaphorical Extension, and Grammaticalization: A Study of Mandarin Chinese -qilai. Conceptual Structure, Discourse, and Language. ed., by Adele Goldberg. CSLI.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Metaphors We Live by",
+                "authors": [
+                    {
+                        "first": "George",
+                        "middle": [],
+                        "last": "Lakoff",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Mark",
+                        "middle": [],
+                        "last": "Johnson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1980,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lakoff, George, and Mark Johnson. 1980. Metaphors We Live by. Chicago: University of Chicago Press.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure",
+                "authors": [
+                    {
+                        "first": "Eve",
+                        "middle": [
+                            "Eliot"
+                        ],
+                        "last": "Sweetser",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Cambrige",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Sweetser, Eve Eliot. 1990. From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambrige: Cambridge University Press.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "type_str": "figure",
+                "num": null,
+                "text": "1995).Figure 1is representative of the mapping relation between Type A and B",
+                "uris": null
+            },
+            "FIGREF1": {
+                "type_str": "figure",
+                "num": null,
+                "text": "only verbs that result in physical and conceptual disappearance may occur in the construction. Then a group of verbs with negative connotation prompts a deduction of constructional meaning. Consequently, the negative sense of the verbs has transferred onto the entire construction, and the constructional meaning is drawn: the speaker's undesirable appraisal of the result. The following figure illustrates the development path from Type B to C:",
+                "uris": null
+            },
+            "FIGREF2": {
+                "type_str": "figure",
+                "num": null,
+                "text": "Figure 2 Semanticization of Context and Emergence of",
+                "uris": null
+            },
+            "FIGREF3": {
+                "type_str": "figure",
+                "num": null,
+                "text": "Figure 2shows the different stages of V-diao construction and their change of mechanism.Finally, a selectional restriction on the V slot exists. The exclusion of positive verbs is conceptually conditioned by the semantics of diao. This suggests, the semantic change and grammaticalisation process of a grammatical unit, or a construction, is conditioned by human physical and experiential basis. Hopefully, this study may serve as a valid argument for the interaction of our language use and grammar, and the conceptual basis of human language. Development of V-diao and Change of Mechanism",
+                "uris": null
+            },
+            "TABREF4": {
+                "html": null,
+                "type_str": "table",
+                "num": null,
+                "content": "<table><tr><td colspan=\"5\">As the construction develops its polysemy, it comes be used in increasingly</td></tr><tr><td colspan=\"5\">wider contexts. At the beginning it only accommodates physical verbs and denotes</td></tr><tr><td colspan=\"5\">physical disappearance. It further proceeds to tolerate less physical verbs and</td></tr><tr><td colspan=\"5\">metaphor allows a sense of conceptual disappearance. Finally, it may apply to a</td></tr><tr><td colspan=\"5\">variety of stative verbs to express speaker attitude. Nevertheless, in spite of its</td></tr><tr><td colspan=\"5\">seemingly free occurrence, some restriction still exists. Consider the following pairs</td></tr><tr><td>for illustration:</td><td/><td/><td/><td/></tr><tr><td>(7) a. wo</td><td>zhengge</td><td>ren</td><td>sha-diao</td><td>le</td></tr><tr><td>I</td><td>entire</td><td>person</td><td>stun-Suffix</td><td>CRS</td></tr><tr><td colspan=\"2\">\"I was entirely stunned.\"</td><td/><td/><td/></tr><tr><td>b. *wo</td><td colspan=\"2\">congming-diao</td><td>le</td><td/></tr><tr><td>I</td><td colspan=\"2\">smart-Suffix</td><td>CRS</td><td/></tr><tr><td>(8) a. dongxi</td><td colspan=\"2\">langfie-diao</td><td>le</td><td/></tr><tr><td>thing</td><td colspan=\"2\">waste-Suffix</td><td>CRS</td><td/></tr><tr><td colspan=\"2\">\"The thing is wasted.\"</td><td/><td/><td/></tr><tr><td>b. *dongxi</td><td colspan=\"2\">jenxi-diao</td><td>le</td><td/></tr><tr><td>thing</td><td colspan=\"2\">cherish-Suffix</td><td>CRS</td><td/></tr><tr><td colspan=\"5\">From the above pairs, it is evident that the V slot does not allow verbs with</td></tr><tr><td colspan=\"5\">positive connotation. It seems that the semantics of positive verbs clashes with that</td></tr><tr><td colspan=\"5\">of the entire construction. Why is this the case? What is basis of such selectional</td></tr><tr><td>restriction?</td><td/><td/><td/><td/></tr></table>",
+                "text": ""
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-2005.json

@@ -0,0 +1,1429 @@
+{
+    "paper_id": "O01-2005",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:40.298708Z"
+    },
+    "title": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion",
+    "authors": [
+        {
+            "first": "Jian",
+            "middle": [],
+            "last": "Zhang",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        },
+        {
+            "first": "Jianfeng",
+            "middle": [],
+            "last": "Gao",
+            "suffix": "",
+            "affiliation": {},
+            "email": "jfgao@microsoft.com"
+        },
+        {
+            "first": "Ming",
+            "middle": [],
+            "last": "Zhou",
+            "suffix": "",
+            "affiliation": {},
+            "email": "mingzhou@microsoft.com"
+        },
+        {
+            "first": "Jiaxing",
+            "middle": [],
+            "last": "Wang",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "Fusion and clustering are two approaches to improving the effectiveness of information retrieval. In fusion, ranked lists are combined together by various means. The motivation is that different IR systems will complement each other, because they usually emphasize different query features when determining relevance and retrieve different sets of documents. In clustering, documents are clustered either before or after retrieval. The motivation is that similar documents tend to be relevant to the same query so that this approach is likely to retrieve more relevant documents by identifying clusters of similar documents. In this paper, we present a novel fusion technique that can be combined with clustering to achieve consistent improvements over conventional approaches. Our method involves three steps: (1) clustering similar documents, (2) re-ranking retrieval results, and (3) combining retrieval results.",
+    "pdf_parse": {
+        "paper_id": "O01-2005",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "Fusion and clustering are two approaches to improving the effectiveness of information retrieval. In fusion, ranked lists are combined together by various means. The motivation is that different IR systems will complement each other, because they usually emphasize different query features when determining relevance and retrieve different sets of documents. In clustering, documents are clustered either before or after retrieval. The motivation is that similar documents tend to be relevant to the same query so that this approach is likely to retrieve more relevant documents by identifying clusters of similar documents. In this paper, we present a novel fusion technique that can be combined with clustering to achieve consistent improvements over conventional approaches. Our method involves three steps: (1) clustering similar documents, (2) re-ranking retrieval results, and (3) combining retrieval results.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "In terms of the overall performance on a large query set, none of the typical IR systems outperform others substantially, while for each individual query, the performance that different systems achieve varies greatly [Voorhees 1997 ]. This observation leads to the idea of combining results obtained by different IR systems to improve overall performance.",
+                "cite_spans": [
+                    {
+                        "start": 217,
+                        "end": 231,
+                        "text": "[Voorhees 1997",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Fusion is a technique that combines retrieval results (or ranked lists) obtained by different systems. However, conventional fusion techniques only consider retrieval results, while the information embedded in the document collection (e.g. the similarity between documents) is ignored. On the other hand, document clustering applies the structure of a document collection, but it usually considers each individual ranked list separately and is not able to take advantage of multiple ranked lists.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we present a novel fusion technique that can be combined with clustering. Given multiple retrieval results obtained by different IR systems, we first perform clustering on each ranked list and obtain a set of clusters. We then identify the clusters that contain the most relevant documents. Each of these clusters is evaluated based on a metric called reliability. Documents in reliable clusters are re-ranked. That is, we set higher scores for these documents. Finally, a conventional fusion method is applied to combine multiple retrieval results, which are re-ranked. Our experiments on the TREC-5 Chinese collection show that the above approach achieves consistent improvements over conventional approaches.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "The remainder of this paper is organized as follows. Section 2 gives a brief survey of related work. In Section 3, we describe our method in detail. In Section 4, a series of experiments are presented to show the effectiveness of our approach. Finally, we present our conclusions in Section 5.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Fusion and clustering have been important research topics for many researchers.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": "Fox and Shaw [Fox 1994 ] reported on their work on result sets fusion. Their method for combining the evidence from multiple retrieval runs is based on document-query similarities in different sets. Five combining strategies were investigated, as summarized in Table 1 . In their experiments, CombSUM and CombMNZ were better than the others. Thompson's work [Thompson 1990 ] includes assigning to each ranked list a variable weight based on the prior performance of the system. His idea is that a retrieval system should be considered preferable to others if its prior performance is better. Thompson's results were slightly better than Fox's.",
+                "cite_spans": [
+                    {
+                        "start": 13,
+                        "end": 22,
+                        "text": "[Fox 1994",
+                        "ref_id": "BIBREF2"
+                    },
+                    {
+                        "start": 358,
+                        "end": 372,
+                        "text": "[Thompson 1990",
+                        "ref_id": "BIBREF8"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 261,
+                        "end": 268,
+                        "text": "Table 1",
+                        "ref_id": "TABREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": "Bartell [Bartell 1994 ] used numerical optimization techniques to determine optimal scalars (weights) for a linear combination of results. The idea is similar to Thompson's except that Bartell obtained the optimal scalars from training data, while Thompson constructed scalars based on their prior performance. Bartell achieved good results on a relatively small collection (less than 50MB).",
+                "cite_spans": [
+                    {
+                        "start": 8,
+                        "end": 21,
+                        "text": "[Bartell 1994",
+                        "ref_id": "BIBREF0"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": "To perform fusion more effectively, researchers began to investigate whether two result sets are suitable for fusion by examining some critical characteristics. Lee [Lee 1997 ] found that the overlap of the result sets was an important factor for fusion. Overlap ratios of relevant and non-relevant documents are calculated as follows:",
+                "cite_spans": [
+                    {
+                        "start": 165,
+                        "end": 174,
+                        "text": "[Lee 1997",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": ", 2 , 2 B A common overlap B A common overlap N N N N R R R R + \u00d7 = + \u00d7 =",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": "where A R and A N are, respectively, the numbers of relevant and irrelevant documents in result set into our fusion approach.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": "Vogt [Vogt 1998 [Vogt , 1999 tested different linear combinations of several results from TREC-5. 36,600 result pairs were tested. A linear regression of several potential indicators was performed to determine the potential improvement for result sets to be fused. Thirteen factors including measures of individual inputs, such as average precision/recall, and some pairwise factors, such as overlap and unique document counts, were considered. Vogt concluded that the characteristics for effective fusion are: (1) at least one result has high precision/recall; (2) a high overlap of relevant documents and a low overlap of non-relevant documents;",
+                "cite_spans": [
+                    {
+                        "start": 5,
+                        "end": 15,
+                        "text": "[Vogt 1998",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 16,
+                        "end": 28,
+                        "text": "[Vogt , 1999",
+                        "ref_id": "BIBREF11"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": "(3) similar distributions of relevance scores; and (4) each retrieval system ranks relevant documents differently. Conclusion (1) and(2) are also confirmed by our experiments, as will be shown in Section 4.3.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": "Clustering is now considered to be a useful information retrieval method for not only documents categorization but also interactive retrieval. The use of clustering in information retrieval is based on the Clustering Hypothesis [Rijsbergen, 1979] : \"closely associated documents tend to be relevant to the same requests\". Hearst [Hearst 1996] showed that this hypothesis holds for a set of documents returned by a retrieval system. According to this hypothesis, if we do a good job of clustering the retrieved documents, we will likely separate the relevant and non-relevant documents into different groups. If we can direct the user to the correct group of documents, we can enhance the likelihood of finding interesting information for the user. Previous works [Cutting et al, 1992] , [Leuski 1999] and [Leuski 2000 ] focused on clustering documents and let users select the clusters they were interested in. Their approaches are interactive. Most of the clustering methods mentioned above work on individual ranked lists and do not take advantage of multiple ranked lists.",
+                "cite_spans": [
+                    {
+                        "start": 228,
+                        "end": 246,
+                        "text": "[Rijsbergen, 1979]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 329,
+                        "end": 342,
+                        "text": "[Hearst 1996]",
+                        "ref_id": "BIBREF3"
+                    },
+                    {
+                        "start": 763,
+                        "end": 784,
+                        "text": "[Cutting et al, 1992]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 787,
+                        "end": 800,
+                        "text": "[Leuski 1999]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 805,
+                        "end": 817,
+                        "text": "[Leuski 2000",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": "In this paper, we combine clustering with fusion. Our approach differs from interactive approaches in three ways. First, we use two or more ranked lists, while others usually use one in clustering. Second, user interactive input is not needed in our approach. Third, we provide a ranked list of documents to the user instead of a set of clusters.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Related Work",
+                "sec_num": "2."
+            },
+            {
+                "text": "Our method is based on two hypotheses:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Fusion with Clustering",
+                "sec_num": "3."
+            },
+            {
+                "text": "Clustering Hypothesis: Documents that are relevant to the same query can be clustered together since they tend to be more similar to each other than to non-relevant documents.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Fusion with Clustering",
+                "sec_num": "3."
+            },
+            {
+                "text": "Fusion Hypothesis: Different ranked lists usually have a high overlap of relevant documents and a low overlap of non-relevant documents.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Fusion with Clustering",
+                "sec_num": "3."
+            },
+            {
+                "text": "The Clustering Hypothesis suggests that we might be able to roughly separate relevant documents from non-relevant documents with a proper clustering algorithm. Relevant documents can be clustered into one or several clusters, and these clusters will contain more relevant documents than others. We call such a cluster a reliable cluster.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 113",
+                "sec_num": null
+            },
+            {
+                "text": "The Fusion hypothesis presents the idea of identifying reliable clusters. The reliable clusters from different ranked lists usually have a high overlap. Therefore, the more relevant documents a cluster contains, the more reliable the cluster is. We will describe the computation of reliability in detail in Section 3.3. Our approach consists of three steps. First, we cluster each ranked list. Then, we identify the reliable clusters and adjust the relevance value of each document according to the reliability of the cluster. Finally, we use CombSUM to combine the adjusted ranked lists and present the result to user.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 113",
+                "sec_num": null
+            },
+            {
+                "text": "In the following sections, we will describe our approach in more detail. For conciseness, we will use some symbols to present our approach, which are listed in Table 2 with their explanations. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 160,
+                        "end": 167,
+                        "text": "Table 2",
+                        "ref_id": "TABREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 113",
+                "sec_num": null
+            },
+            {
+                "text": "q A query d A document A RL , B RL",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 113",
+                "sec_num": null
+            },
+            {
+                "text": "Ranked list returned by retrieval systems A and B, respectively",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 113",
+                "sec_num": null
+            },
+            {
+                "text": "i A C , i th cluster in A RL ) , ( _ , , j B i A C C CC Sim Similarity between i A C , and j B C , ) , ( _ ,i A C q qC Sim Similarity between query q and i A C , ) , ( _ j i d d dd Sim Similarity between two documents, i d and j d ) ( ,i A C r Reliability of cluster i A C , ) (d rel A Relevance score of document d given by retrieval system A ) ( * d rel A Adjusted relevance score of document d ) (d rel",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 113",
+                "sec_num": null
+            },
+            {
+                "text": "Final relevance score of document d",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 113",
+                "sec_num": null
+            },
+            {
+                "text": "The goal of clustering is to separate relevant documents from non-relevant documents. To accomplish this, we need to define a measure for the similarity between documents and design a corresponding clustering algorithm.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Clustering",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "In our experiments, we used the vector space model to represent documents. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Similarity between documents",
+                "sec_num": "3.1.1"
+            },
+            {
+                "text": ", )] / log( ) 0 . 1 ) [(log( ) / log( ] 0 . 1 ) [log( 2 \u2211 \u00d7 + \u00d7 + = j j ij k ik ik n N f n N f w (1)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Similarity between documents",
+                "sec_num": "3.1.1"
+            },
+            {
+                "text": "where ik f is the occurrence frequency of term k t in document i d , N is the total number of documents in the collection and k n is the number of documents that contain term k t .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Similarity between documents",
+                "sec_num": "3.1.1"
+            },
+            {
+                "text": "Actually, this is one of the most frequently used tf*idf weighting schemes in IR.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Similarity between documents",
+                "sec_num": "3.1.1"
+            },
+            {
+                "text": "For any two documents i d and j d , the cosine measure as given below is used to determine their similarity:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 115",
+                "sec_num": null
+            },
+            {
+                "text": ". ) ( ) , ( _ 2 2 \u2211 \u2211 \u2211 \u00d7 \u00d7 = k jk k ik k jk ik j i w w w w d d dd Sim (2)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 115",
+                "sec_num": null
+            },
+            {
+                "text": "There are many clustering algorithms for document clustering. Our goal is to cluster a small collection of documents returned by an individual retrieval system. Since the size of the collection was 1,000 in our experiments, the complexity of the clustering algorithm was not a serious problem. The ideal result is obtained when clustering gathers all relevant documents into one cluster and all non-relevant documents into the other cluster. However, this is unlikely to happen. In fact, relevant documents are usually distributed in several clusters. After clustering, each ranked list is composed of a set of clusters, say 1",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Clustering algorithm",
+                "sec_num": "3.1.2"
+            },
+            {
+                "text": "C , 2 C \u2026 n C .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Clustering algorithm",
+                "sec_num": "3.1.2"
+            },
+            {
+                "text": "The size of a cluster is the number of documents in the cluster. The clustering algorithm shown in Fig.2 cannot guarantee that the clusters will be of identical size. This causes many problems because the overlap depends on the size of each cluster.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 99,
+                        "end": 104,
+                        "text": "Fig.2",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Size of a cluster",
+                "sec_num": "3.1.3"
+            },
+            {
+                "text": "To solve this problem, we force the clusters to have the same size using the following approach. For clusters that contain a larger number of documents than the average, we remove the documents that are far from the cluster's centroid. These removed documents are added to clusters that are smaller than average 2 .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Size of a cluster",
+                "sec_num": "3.1.3"
+            },
+            {
+                "text": "Since all the clusters are of the same size, the size of a cluster becomes a parameter in our algorithm. Thus, we need to set this parameter to an optimal value to achieve the best performance. We will report experiments conducted to determine this value in Section 4.3.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Size of a cluster",
+                "sec_num": "3.1.3"
+            },
+            {
+                "text": "After clustering each ranked list, we obtain a group of clusters, each of which contains more or less relevant documents. Through re-ranking, we expect to determine reliable clusters and adjust the relevance scores of the documents in each ranked list such that the relevance scores become more reasonable. To identify reliable clusters, we assign to each cluster a reliability score. According to the Fusion Hypothesis, we use the overlap between clusters to compute the reliability of a cluster. The reliability",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": ") ( ,i A C r of cluster i A C",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": ", is computed as follows (see Table 2 for definitions of the symbols):",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 30,
+                        "end": 37,
+                        "text": "Table 2",
+                        "ref_id": "TABREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": ", ) , ( _ ) , ( _ ) , ( _ ) ( , , , , , \u2211 \u2211 \u23a5 \u23a5 \u23a5 \u23a6 \u23a4 \u23a2 \u23a2 \u23a2 \u23a3 \u23a1 = j j B i A t t B j B i A C C CC Sim C q qC Sim C q qC Sim C r (3) where , ) , ( _ , , , , j B i A j B i A C C C C CC Sim \u2229 = (4) . ) ( ) , ( _ , , , i A C d A i A C d rel C q qC Sim i A \u2211 \u2208 = (5) 2",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "The size of a cluster and the number of clusters are critical issues in clustering and have been studied by many researchers. This paper focuses on how to combine fusion and clustering together and shows the potential of this combination approach. Therefore, we use a very simple method to solve the problem. Our clustering algorithm is also very simple. Our future work will be to investigate the impacts of different algorithms.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "In equation 4, the similarity of two clusters is estimated based on the common documents they both contain. In equation 5, the similarity between a query and a cluster is estimated based on the average relevance score of the documents that the cluster contains. In equation 3, for each cluster",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "i A C , in A RL , its reliability ) ( ,i A C r",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "is defined as the weighted sum of the similarity between cluster Ai C and all the clusters in B RL . The intuition underlying this formula is that the more similar two clusters are, the more reliable they are, as illustrated in Fig.1 .",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 228,
+                        "end": 233,
+                        "text": "Fig.1",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "Since reliability represents the precision of a cluster, we use it to adjust the relevance score of the documents in each cluster. Formula (6) adjusts the relevance score of a document in a highly reliable cluster:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": ")], ( 1 [ ) ( ) ( , * t A A A C r d rel d rel + \u00d7 = (6) where t A C d , \u2208",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": ".",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Re-ranking",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "So far, each original ranked list has been adjusted by means of clustering and re-ranking. We next combine these improved ranked lists together using the following formula (i.e. CombSUM in [Fox 1994 ]):",
+                "cite_spans": [
+                    {
+                        "start": 189,
+                        "end": 198,
+                        "text": "[Fox 1994",
+                        "ref_id": "BIBREF2"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Fusion",
+                "sec_num": "3.3"
+            },
+            {
+                "text": "). ( ) ( ) (",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Fusion",
+                "sec_num": "3.3"
+            },
+            {
+                "text": "* * d rel d rel d rel B A + = (7)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Fusion",
+                "sec_num": "3.3"
+            },
+            {
+                "text": "In equation 7, the combined relevance of document d is the sum of all the adjusted relevance values that have been computed in the previous steps.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Fusion",
+                "sec_num": "3.3"
+            },
+            {
+                "text": "In this section, we will present the results of our experiments. We will first describe our experimental settings in Section 4.1. In Section 4.2, we will verify the two hypotheses described in Section 3 using the results of some experiments. In Section 4.3, we will compare our approach with the other three conventional fusion methods. Finally, we will examine the impact of cluster size.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experimental Results",
+                "sec_num": "4."
+            },
+            {
+                "text": "We used several retrieval results from the TREC-5 Chinese information retrieval track in our fusion experiments. The document collection contains articles published in the People's Daily and news released by the Xinhua News Agency. Some statistical characteristics of the collection are summarized in Tables 3. The 10 groups who took part in TREC-5 Chinese provided 20 retrieval results. We randomly picked seven ranked lists for our fusion experiments. The tags and average precision are listed in Table 4 . It is noted that the average precision is similar except for HIN300. Since the ranges of similarity values of the different retrieval results were quite different, we normalized each retrieval result before combining them. The bound of each retrieval result was mapped to [0,1] using the following formula [Lee 1997 ",
+                "cite_spans": [
+                    {
+                        "start": 815,
+                        "end": 824,
+                        "text": "[Lee 1997",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 499,
+                        "end": 506,
+                        "text": "Table 4",
+                        "ref_id": "TABREF5"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Experiment settings",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "We will first examine the two hypotheses we mentioned in Section 3.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Examining the hypotheses",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "In relation to Clustering Hypothesis, we clustered each ranked list into 10 clusters using our clustering algorithm. Table 5 shows some statistical information for the clustering results. The first row lists four kinds of clusters containing no, 1, 2-10 and more than 10 relevant document(s). The second row shows the corresponding percentage of each kind of cluster.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 117,
+                        "end": 124,
+                        "text": "Table 5",
+                        "ref_id": "TABREF7"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Examining the hypotheses",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "The third row shows the percentage of relevant documents in each kind of cluster.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 119",
+                "sec_num": null
+            },
+            {
+                "text": "From Table 5 , we can make two observations. First, about 50% of the clusters contain 1 or no relevant document. Second, most relevant documents (more than 60%) are in a small number of clusters (about 7%). According to these observations, we can draw the conclusion that relevant documents are concentrated in a few clusters.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 5,
+                        "end": 12,
+                        "text": "Table 5",
+                        "ref_id": "TABREF7"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 119",
+                "sec_num": null
+            },
+            {
+                "text": "Thus, in our experiments, the Clustering Hypothesis holds in terms of the initial retrieval result when a proper algorithm is adopted. for each combination pair. Table 6 lists some results. The last row shows that the average overlap R is 0.7688, while the corresponding average overlap N is 0.3351. It turns out that the Fusion Hypothesis holds for the retrieval results we obtained. Table 6 will also be used in Section 4.3 to confirm that overlap R is the most important factor determining the performance of fusion. We mark those rows whose overlap R scores are higher than 0.80 with the character *. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 162,
+                        "end": 169,
+                        "text": "Table 6",
+                        "ref_id": "TABREF8"
+                    },
+                    {
+                        "start": 385,
+                        "end": 392,
+                        "text": "Table 6",
+                        "ref_id": "TABREF8"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 119",
+                "sec_num": null
+            },
+            {
+                "text": "First, we studied three combination methods that were proposed by Fox, namely, CombMAX, CombSUM, and CombMNZ. Their fusion results for the same data set are listed in Table 7 . The last row lists the average precision of each combination strategy. Since the average precision of the individual retrieval systems is 0.3086 (see Table 4 ), each of these three fusion methods has improved significantly in terms of the average precision. CombSUM appears to be the best one among them. This confirms the observation in [Fox 1994 ].",
+                "cite_spans": [
+                    {
+                        "start": 515,
+                        "end": 524,
+                        "text": "[Fox 1994",
+                        "ref_id": "BIBREF2"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 167,
+                        "end": 174,
+                        "text": "Table 7",
+                        "ref_id": "TABREF9"
+                    },
+                    {
+                        "start": 327,
+                        "end": 334,
+                        "text": "Table 4",
+                        "ref_id": "TABREF5"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Comparison with conventional fusion methods",
+                "sec_num": "4.3"
+            },
+            {
+                "text": "Then, we compared the performance of our approach with that of the other three methods, as shown in the last row in Table 7 . Our new approach achieved 3% improvement over CombSUM. We also find that among all the 21 combination pairs, 17 of them are improved, compared to the results obtained using the CombSUM approach. We mark these rows with the character *. Comparing the results shown in Table 7 with those listed in Table 6 , we find that the pairs with a overlap R of over 0.80 correspond to better combination performance. We call this kind of pair a combinable pair. For example, BrklyCH1 & CLCHNA is a combinable pair. Although the average combination performance is 0.3654 (using our approach), almost all the combinable pairs exceed the average performance 3 . This again confirms the conclusion in both [Lee 1997] and [Vogt 1998 ] that the performance of fusion heavily depends on overlap R . It also reveals the limitation of our approach and of other linear fusion techniques in that a high overlap of relevant documents is a pre-requisite for performance enhancement. For those pairs that don't satisfy this pre-requisite, normal fusion may even decrease retrieval performance.",
+                "cite_spans": [
+                    {
+                        "start": 816,
+                        "end": 826,
+                        "text": "[Lee 1997]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 831,
+                        "end": 841,
+                        "text": "[Vogt 1998",
+                        "ref_id": "BIBREF10"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 116,
+                        "end": 123,
+                        "text": "Table 7",
+                        "ref_id": "TABREF9"
+                    },
+                    {
+                        "start": 393,
+                        "end": 400,
+                        "text": "Table 7",
+                        "ref_id": "TABREF9"
+                    },
+                    {
+                        "start": 422,
+                        "end": 429,
+                        "text": "Table 6",
+                        "ref_id": "TABREF8"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Comparison with conventional fusion methods",
+                "sec_num": "4.3"
+            },
+            {
+                "text": "We also compared our approach with the optimal linear combination. Since ranked lists 3 \"gmu96ca1 & gmu96cm1\" is an exception because their related overlap N score is very high.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 121",
+                "sec_num": null
+            },
+            {
+                "text": "J. Zhang et al.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "122",
+                "sec_num": null
+            },
+            {
+                "text": "are combined linearly, only the ratio of the two weights affects the final performance:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "122",
+                "sec_num": null
+            },
+            {
+                "text": ". B A combined wRL RL RL + =",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "122",
+                "sec_num": null
+            },
+            {
+                "text": "CombSUM can be taken as a special case of linear combination where w is set to be 1. When the relevant documents are known, the weight w can be optimized using some numerical method. In our experiment, the weight w was optimized using golden section search [Press 1992 ]. This approach was adopted in [Vogt 1998 ]. The average precision for the optimal linear combination we obtained is 0.3714. As shown in Fig.3 , our approach performs better than CombSUM and CombMAX and is very close to CombBest. To summarize, we can draw three conclusions from the above experiments. First, in most cases, our new approach shows better performance than most of the conventional methods, including CombSUM and CombMNZ. Second, overlap R strongly affects the performance of linear fusion. Third, the performance of our approach is very close to that of the optimal linear combination approach.",
+                "cite_spans": [
+                    {
+                        "start": 257,
+                        "end": 268,
+                        "text": "[Press 1992",
+                        "ref_id": "BIBREF13"
+                    },
+                    {
+                        "start": 301,
+                        "end": 311,
+                        "text": "[Vogt 1998",
+                        "ref_id": "BIBREF10"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 407,
+                        "end": 412,
+                        "text": "Fig.3",
+                        "ref_id": "FIGREF4"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "122",
+                "sec_num": null
+            },
+            {
+                "text": "We also studied the impact of cluster size. Table 8 shows the experimental results. When the cluster size varied from 200 to 5, the average precision did not change much. The maximum value was 0.3675 when the cluster size was 25 and the minimum value was 0.3621",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 44,
+                        "end": 51,
+                        "text": "Table 8",
+                        "ref_id": "TABREF10"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Impact of cluster size",
+                "sec_num": "4.4"
+            },
+            {
+                "text": "when the cluster size was 200. This shows that the cluster size setting has very little impact in our approach. Another interesting question is what will happen when the cluster size is set to 1000 or 1.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 123",
+                "sec_num": null
+            },
+            {
+                "text": "When the cluster size is set to 1000, each ranked list becomes a single cluster. Then, the reliability of A C and B C can be computed as follows:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 123",
+                "sec_num": null
+            },
+            {
+                "text": ". ) , ( _ ) ( ) ( B A B A B A C C C C CC Sim C r C r \u2229 = = = Since ) ( A C r and ) ( B C r",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 123",
+                "sec_num": null
+            },
+            {
+                "text": "are equal, the re-ranking and fusion step becomes a normal CombSUM approach, and the average precision is equal to that of the CombSUM approach.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 123",
+                "sec_num": null
+            },
+            {
+                "text": "When the cluster size is set to 1, each document forms a cluster by itself. Those documents appearing in both ranked lists will be improved. For those documents that only appear in one ranked list, their relevance will remain unchanged. On the other hand, the relevance score of those documents that appear in both ranked lists will be improved with a . The final result will be close to that of the CombSUM approach because this factor is close to 1.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 123",
+                "sec_num": null
+            },
+            {
+                "text": "The impact of the cluster size setting is illustrated in Fig.4 . From this figure, we find that fusion combined with clustering is consistently better than the approaches that do not include clustering (where cluster size = 1000). We find that a setting size to 25 gives the best combination when the ranked list has a size of 1,000. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 57,
+                        "end": 62,
+                        "text": "Fig.4",
+                        "ref_id": "FIGREF6"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Improving the Effectiveness of Information Retrieval with Clustering and Fusion 123",
+                "sec_num": null
+            },
+            {
+                "text": "Combining multiple retrieval results is certainly a practical technique for improving the overall performance of information retrieval systems. In this paper, we have proposed a novel fusion method that can be combined with document clustering to improve retrieval performance. Our approach consists of three steps. First, we apply clustering to the initial ranked document lists to obtain a list of document clusters. Then, we identify reliable clusters and adjust each ranked list separately using our re-ranking approach. Finally, conventional fusion is carried out to produce an adjusted ranked list.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "5."
+            },
+            {
+                "text": "Since our approach is based on two hypotheses, we first verified them by means of experiments. We also compared our approach with other conventional approaches. The results show that each of them achieves some improvement, and that our approach compares favorably with them. We also investigated the impact of cluster size. We found that our approach is rather stable under variation in the size of clusters.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "5."
+            },
+            {
+                "text": "Although our method showed good performance in our experiments, we believe it still can be improved further. A better clustering algorithm for identifying more reliable clusters and more elaborate formula for re-ranking ranked lists should lead to further improvement. These will be topics for our future work.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "5."
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Automatic Combination of Multiple Ranked Retrieval Systems",
+                "authors": [
+                    {
+                        "first": "B",
+                        "middle": [
+                            "T"
+                        ],
+                        "last": "Bartell",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Cottrell",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Belew",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "Proceedings of the 17th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "173--181",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Bartell,B.T., Cottrell,G.W., and Belew,R.K., \"Automatic Combination of Multiple Ranked Retrieval Systems,\" Proceedings of the 17th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval, 1994, pp. 173-181.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Scatter/gather: A Cluster-based Approach to Browsing Large Document Collections",
+                "authors": [
+                    {
+                        "first": "D",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Cutting",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Karger",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "O"
+                        ],
+                        "last": "Pedersen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Tukey",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "Proceedings of the 15th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "126--135",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "D.R.Cutting, D.R.Karger, J.O.Pedersen, and J.W.Tukey, \"Scatter/gather: A Cluster-based Approach to Browsing Large Document Collections,\" Proceedings of the 15th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval, 1992, pp. 126-135.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Combination of Multiple Searches",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Fox",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Shaw",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "The Second Text Retrieval Conference (TREC2), NIST Special Publication 500-215",
+                "volume": "",
+                "issue": "",
+                "pages": "243--252",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Fox,E. and Shaw,J., \"Combination of Multiple Searches,\" The Second Text Retrieval Conference (TREC2), NIST Special Publication 500-215, 1994, pp. 243-252.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Reexamining the Cluster Hypothesis: Scatter/Gather on Retrieval Results",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Hearst",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "O"
+                        ],
+                        "last": "Pedersen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Proceedings of the 19th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "76--82",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Hearst,M.A., and Pedersen,J.O., \"Reexamining the Cluster Hypothesis: Scatter/Gather on Retrieval Results,\" Proceedings of the 19th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval, 1996, pp. 76-82.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Analyses of Multiple Evidence Combination",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Lee",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the 20th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "267--276",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "J.H.Lee. \"Analyses of Multiple Evidence Combination.,\" Proceedings of the 20th Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval, 1997, pp. 267-276.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "The Best of Both Worlds: Combining Ranked List and Clustering",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Leuski",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Allan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "A.Leuski and J.Allan, \"The Best of Both Worlds: Combining Ranked List and Clustering,\" CIIR Technical Report IR-172, 1999, http://cobar.cs.umass.edu/pubfiles/ir-172.ps.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Improving Interactive Retrieval by Combining Ranked List and Clustering",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Leuski",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Allan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Informations Assistee par Ordinateur = Computer-Assisted Information Retrieval)",
+                "volume": "",
+                "issue": "",
+                "pages": "665--681",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "A.Leuski and J.Allan, \"Improving Interactive Retrieval by Combining Ranked List and Clustering,\" Proceedings of RIAO(Recherche d'Informations Assistee par Ordinateur = Computer-Assisted Information Retrieval) 2000 Conference, 2000, pp. 665-681.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "A Combination of Expert Opinion Approach to Probabilistic Information Retrieval, part I: The Conceptual Model",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Thompson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Information Processing and Management",
+                "volume": "26",
+                "issue": "3",
+                "pages": "371--382",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Thompson,P., \"A Combination of Expert Opinion Approach to Probabilistic Information Retrieval, part I: The Conceptual Model,\" Information Processing and Management, 26(3) 1990, pp. 371-382.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Using Relevance to Train a Linear Mixture of Experts",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Vogt",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [],
+                        "last": "Cottrell",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Belew",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "B",
+                        "middle": [],
+                        "last": "Bartell",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the 5th Text Retrieval Conference (TREC5), NIST Special Publication 500-238",
+                "volume": "",
+                "issue": "",
+                "pages": "503--516",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Vogt,C., Cottrell,G., Belew,R. and Bartell,B., \"Using Relevance to Train a Linear Mixture of Experts,\" Proceedings of the 5th Text Retrieval Conference (TREC5), NIST Special Publication 500-238, 1997, pp. 503-516.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Predicting the Performance of Linearly Combined IR Systems",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Vogt",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [],
+                        "last": "Cottrell",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the 21st Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval",
+                "volume": "",
+                "issue": "",
+                "pages": "190--196",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Vogt,C. and G.Cottrell., \"Predicting the Performance of Linearly Combined IR Systems,\" Proceedings of the 21st Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval, 1998, pp. 190-196.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Fusion Via a Linear Combination of Scores",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Vogt",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [],
+                        "last": "Cottrell",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Information Retrieval",
+                "volume": "1",
+                "issue": "2-3",
+                "pages": "151--173",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Vogt,C. and Cottrell,G., \"Fusion Via a Linear Combination of Scores,\" Information Retrieval, 1(2-3), 1999, pp. 151-173.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "Overview of the Sixth Text Retrieval Conference (TREC-6)",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Voorhees",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Harman",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "NIST Special Publication",
+                "volume": "",
+                "issue": "",
+                "pages": "1--24",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "E.Voorhees, D.Harman, \"Overview of the Sixth Text Retrieval Conference (TREC-6),\" NIST Special Publication 500-240, 1997. pp. 1-24.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Numerical Recipes in C -The Art of Scientific Computing",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Press",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Teukolsky",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "T"
+                        ],
+                        "last": "Vetterling",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "B",
+                        "middle": [
+                            "P"
+                        ],
+                        "last": "Flannery",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Press,W.H., Teukolsky,S.A., Vetterling,W.T., and Flannery,B.P., Numerical Recipes in C - The Art of Scientific Computing, Cambridge University Press, 1992.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "type_str": "figure",
+                "uris": null,
+                "num": null,
+                "text": "shows the basis idea behind our approach. Two clusters (a1 and b1) from different ranked lists that have the largest overlap are identified as reliable clusters."
+            },
+            "FIGREF1": {
+                "type_str": "figure",
+                "uris": null,
+                "num": null,
+                "text": "Clustering results of two ranked lists."
+            },
+            "FIGREF2": {
+                "type_str": "figure",
+                "uris": null,
+                "num": null,
+                "text": "shows our clustering algorithm. The LoopThreshold and ShiftThreshold value were set to 10 in our experiments.Randomly set document i d to cluster j C ; LoopCount =0; ShiftCount = 1000; While (LoopCount < LoopThreshold and ShiftCount > ShiftThreshold) Do Construct the centroid of each cluster, i.e. to its nearest cluster(the distance is determined by the similarity between i d and the centroid of the cluster); ShiftCount = the number of documents shifted to other cluster;LoopCount++; Algorithm for document clustering."
+            },
+            "FIGREF4": {
+                "type_str": "figure",
+                "uris": null,
+                "num": null,
+                "text": "Performance of different approaches."
+            },
+            "FIGREF6": {
+                "type_str": "figure",
+                "uris": null,
+                "num": null,
+                "text": "Impact of cluster size."
+            },
+            "TABREF0": {
+                "text": "Formulas proposed byFox & Shaw.",
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td>Name</td><td colspan=\"4\">Combined Similarity =</td></tr><tr><td>CombMAX</td><td colspan=\"4\">MAX(Individual Similarities)</td></tr><tr><td>CombMIN</td><td colspan=\"4\">MIN(Individual Similarities)</td></tr><tr><td>CombSUM</td><td colspan=\"4\">SUM(Individual Similarities)</td></tr><tr><td>CombANZ</td><td colspan=\"3\">dual SUM(Indivi</td><td>es) Similariti</td></tr><tr><td/><td>Number</td><td>of</td><td colspan=\"2\">Nonzero</td><td>es Similariti</td></tr><tr><td>CombMNZ</td><td colspan=\"4\">SUM(Individual Similarities) * Number of Nonzero</td></tr><tr><td/><td/><td colspan=\"3\">Similarities</td></tr></table>"
+            },
+            "TABREF1": {
+                "text": "Lee observed that fusion works well for result sets that have a high overlap R",
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td colspan=\"3\">112</td><td/><td>J. Zhang et al.</td></tr><tr><td/><td/><td/><td/><td>and a low</td></tr><tr><td colspan=\"2\">N</td><td>overlap</td><td colspan=\"2\">. Inspired by this observation, we also incorporate common R</td></tr><tr><td/><td/><td/><td>A RL 1 . common R</td><td>is the number of common relevant documents in</td><td>A RL and B RL .</td></tr><tr><td colspan=\"2\">N</td><td>common</td><td colspan=\"2\">is the number of common irrelevant documents in</td><td>A RL and B RL .</td></tr><tr><td>1</td><td colspan=\"4\">A RL means ranked list returned by retrieval system A.</td></tr></table>"
+            },
+            "TABREF2": {
+                "text": "Notations.",
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td>Symbol</td><td>Explanation</td></tr></table>"
+            },
+            "TABREF4": {
+                "text": "Characteristics of the TREC-5 Chinese collection.",
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td>Number of docs</td><td>164,811</td></tr><tr><td>Total size (Mega Bytes)</td><td>170</td></tr><tr><td>Average doc length (Characters)</td><td>507</td></tr><tr><td>Number of queries</td><td>28</td></tr><tr><td>Average query length (Characters)</td><td>119</td></tr><tr><td>Average number of relevant docs/query</td><td>93</td></tr></table>"
+            },
+            "TABREF5": {
+                "text": "Average precision of individual retrieval system",
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td>Ranked list</td><td>AvP (11 pt)</td></tr><tr><td>BrklyCH1</td><td>0.3568</td></tr><tr><td>CLCHNA</td><td>0.2702</td></tr><tr><td>Cor5C1vt</td><td>0.3647</td></tr><tr><td>HIN300</td><td>0.1636</td></tr><tr><td>City96c1</td><td>0.3256</td></tr><tr><td>Gmu96ca1</td><td>0.3218</td></tr><tr><td>gmu96cm1</td><td>0.3579</td></tr><tr><td>Average :</td><td>0.3086</td></tr></table>"
+            },
+            "TABREF7": {
+                "text": "Distribution of relevant docs.",
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td>Different kinds of</td><td>Containing</td><td>Containing</td><td>Containing</td><td>Containing</td></tr><tr><td>clusters</td><td>no relevant</td><td>1 relevant</td><td>2-10 relevant</td><td>&gt;10 relevant</td></tr><tr><td/><td>doc</td><td>doc</td><td>docs</td><td>docs</td></tr><tr><td>Percentage of each</td><td/><td/><td/><td/></tr><tr><td>kind of cluster</td><td>38.3%</td><td>15.0%</td><td>35.0%</td><td>7.0%</td></tr><tr><td>Percentage of</td><td/><td/><td/><td/></tr><tr><td>relevant docs</td><td/><td/><td/><td/></tr><tr><td>contained in this kind of cluster</td><td>0%</td><td>3.7%</td><td>35.8%</td><td>60.5%</td></tr><tr><td colspan=\"3\">To test the Fusion Hypothesis, we computed overlap R</td><td>and overlap N</td><td/></tr></table>"
+            },
+            "TABREF8": {
+                "text": "",
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td>overlap R</td><td>and overlap N</td><td colspan=\"3\">values of combination pairs.</td></tr><tr><td colspan=\"2\">Combination pair</td><td>overlap R</td><td>N</td><td>overlap</td></tr><tr><td colspan=\"2\">BrklyCH1 &amp; CLCHNA</td><td>* 0.8542</td><td/><td>0.3398</td></tr><tr><td colspan=\"2\">BrklyCH1 &amp; Cor5C1vt</td><td>* 0.9090</td><td/><td>0.4393</td></tr><tr><td colspan=\"2\">BrklyCH1 &amp; HIN300</td><td>0.4985</td><td/><td>0.2575</td></tr><tr><td colspan=\"2\">BrklyCH1 &amp; City96c1</td><td>* 0.8996</td><td/><td>0.4049</td></tr><tr><td colspan=\"2\">BrklyCH1 &amp; Gmu96ca1</td><td>* 0.8784</td><td/><td>0.3259</td></tr><tr><td colspan=\"2\">BrklyCH1 &amp; gmu96cm1</td><td>* 0.8871</td><td/><td>0.3292</td></tr><tr><td colspan=\"2\">CLCHNA &amp; Cor5C1vt</td><td>* 0.8728</td><td/><td>0.4118</td></tr><tr><td colspan=\"2\">CLCHNA &amp; HIN300</td><td>0.4652</td><td/><td>0.2172</td></tr><tr><td colspan=\"2\">CLCHNA &amp; City96c1</td><td>* 0.8261</td><td/><td>0.2668</td></tr><tr><td colspan=\"2\">CLCHNA &amp; Gmu96ca1</td><td>* 0.8447</td><td/><td>0.3090</td></tr><tr><td colspan=\"2\">CLCHNA &amp; gmu96cm1</td><td>* 0.8585</td><td/><td>0.3412</td></tr><tr><td colspan=\"2\">Cor5C1vt &amp; HIN300</td><td>0.4961</td><td/><td>0.2392</td></tr><tr><td colspan=\"2\">Cor5C1vt &amp; City96c1</td><td>* 0.8763</td><td/><td>0.2943</td></tr><tr><td colspan=\"2\">Cor5C1vt &amp; Gmu96ca1</td><td>* 0.9193</td><td/><td>0.4742</td></tr><tr><td colspan=\"2\">Cor5C1vt &amp; gmu96cm1</td><td>* 0.9185</td><td/><td>0.4525</td></tr><tr><td colspan=\"2\">HIN300 &amp; City96c1</td><td>0.4813</td><td/><td>0.1555</td></tr><tr><td colspan=\"2\">HIN300 &amp; Gmu96ca1</td><td>0.4636</td><td/><td>0.1854</td></tr><tr><td colspan=\"2\">HIN300 &amp; gmu96cm1</td><td>0.4701</td><td/><td>0.2004</td></tr><tr><td colspan=\"2\">City96c1 &amp; Gmu96ca1</td><td>* 0.8698</td><td/><td>0.2854</td></tr><tr><td colspan=\"2\">City96c1 &amp; gmu96cm1</td><td>* 0.8860</td><td/><td>0.3005</td></tr><tr><td colspan=\"2\">Gmu96ca1 &amp; gmu96cm1</td><td>* 0.9687</td><td/><td>0.8064</td></tr><tr><td>Average</td><td/><td>0.7688</td><td/><td>0.3351</td></tr></table>"
+            },
+            "TABREF9": {
+                "text": "Average precision of each combination pair.",
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td>Combination pair</td><td>Comb MAX</td><td>Comb SUM</td><td>Comb MNZ</td><td>Our Approach (Cluster size=100)</td></tr><tr><td>BrklyCH1 &amp; CLCHNA</td><td>0.3401</td><td>0.3627</td><td>0.3549</td><td>* 0.3755</td></tr><tr><td>BrklyCH1 &amp; Cor5C1vt</td><td>0.3832</td><td>0.3976</td><td>0.3961</td><td>* 0.4107</td></tr><tr><td>BrklyCH1 &amp; HIN300</td><td>0.3560</td><td>0.3243</td><td>0.2618</td><td>0.3107</td></tr><tr><td>BrklyCH1 &amp; city96c1</td><td>0.3650</td><td>0.3833</td><td>0.3856</td><td>* 0.3912</td></tr><tr><td>BrklyCH1 &amp; gmu96ca1</td><td>0.3753</td><td>0.4028</td><td>0.3999</td><td>* 0.4022</td></tr><tr><td>BrklyCH1 &amp; gmu96cm1</td><td>0.3979</td><td>0.4234</td><td>0.4201</td><td>* 0.4243</td></tr><tr><td>CLCHNA &amp; Cor5C1vt</td><td>0.3434</td><td>0.3560</td><td>0.3492</td><td>* 0.3707</td></tr><tr><td>CLCHNA &amp; HIN300</td><td>0.2746</td><td>0.2478</td><td>0.2154</td><td>0.2579</td></tr><tr><td>CLCHNA &amp; city96c1</td><td>0.3007</td><td>0.3459</td><td>0.3573</td><td>* 0.3931</td></tr><tr><td>CLCHNA &amp; gmu96ca1</td><td>0.3269</td><td>0.3667</td><td>0.3634</td><td>* 0.3690</td></tr><tr><td>CLCHNA &amp; gmu96cm1</td><td>0.3555</td><td>0.3864</td><td>0.3783</td><td>* 0.3883</td></tr><tr><td>Cor5C1vt &amp; HIN300</td><td>0.3778</td><td>0.3081</td><td>0.2520</td><td>0.3139</td></tr><tr><td>Cor5C1vt &amp; city96c1</td><td>0.3709</td><td>0.4091</td><td>0.4104</td><td>* 0.4285</td></tr><tr><td>Cor5C1vt &amp; gmu96ca1</td><td>0.3568</td><td>0.3684</td><td>0.3676</td><td>* 0.3724</td></tr><tr><td>Cor5C1vt &amp; gmu96cm1</td><td>0.3831</td><td>0.3926</td><td>0.3911</td><td>* 0.3975</td></tr><tr><td>HIN300 &amp; city96c1</td><td>0.2616</td><td>0.2565</td><td>0.2444</td><td>0.3036</td></tr><tr><td>HIN300 &amp; gmu96ca1</td><td>0.3466</td><td>0.2942</td><td>0.2464</td><td>0.2954</td></tr><tr><td>HIN300 &amp; gmu96cm1</td><td>0.3764</td><td>0.3205</td><td>0.2613</td><td>0.3150</td></tr><tr><td>city96c1 &amp; gmu96ca1</td><td>0.3310</td><td>0.3764</td><td>0.3854</td><td>* 0.3939</td></tr><tr><td>city96c1 &amp; gmu96cm1</td><td>0.3595</td><td>0.3970</td><td>0.4047</td><td>* 0.4090</td></tr><tr><td>gmu96ca1 &amp; gmu96cm1</td><td>0.3451</td><td>0.3514</td><td>0.3511</td><td>* 0.3505</td></tr><tr><td>Average:</td><td>0.3489</td><td>0.3557</td><td>0.3426</td><td>0.3654</td></tr></table>"
+            },
+            "TABREF10": {
+                "text": "Impact of cluster size.",
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td>Size of Cluster</td><td>200</td><td>100</td><td>50</td><td>25</td><td>10</td><td>5</td></tr><tr><td>11pt AvP</td><td colspan=\"6\">0.3621 0.3654 0.3661 0.3675 0.3668 0.3661</td></tr></table>"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-3001.json

@@ -0,0 +1,869 @@
+{
+    "paper_id": "O01-3001",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:32.971537Z"
+    },
+    "title": "Metaphorical Transfer and Pragmatic Strengthening 1 : On the Development of V-diao in Mandarin *",
+    "authors": [
+        {
+            "first": "Wei-Lun",
+            "middle": [],
+            "last": "Louis",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Chengkung University",
+                "location": {
+                    "settlement": "Tainan",
+                    "country": "Taiwan"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "",
+            "middle": [],
+            "last": "Lu",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Chengkung University",
+                "location": {
+                    "settlement": "Tainan",
+                    "country": "Taiwan"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Louis",
+            "middle": [
+                "W L"
+            ],
+            "last": "Lu",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Taiwan University",
+                "location": {
+                    "settlement": "Taipei",
+                    "country": "Taiwan"
+                }
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "In this synchronic study, I shall adopt a corpus-based approach to investigate the semantic change of V-diao in Mandarin. Semantically, V-diao constructions fall into three categories: A) Physical disappearance from its original position, with the V slot filled by physical verbs, such as tao-diao \"escape,\" diu-diao \"throw away,\" and so on. B) Disappearance from a certain conceptual domain, rather than from the physical space, with the V slot filled by less physically perceivable verbs, such as jie-diao \"quit,\" wang-diao \"forget,\" and the like. C) The third category of V-diao involves the speaker's subjective, always negative, attitude toward the result. Examples include: lan-diao \"rot,\" ruan-diao \"soften,\" huang-diao \"yellow,\" and so forth.",
+    "pdf_parse": {
+        "paper_id": "O01-3001",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "In this synchronic study, I shall adopt a corpus-based approach to investigate the semantic change of V-diao in Mandarin. Semantically, V-diao constructions fall into three categories: A) Physical disappearance from its original position, with the V slot filled by physical verbs, such as tao-diao \"escape,\" diu-diao \"throw away,\" and so on. B) Disappearance from a certain conceptual domain, rather than from the physical space, with the V slot filled by less physically perceivable verbs, such as jie-diao \"quit,\" wang-diao \"forget,\" and the like. C) The third category of V-diao involves the speaker's subjective, always negative, attitude toward the result. Examples include: lan-diao \"rot,\" ruan-diao \"soften,\" huang-diao \"yellow,\" and so forth.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "V-diao is traditionally termed a resultative compound, indicating the result of an action [Li and Thompson 1981] . However, a close examination of linguistic data indicates that the semantics of V-diao cannot be calculated by simply putting its components together. In this paper, I shall focus on the semantics of diao and try to tackle V-diao at a lexical level to see whether such lexical analysis works.",
+                "cite_spans": [
+                    {
+                        "start": 90,
+                        "end": 112,
+                        "text": "[Li and Thompson 1981]",
+                        "ref_id": "BIBREF8"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Semantic Classification of V-diao",
+                "sec_num": "1."
+            },
+            {
+                "text": "The V-diao construction comprises a verb (be it action or stative) and a verbal suffix -diao. It gives the final state of the agent, if used intransitively, and of the receiver of the action, in transitive cases. It may represent: A) physical disappearance of an entity from its original position, B) disappearance from a certain conceptual domain, and C) the speaker's subjective evaluation of the result of an event, as in (1)-(3), respectively:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Semantic Classification of V-diao",
+                "sec_num": "1."
+            },
+            {
+                "text": "(1) ta qiaoqiao pao-diao le he quietly run away CRS \"He ran away quietly.\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Semantic Classification of V-diao",
+                "sec_num": "1."
+            },
+            {
+                "text": "(2) ta jie-diao le nage huai xiguan he get rid of Perf that bad habit \"He got rid of that bad habit.\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Semantic Classification of V-diao",
+                "sec_num": "1."
+            },
+            {
+                "text": "(3) diennau zuotien huai-diao le computer yesterday break down CRS \"The computer broke down yesterday.\" I shall begin this paper with a close look at the semantics of the foregoing types of V-diao, especially the last one. This is because the Type C construction involves an intriguing phenomenon: interpretation of a negative result cannot be arrived at by directly adding the suffix -diao to any verb. It is worth noting that, synchronically, the semantics of diao denote a downward movement. It is, thus, reasonable to claim that the negative interpretation may derive from the human experiential basis of space.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Semantic Classification of V-diao",
+                "sec_num": "1."
+            },
+            {
+                "text": "On the Development of V-diao in Mandarin 3",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Semantic Classification of V-diao",
+                "sec_num": "1."
+            },
+            {
+                "text": "It is reported that a suffix in a resultative verb compound in Mandarin indicates the sequel of an action [Li and Thompson 1981] . The first kind of -diao gives the final state, i.e., physical absence, of the agent or the patient. This kind of -diao is mostly affixed to easily perceivable physical action verbs such as pao \"run,\" as in (1), diu \"throw,\" shao \"burn,\" and so on.",
+                "cite_spans": [
+                    {
+                        "start": 106,
+                        "end": 128,
+                        "text": "[Li and Thompson 1981]",
+                        "ref_id": "BIBREF8"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Type A: Physical Disappearance",
+                "sec_num": "1.1"
+            },
+            {
+                "text": "The second sort of V-diao also denotes the result of an action. However, this differs from type A in the sense that it represents a less \"concrete\" disappearance. It is often attached to low transitive verbs, without obvious physical motion, and accompanies an abstract noun phrase. Consider example (2) again:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Type B: Disappearance from a Conceptual Domain",
+                "sec_num": "1.2"
+            },
+            {
+                "text": "(2) ta jie-diao le nage huai xiguan he get rid of Perf that bad habit \"He got rid of that bad habit.\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Type B: Disappearance from a Conceptual Domain",
+                "sec_num": "1.2"
+            },
+            {
+                "text": "A bad habit is an abstract entity. The abandonment of it by the agent is almost physically undetectable. But how can one perceive its existence and absence? Also, from where does the habit disappear?",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Type B: Disappearance from a Conceptual Domain",
+                "sec_num": "1.2"
+            },
+            {
+                "text": "This has everything to do with our conceptual system. We experience many things, through sight and touch, as having distinct physical shapes and boundaries. We thus tend to project physical shapes and boundaries on them, conceptualising them as entities and imposing on them physical characteristics such as existence and disappearance, even though we can never really feel them with our hands or sense them with our eyes or nose [Lakoff and Johnson 1980] . Further details concerning Type B and metaphorical transfer will be addressed in the next section.",
+                "cite_spans": [
+                    {
+                        "start": 430,
+                        "end": 455,
+                        "text": "[Lakoff and Johnson 1980]",
+                        "ref_id": "BIBREF7"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Type B: Disappearance from a Conceptual Domain",
+                "sec_num": "1.2"
+            },
+            {
+                "text": "In this case, a habit is conceptualised as a physical entity. It can fade out, can be done away with, and can finally disappear from our conceptual domain as physical things do from a physical space. Thus, Type B seems to represent the final state of, usually, a non-physical action, i.e., an abstract entity being done away with, finally disappearing from one's conceptual domain.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Type B: Disappearance from a Conceptual Domain",
+                "sec_num": "1.2"
+            },
+            {
+                "text": "Type C V-diao denotes a somewhat negative evaluation of the result in question. It often co-occurs with verbs with negative connotation, such as lan-diao \"rot,\" si-diao \"die,\" shu-diao \"lose,\" etc. However, its negative meaning does not seem to come from the preceding verb in every case. Consider the following instances (4) and ( \"Vegetables won't be fresh if they turn yellow.\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Type C: Evaluative Function from the Speaker",
+                "sec_num": "1.3"
+            },
+            {
+                "text": "In (4) and (5), the words huang \"yellow\" and ruan \"soft\" do not themselves carry negative meanings, but the entire phrase clearly involve one's unfavourable attitude toward the final state of the vegetables and cookies. In the following sections, I shall examine the semantic change of -diao and try to account for the emergence of its unfavourable interpretation.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Type C: Evaluative Function from the Speaker",
+                "sec_num": "1.3"
+            },
+            {
+                "text": "Two main sources provide examples discussed to illuminate this search. The written source mostly comes from the Academia Sinica Corpus, with a complete tagging system. The spoken source comprises the Taida Spoken Corpus, together with another eight hours of transcribed data 2 . The spoken part amounts to an entire length of sixteen hours of conversational Mandarin. In sum, we collected a total of one hundred and eighty-nine tokens of -diao, excluding its use as a main verb such as xiao-diao-da-ya ( ), diao-tao ( ), and so on. Also, when our argument called for constructed examples, native speakers, inclusive of the author himself, were consulted. Two interesting observations on the corpora are left unaddressed due to the limited scope of the current study. First, the approximate portion of main verbs is much higher in our written corpus than that in our spoken corpus (around 4:1). Second, the development of -diao seems to match the tendency of subjectification proposed by Traugott [1989 Traugott [ , 1995 . However, these issues are not closely related to the current study and will, thus, be left out of this research.",
+                "cite_spans": [
+                    {
+                        "start": 987,
+                        "end": 1001,
+                        "text": "Traugott [1989",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 1002,
+                        "end": 1019,
+                        "text": "Traugott [ , 1995",
+                        "ref_id": "BIBREF11"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Data and Methodology",
+                "sec_num": "1.4"
+            },
+            {
+                "text": "It is argued that, when a grammatical meaning is derived from its source, there often exists a metaphorical relation between the two meanings [Sweetser, 1990; Bybee, Perkins and Pagliuca, 1994] . Such a semantic change takes place to serve a certain functional end in grammar and discourse, as indicated by Heine, Claudi and Hunnemeyer [1991:48] :",
+                "cite_spans": [
+                    {
+                        "start": 142,
+                        "end": 158,
+                        "text": "[Sweetser, 1990;",
+                        "ref_id": "BIBREF9"
+                    },
+                    {
+                        "start": 159,
+                        "end": 193,
+                        "text": "Bybee, Perkins and Pagliuca, 1994]",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 307,
+                        "end": 345,
+                        "text": "Heine, Claudi and Hunnemeyer [1991:48]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Transfer",
+                "sec_num": "2."
+            },
+            {
+                "text": "We try to demonstrate that metaphorical transfer forms one of the main driving forces in the 2 The second source of spoken data was offered by Dr. Lily I-wen Su.",
+                "cite_spans": [
+                    {
+                        "start": 93,
+                        "end": 94,
+                        "text": "2",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Transfer",
+                "sec_num": "2."
+            },
+            {
+                "text": "On the Development of V-diao in Mandarin 5 development of grammatical categories; that is, in order to express more \"abstract\" functions, concrete entities are recruited.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Transfer",
+                "sec_num": "2."
+            },
+            {
+                "text": "The above corresponds to my observations of V-diao: a metaphorical transfer takes place when meaning proceeds from the physical domain to a conceptual domain, denoting metaphorical disappearance.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Transfer",
+                "sec_num": "2."
+            },
+            {
+                "text": "The above claim seems to be verified in the development of -diao. The meaning of Type A is the most concrete and physical one, since it indicates a salient result after some physical action is carried out. Type B, on the other hand, denotes disappearance from our mental space instead of from a physical space. Now consider (6) a typical instance of such metaphorical transfer: 6 Pao-bu-diao in (6a) denotes the unsuccessful outcome of the agent's escape. The agent fails to escape and does not disappear. In (6b), the meaning is that the landmark \"a hundred thousand\" is certain to be met. However, not every single case of Type B has a counterpart in A. Actually, most Type B constructions do not. Pao-bu-diao is simply a case employed to illustrate the metaphorical relation of the polysemy between Type A and B. In most cases of Type B V-diao, the V slot is filled by less physical verbs, such as jie \"get rid of\" in (2), hulue \"ignore,\" wang \"forget,\" and so on.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type A to Type B: Metaphor at Work",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "In this section, I have shown that the physical \"resultative compound\" V-diao has undergone a metaphorical transfer and developed the sense of disappearance from a conceptual domain. Thus, it makes perfect sense to conclude that the polysemy in this case is at least partly contributed by metaphor, since disappearance is a common feature of Types A and B. The following figure indicates the mapping relation between Type A and Type B: ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "Type A (source domain) Type B (target domain) mapping relation disappearance of ----------------------------------------> disappearance of physical shape conceptual shape",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "Other than metaphor, pragmatic strengthening is claimed to be a major mechanism of semantic change [Hopper and Traugott, 1993; Bybee, Perkins and Pagliuca, 1994] . In such changes, context plays a crucial role. Frequent use of a grammatical or lexical unit in a particular context may lead to the inference that the context is an incorporated part of its meaning. Goossens' research on Old English modals [1982] indicates that there rarely are \"real\" epistemic markers in OE, and that possibility markers frequently combine with adverbs to express epistemic functions. That is, speakers can generalise and extract the epistemic meanings from the context and impose them on modals. This suggests that frequent co-occurrence with a particular context may \"colour\" the semantics of a grammatical unit.",
+                "cite_spans": [
+                    {
+                        "start": 99,
+                        "end": 126,
+                        "text": "[Hopper and Traugott, 1993;",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 127,
+                        "end": 161,
+                        "text": "Bybee, Perkins and Pagliuca, 1994]",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 405,
+                        "end": 411,
+                        "text": "[1982]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pragmatic Strengthening",
+                "sec_num": "3."
+            },
+            {
+                "text": "In this section, I will demonstrate that the final stage of development of V-diao is based on such a mechanism. Now let us see how language use and context collaborate to produce semantic change in the case of V-diao.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pragmatic Strengthening",
+                "sec_num": "3."
+            },
+            {
+                "text": "In Type C -diao, the sense of disappearance is retained, but there seems to exist something more than the combination of the verbal sense and disappearance. In general, these phrases involve unfavorable assessment on the part of the speaker. That is, the speaker obviously does not favour the change of state.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type B to Type C: Semanticisation of Context",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "It is noteworthy that Type C can be further divided into two subtypes based on the verb in the V slot: 1) verbs with negative connotation, such as lan \"rot,\" si \"die,\" po \"break,\" shu \"lose,\" and so on; 2) neutral verbs, such as huang \"yellow,\" ya \"croak,\" ruan \"soft,\" and so on. This classification highly pertains to the semantic change addressed in the current research. Let us see how.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type B to Type C: Semanticisation of Context",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "Initially, only the former combinations are formed. They simply denote a metaphorical disappearance, labeled Type B. As the frequency of use increases, the speakers tend to associate the construction with the adverse image related to negative verbs. Such frequent collocation of negative verbs and -diao may invite the generalisation that the suffix is applied to express one's unfavourable appraisal of the situation at issue. The context is, thus, \"semanticized\" [Hopper and Traugott, 1993:75] and is transferred onto -diao. Consequently, the construction may accommodate neutral stative verbs in \"Vegetables won't be fresh if they turn yellow.\"",
+                "cite_spans": [
+                    {
+                        "start": 465,
+                        "end": 495,
+                        "text": "[Hopper and Traugott, 1993:75]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type B to Type C: Semanticisation of Context",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "Huang and ruan themselves do not signal negativity. The adverse meaning is subtly signalled and triggered by the repetitive occurrence of negative verbs in the position. In other words, the emergence of the speaker's negative attitude derives neither from the suffix denoting disappearance, nor from the verb preceding it, but could have been generalised from the constant collocation of negative words and -diao. Now, even neutral verbs may fit into the V slot and yield negative assessment. However, no positive verbs may combine with -diao. Details of this co-occurrence restriction will be given in the next section.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "From Type B to Type C: Semanticisation of Context",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "Pragmatic strengthening is one of the driving forces of semantic change, and I have proven that it plays a crucial role in the development of V-diao as well. First, only verbs that result in physical and conceptual disappearance may occur in the construction. Among them, a group of verbs with negative connotation prompt the deduction of negative connotation. Consequently, the negative sense of the verb is transferred to the entire phrase, resulting in the speaker's unfavorable appraisal of the result. The following figure illustrates the development path from Type B to Type C:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "Type B Type C negative meaning meaning is transferred comes from the context -- ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "-----------------------------------> to diao",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "As the polysemy of V-diao develop, its use broaden to increasingly wider contexts. At first, it only accommodates physical verbs and denotes physical disappearance. It then proceeds to tolerate less physical verbs and metaphorically allows a sense of conceptual disappearance. Finally, it may be applied to a variety of stative verbs to express the speaker's attitude. Nevertheless, in spite of its seemingly free occurrence, some restrictions still exist. Consider the following pairs for the purpose of illustration: 7 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conceptual Structure and Selectional Restriction",
+                "sec_num": "4."
+            },
+            {
+                "text": "I have argued for metaphor as the driving force of semantic change in the development of V-diao. The metaphorical transfer discussed in section two must obey the orientational metaphor GOOD IS UP; BAD IS DOWN proposed by Lakoff and Johnson [1980:16] :",
+                "cite_spans": [
+                    {
+                        "start": 221,
+                        "end": 249,
+                        "text": "Lakoff and Johnson [1980:16]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "Physical basis for personal well-being: Happiness, health, life, and control-the things that principally characterize what is good for a person-are all UP. Also, C. R. Huang's previous studies on Mandarin -qilai constructions indicate that the development of grammatical units cannot contradict the metaphor that they are based on, and that the collocations of -qilai and verbs are conceptually restricted on a semantic basis [Chang 1994, Huang and Chang 1996] . The following observations concerning V-diao correspond to this claim.",
+                "cite_spans": [
+                    {
+                        "start": 426,
+                        "end": 448,
+                        "text": "[Chang 1994, Huang and",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 449,
+                        "end": 460,
+                        "text": "Chang 1996]",
+                        "ref_id": "BIBREF6"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "The physical and experiential basis for DOWN IS BAD is also evident in our language use and conceptual system. Synchronically, the most basic meaning of diao is physical dropping / falling, signaling downward movement. It follows that diao can relate to something bad in our conceptual system. Whether it is grammaticalised or not, diao should never override the conceptual restriction to modify something good. In other words, if the metaphor DOWN IS BAD is truly at work, it seems On the Development of V-diao in Mandarin 9 rather natural for V-diao not to accommodate a verb with positive connotation. Thus, the conceptual / cognitive restriction can fully account for the intrinsic incompatibility of positive verbs with V-diao.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "The above semantic restriction is critical in the development from Type B to Type C V-diao; without it, later unfolding would be impossible. Language users generalise the negative meaning of -diao from a previous existing pattern. The constraint must have existed prior to the semanticisation of context. Otherwise, without such a selectional restriction, the meaning would fail to emerge, since positive verbs would intervene. Therefore, it is safe to say that this constraint metaphorically shapes, or at least partly contributes to, the semantic shift of V-diao.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Metaphorical Basis of Selectional Restriction",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "In this section, the incompatibility of positive verbs with -diao has been explored from a semantic viewpoint. The meaning of diao conceptually constrains the verbs it co-occurs with, which proves the metaphorical nature of our conceptual system. Also, this selectional restriction results in the existing pattern, which in turn results in the negative meaning of -diao. This metaphorical condition, thus, reflects interaction between the grammar and conceptual system.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Summary",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "In this study, I have classified V-diao constructions according to their semantics. In the second section, metaphorical transfer has been proposed as an important mechanism involved in the development of V-diao. Further, I have discussed how pragmatic strengthening enables language users to arrive at the negative meaning of -diao. Figure 3 shows different stages of V-diao and the change of mechanism.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 333,
+                        "end": 341,
+                        "text": "Figure 3",
+                        "ref_id": "FIGREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "5."
+            },
+            {
+                "text": "Finally, I have shown that a selectional restriction on the V slot exists. The exclusion of positive verbs is conceptually conditioned by the semantics of diao. This suggests that the semantic change and grammaticalisation process of a grammatical unit is conditioned by human experiential basis. Hopefully, this study will serve as a valid argument for the interaction between our language use and grammar, and for a conceptual basis of human language.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "5."
+            },
+            {
+                "text": "(metaphor) conceptual (strengthening) negative evaluation ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "TYPE A -------------------------> TYPE B -----------------------------> TYPE C Physical",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of the World",
+                "authors": [
+                    {
+                        "first": "Joan",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Bybee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Revere",
+                        "middle": [],
+                        "last": "Perkins",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "William",
+                        "middle": [],
+                        "last": "Pagliuca",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Bybee, Joan L., Revere Perkins, and William Pagliuca. 1994. The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of the World. Chicago: The University of Chicago Press.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "V-qi-lai Constructions in Mandarin Chinese: A Study of Their Semantics and Syntax",
+                "authors": [
+                    {
+                        "first": "Shen-Ming",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chang, Shen-ming. 1994. V-qi-lai Constructions in Mandarin Chinese: A Study of Their Semantics and Syntax. M. A. Thesis. National Tsing Hua University.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Regularity and Idiomaticity in Grammatical Constructions: The Case of Let Alone",
+                "authors": [
+                    {
+                        "first": "Charles",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Fillmore",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Paul",
+                        "middle": [],
+                        "last": "Kay",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Mary",
+                        "middle": [],
+                        "last": "Catherine",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "O'",
+                        "middle": [],
+                        "last": "Connor",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1988,
+                "venue": "Language",
+                "volume": "64",
+                "issue": "",
+                "pages": "501--539",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Fillmore, Charles J., Paul Kay, and Mary Catherine O'Connor. 1988. \"Regularity and Idiomaticity in Grammatical Constructions: The Case of Let Alone.\" Language 64:501-38",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "On the Development of the Modals and of the Epistemic Functions in English",
+                "authors": [
+                    {
+                        "first": "Louis",
+                        "middle": [],
+                        "last": "Goossens",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1982,
+                "venue": "Papers from the Fifth International Conference on Historical Linguistics",
+                "volume": "",
+                "issue": "",
+                "pages": "74--84",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Goossens, Louis. 1982. \"On the Development of the Modals and of the Epistemic Functions in English.\" Papers from the Fifth International Conference on Historical Linguistics, ed. by Anders Ahlqvist, 74-84. Amsterdam: Benjamins.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "From Cognition to Grammar --Evidence from African Languages",
+                "authors": [
+                    {
+                        "first": "Bernd",
+                        "middle": [],
+                        "last": "Heine",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ulrike",
+                        "middle": [],
+                        "last": "Claudi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Friederike",
+                        "middle": [],
+                        "last": "Hunnemeyer",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "",
+                "volume": "1",
+                "issue": "",
+                "pages": "149--87",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Heine, Bernd, Ulrike Claudi, and Friederike Hunnemeyer. 1991. \"From Cognition to Grammar -- Evidence from African Languages.\" Approaches to Grammaticalization. eds. by Traugott and Heine, Vol. 1, 149-87.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Metaphor, Metaphorical Extension, and Grammaticalization: A Study of Mandarin Chinese -qilai",
+                "authors": [
+                    {
+                        "first": "Chu",
+                        "middle": [
+                            "-"
+                        ],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Shen-Ming",
+                        "middle": [],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Conceptual Structure, Discourse, and Language",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Huang, Chu-ren and Shen-ming Chang. 1996. \"Metaphor, Metaphorical Extension, and Grammaticalization: A Study of Mandarin Chinese -qilai.\" Conceptual Structure, Discourse, and Language. ed., by Adele Goldberg. CSLI.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Metaphors We Live by",
+                "authors": [
+                    {
+                        "first": "George",
+                        "middle": [],
+                        "last": "Lakoff",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Mark",
+                        "middle": [],
+                        "last": "Johnson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1980,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lakoff, George, and Mark Johnson. 1980. Metaphors We Live by. Chicago: University of Chicago Press.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Mandarin Chinese: A Functional Reference Grammar",
+                "authors": [
+                    {
+                        "first": "Charles",
+                        "middle": [],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Sandra",
+                        "middle": [],
+                        "last": "Thompson",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1981,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Li, Charles, and Sandra Thompson. 1981. Mandarin Chinese: A Functional Reference Grammar. Los Angeles: University of California Press.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure",
+                "authors": [
+                    {
+                        "first": "Eve",
+                        "middle": [
+                            "Eliot"
+                        ],
+                        "last": "Sweetser",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "Cambrige",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Sweetser, Eve Eliot. 1990. From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambrige: Cambridge University Press.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "On the Rise of Epistemic Meanings in English: An Example of Subjectification in Semantic Change",
+                "authors": [
+                    {
+                        "first": "Elizabeth",
+                        "middle": [],
+                        "last": "Traugott",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Closs",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1989,
+                "venue": "Language",
+                "volume": "65",
+                "issue": "",
+                "pages": "31--55",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Traugott, Elizabeth Closs. 1989. \"On the Rise of Epistemic Meanings in English: An Example of Subjectification in Semantic Change.\" Language 65:31-55.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Subjectification in Grammaticalisation",
+                "authors": [
+                    {
+                        "first": "Elizabeth",
+                        "middle": [],
+                        "last": "Traugott",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Closs",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "31--55",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Traugott, Elizabeth Closs. 1995. \"Subjectification in Grammaticalisation.\" Subjectivity and Subjectivisation, eds. by Dieter Stein and Susan Wright, 31-55. Cambridge: Cambridge University Press.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "type_str": "figure",
+                "text": "Metaphorical Transfer Between Types A and B V-diao",
+                "uris": null,
+                "num": null
+            },
+            "FIGREF1": {
+                "type_str": "figure",
+                "text": "Semanticisation of the Context in V-diao",
+                "uris": null,
+                "num": null
+            },
+            "FIGREF2": {
+                "type_str": "figure",
+                "text": "Different Stages of V-diao and Change of Mechanism",
+                "uris": null,
+                "num": null
+            },
+            "TABREF2": {
+                "type_str": "table",
+                "html": null,
+                "text": "On the Development of V-diao in Mandarin 7 the V slot and still gain a negative interpretation. See (4) and (5) again for the purpose of illustration:",
+                "content": "<table><tr><td>(4)</td><td>binggan</td><td colspan=\"2\">ruan-diao jiu</td><td>bu</td><td>hauchi</td><td>le</td></tr><tr><td/><td>cookie</td><td>soften</td><td>PARTICLE</td><td>not</td><td>tasty</td><td>CRS</td></tr><tr><td/><td colspan=\"4\">\"Cookies won't taste good if they become soft.\"</td><td/></tr><tr><td>(5)</td><td>cai</td><td colspan=\"2\">huang-diao jiu</td><td>bu</td><td>xinxien</td><td>le</td></tr><tr><td/><td colspan=\"2\">vegetable yellow</td><td>PARTICLE</td><td>not</td><td>fresh</td><td>CRS</td></tr></table>",
+                "num": null
+            },
+            "TABREF3": {
+                "type_str": "table",
+                "html": null,
+                "text": "pairs, it is evident that the V slot does not allow verbs with positive connotation. It seems that the semantics of positive verbs clashes with that of the entire construction. Why is this the case? What is basis of this selectional restriction?",
+                "content": "<table><tr><td>a. wo</td><td colspan=\"2\">zhengge</td><td>ren</td><td>sha-diao</td><td>le</td></tr><tr><td>I</td><td>entire</td><td/><td>person</td><td>dumb-Suffix</td><td>CRS</td></tr><tr><td colspan=\"4\">\"I was entirely stunned.\"</td><td/></tr><tr><td>b. *wo</td><td colspan=\"3\">congming-diao</td><td>le</td></tr><tr><td>I</td><td colspan=\"2\">smart-Suffix</td><td/><td>CRS</td></tr><tr><td>(8) a. dongxi</td><td/><td colspan=\"2\">langfei-diao</td><td>le</td></tr><tr><td>thing</td><td/><td colspan=\"2\">waste-Suffix</td><td>CRS</td></tr><tr><td colspan=\"4\">\"The thing is wasted.\"</td><td/></tr><tr><td colspan=\"2\">b. *dongxi</td><td colspan=\"2\">zhenxi-diao</td><td>le</td></tr><tr><td>thing</td><td/><td colspan=\"2\">cherish-Suffix</td><td>CRS</td></tr><tr><td colspan=\"2\">From the above</td><td/><td/><td/></tr></table>",
+                "num": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-3002.json

@@ -0,0 +1,1450 @@
+{
+    "paper_id": "O01-3002",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:09:36.520130Z"
+    },
+    "title": "A Simple Method for Chinese Video OCR and Its Application to Question Answering",
+    "authors": [
+        {
+            "first": "Chuan-Jie",
+            "middle": [],
+            "last": "Lin",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Taiwan University",
+                "location": {
+                    "settlement": "Taipei",
+                    "country": "TAIWAN, R.O.C"
+                }
+            },
+            "email": "cjlin@nlg2.csie.ntu.edu.tw"
+        },
+        {
+            "first": "Che-Chia",
+            "middle": [],
+            "last": "Liu",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Taiwan University",
+                "location": {
+                    "settlement": "Taipei",
+                    "country": "TAIWAN, R.O.C"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Hsin-Hsi",
+            "middle": [],
+            "last": "Chen",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Taiwan University",
+                "location": {
+                    "settlement": "Taipei",
+                    "country": "TAIWAN, R.O.C"
+                }
+            },
+            "email": "hh_chen@csie.ntu.edu.tw"
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "Captions in videos contain valuable information for video retrieval. Although texts in captions can be obtained easily in the new image compression formats like MPEG2, there still are many video programs encoded in older formats. Thus, video OCR is indispensable for content-based video retrieval. This paper proposes a simple video OCR method for Chinese captions, including image capturing, caption region deciding, background removing, character segmentation, OCR and post-processing. We employed Discovery Channel films as training and testing corpus. In an outside test, the accuracy of the video OCR was 84.1%. The hardware used in the experiment consisted of a computer with a P4-1.7G CPU, 256MB RAM and a 40G, 7200rpm hard disk. On average, it took 29 minutes and 11 seconds to process a film 495MB in size. We also applied the results of video OCR to video retrieval and question answering.",
+    "pdf_parse": {
+        "paper_id": "O01-3002",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "Captions in videos contain valuable information for video retrieval. Although texts in captions can be obtained easily in the new image compression formats like MPEG2, there still are many video programs encoded in older formats. Thus, video OCR is indispensable for content-based video retrieval. This paper proposes a simple video OCR method for Chinese captions, including image capturing, caption region deciding, background removing, character segmentation, OCR and post-processing. We employed Discovery Channel films as training and testing corpus. In an outside test, the accuracy of the video OCR was 84.1%. The hardware used in the experiment consisted of a computer with a P4-1.7G CPU, 256MB RAM and a 40G, 7200rpm hard disk. On average, it took 29 minutes and 11 seconds to process a film 495MB in size. We also applied the results of video OCR to video retrieval and question answering.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "In the new information era, multimedia is widely used, and the amount of existing video data is huge. How to extract the content of video data for further application has become an important issue. The well-known project \"Informedia\" [Wactlar, 2000] in digital library is a typical example. Captions in videos contain valuable information for video retrieval. Although texts in captions can be easily obtained in the new image compression formats like MPEG2, there still are many video programs encoded in older formats. Thus, video OCR is indispensable for content-based video retrieval. This paper proposes a simple video OCR method for Chinese captions and demonstrates its application in video search and question answering.",
+                "cite_spans": [
+                    {
+                        "start": 234,
+                        "end": 249,
+                        "text": "[Wactlar, 2000]",
+                        "ref_id": "BIBREF13"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "OCR research started very early and has achieved many good results. In a traditional OCR system, textual data is scanned and saved as images, and then transformed into text files [Lee and Chen, 1996] . There have also been many researches on handwriting OCR. In contrast, video OCR is more challenging than traditional OCR because we have to recognize small characters on a colorful background instead of black characters on a white background.",
+                "cite_spans": [
+                    {
+                        "start": 179,
+                        "end": 199,
+                        "text": "[Lee and Chen, 1996]",
+                        "ref_id": "BIBREF1"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Several approaches have been proposed to video OCR. Wu et al. [1997 Wu et al. [ , 1998 ] tried to find characters in pictures by means of connected components. Their method performs well on pictures but not films because the background of a film is more complicated, and text will also connect with other objects in the film. Lienhart et al. [1998 Lienhart et al. [ , 2000 found text by means of color segmentation, contrast segmentation, geometry analysis, and texture analysis. Li, Doermann and Kia [2000] adopted a neural network to detect strings in images. Li and Doermann [1999] also employed multiple images to enhance resolution. Smith and Kande [1997] used text and object shifting, and facial recognition to reduce the size of images. Sato et al. [1998] achieved higher OCR correctness by means of image improving and multi-frame integration.",
+                "cite_spans": [
+                    {
+                        "start": 52,
+                        "end": 67,
+                        "text": "Wu et al. [1997",
+                        "ref_id": "BIBREF15"
+                    },
+                    {
+                        "start": 68,
+                        "end": 86,
+                        "text": "Wu et al. [ , 1998",
+                        "ref_id": "BIBREF14"
+                    },
+                    {
+                        "start": 326,
+                        "end": 347,
+                        "text": "Lienhart et al. [1998",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 348,
+                        "end": 372,
+                        "text": "Lienhart et al. [ , 2000",
+                        "ref_id": "BIBREF4"
+                    },
+                    {
+                        "start": 480,
+                        "end": 507,
+                        "text": "Li, Doermann and Kia [2000]",
+                        "ref_id": "BIBREF3"
+                    },
+                    {
+                        "start": 562,
+                        "end": 584,
+                        "text": "Li and Doermann [1999]",
+                        "ref_id": "BIBREF2"
+                    },
+                    {
+                        "start": 638,
+                        "end": 660,
+                        "text": "Smith and Kande [1997]",
+                        "ref_id": "BIBREF11"
+                    },
+                    {
+                        "start": 745,
+                        "end": 763,
+                        "text": "Sato et al. [1998]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "This paper focuses on Chinese captions in videos. Section 2 introduces several issues concerning video OCR and the architecture of our system. Sections 3 to 8 describe each strategy and each module in detail. The performance was evaluated using films made by the Discovery Channel. Section 9 demonstrates an application for question answering. Section 10 presents conclusion.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "There are two kinds of texts in videos, i.e., captions and image texts. Captions often appear at specific positions, such as a textual line in the lower part of a screen, or a vertical text line in the left or right part of a screen. Image texts consist of characters appearing in an image, such as shop signs, automobile registration numbers, etc. They are themselves part of the image, so they change their positions when the camera moves. Captions are narratives or dialogues in a film, so they often carry valuable information. The focus of this paper is how to extract texts in captions.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Architecture",
+                "sec_num": "2."
+            },
+            {
+                "text": "Complex backgrounds often show up behind captions; thus, the first problem is how to remove backgrounds. After backgrounds are removed, the remaining captions are black characters on a white background. Th at will make the following OCR task easier. We also apply a post-processing procedure to improve OCR performance. Figure 1 shows the architecture of the whole system.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 320,
+                        "end": 328,
+                        "text": "Figure 1",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Architecture",
+                "sec_num": "2."
+            },
+            {
+                "text": "To evaluate the performance of the system, some films produced by the Discovery Channel were used as experimental materials. Their topics vary widely from natural science to history, military, adventures and human life.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "A Simple Method for Chinese Video OCR and Its Application to Question Answering 13",
+                "sec_num": null
+            },
+            {
+                "text": "The characteristics of captions are: (1) they are always in a straight line from left to right or up to down; (2) the characters usually have colors which contrast with the background, and often have perceivable borders; (3) they are always in the foreground of the image; (4) they usually consist of two or more characters; (5) the height of the caption region is not often higher than one third of the height of the image, because characters cannot be too large or too small for reading; (6) they have fixed height, width, and size; (7) they have fixed colors. We employ these characteristics to locate captions.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Deciding Caption Regions",
+                "sec_num": "3."
+            },
+            {
+                "text": "Before processing, we first transform the original images into binary images. This technique is often used in video processing. It helps to simplify the background and make the retrieval of captions much easier.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Binary Image",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "When extracting images from a film, we take 2 pictures in a second and save them in the BMP format. In a BMP file, the color of each point is recorded as its RGB-value, (red-value, green-value, blue-value) . Each value ranges in brightness from 0 to 255. Here, 0 indicates the darkest value and 255 the brightest value. Using the RGB-values, we can transform an image into a binary image using the following method:",
+                "cite_spans": [
+                    {
+                        "start": 158,
+                        "end": 205,
+                        "text": "RGB-value, (red-value, green-value, blue-value)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Binary Image",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "Let the binary-threshold be SegColorScore For each point (red-value, green-value, blue-value) in an image:",
+                "cite_spans": [
+                    {
+                        "start": 57,
+                        "end": 93,
+                        "text": "(red-value, green-value, blue-value)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Removing Backgrounds by Means of Multiple Images",
+                "sec_num": null
+            },
+            {
+                "text": "IF red-value, green-value, and blue-value are larger than SegColorScore THEN change the color of this point to black, i.e., (0, 0, 0) ELSE change the color of this point to white, i.e., (255, 255, 255) .",
+                "cite_spans": [
+                    {
+                        "start": 186,
+                        "end": 191,
+                        "text": "(255,",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 192,
+                        "end": 196,
+                        "text": "255,",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 197,
+                        "end": 201,
+                        "text": "255)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Removing Backgrounds by Means of Multiple Images",
+                "sec_num": null
+            },
+            {
+                "text": "In our experiment, SegColorScore was set to 190. Figure 2 shows an example of binary image transformation. The captions are clearly separated from the background. The result is black characters on a white background.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 49,
+                        "end": 57,
+                        "text": "Figure 2",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Removing Backgrounds by Means of Multiple Images",
+                "sec_num": null
+            },
+            {
+                "text": "After performing binary image transformation, we decide where the captions are. Here, we employ another characteristic of captions: if we draw a horizontal line across a caption, the line will go through many vertical lines of Chinese characters. As in printed characters, these vertical lines are often of the same width.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Deciding Caption Regions",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "Consider every point at the same height height i . A sequence of black points is called a segment. In this way, a horizontal line at height i is composed of a set SEGMENT i =(segment i1 , segment i2 , \u2026) of segments. If the difference between the numbers of black points in two neighboring segments is not larger than a predefined threshold (e.g., 3 in this paper), then we say these two segments belong to the same group. Thus, we have a set GROUP i =(group i1 , group i2 , \u2026) at height i . Seg(group ij ) is defined as the number of segments in group ij . Now we define Score As Caption Region (abbreviated as SACR hereafter) of height i as",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Deciding Caption Regions",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "( ) ( ) ij GROUP j ij i group Seg group Seg SACR i 2 1 log \u00d7 \u2211 = = .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Deciding Caption Regions",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "(1)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Deciding Caption Regions",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "Consider the following example. Here, 0 denotes a white point and 1 a black point. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 2 An Example of Binary Image Transformation. A Simple Method for Chinese Video OCR and Its Application to Question Answering 15",
+                "sec_num": null
+            },
+            {
+                "text": "|---------1----------||--------2-------||-----3-----||--4--| Seg(",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 2 An Example of Binary Image Transformation. A Simple Method for Chinese Video OCR and Its Application to Question Answering 15",
+                "sec_num": null
+            },
+            {
+                "text": "= + + + .",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Figure 2 An Example of Binary Image Transformation. A Simple Method for Chinese Video OCR and Its Application to Question Answering 15",
+                "sec_num": null
+            },
+            {
+                "text": "Assume that the height of an image is m. We calculate m SACR's for the height levels and compute the average SACR . The height levels that have SACR's higher than the average one are in the caption region. Figures 3 and 4 show two examples. On the left side is the original image; in the middle is its binary image; and on the right side is the corresponding SACR of each height level, where the x-axis denotes the height, the y-axis denotes the SACR value, the solid vertical line is SACR , and the horizontal dashed lines denote the caption regions.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 206,
+                        "end": 221,
+                        "text": "Figures 3 and 4",
+                        "ref_id": "FIGREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Figure 2 An Example of Binary Image Transformation. A Simple Method for Chinese Video OCR and Its Application to Question Answering 15",
+                "sec_num": null
+            },
+            {
+                "text": "The experiment was performed on three Discovery films: \"Lightening,\" \"Animals in the Wild,\" and \"Whales.\" There we re 69, 66, and 41 sentences in captions, respectively. The first 500 images of each film were extracted as experiment data. As shown in Table 1 , the precision rates obtained were 76.7%, 39.8% and 82.0%, respectively, but the recall rates were nearly 100%. Errors occurred in cases like the stone road shown in Figure 4 . The white stone road in the image had many black segments of the same width, so it was misjudged as a caption region. Such misjudgments can be filtered out in the OCR processing stage. Hence, the recall rate is more important here for retrieving all the captions. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 251,
+                        "end": 258,
+                        "text": "Table 1",
+                        "ref_id": "TABREF2"
+                    },
+                    {
+                        "start": 426,
+                        "end": 434,
+                        "text": "Figure 4",
+                        "ref_id": "FIGREF3"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Evaluation",
+                "sec_num": "3.3"
+            },
+            {
+                "text": "When we adjusted the binary image threshold SegColorScore, we found an interesting phenomenon: if SegColorScore was set too low, the background could not be removed very well; on the other hand, if it was set too high, the background was removed, but the captions were too unclear to do OCR. The value 190 used in the previous module resulted in very unclear images.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Removing Backgrounds within Single Images",
+                "sec_num": "4."
+            },
+            {
+                "text": "To do OCR more precisely, we have to keep the character clear while removing all the background. In this section, we will propose a method for removing backgrounds within single images by employing the difference between the captions and the background. How information from multiple images is used to remove backgrounds will be discussed in the next section.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Removing Backgrounds within Single Images",
+                "sec_num": "4."
+            },
+            {
+                "text": "During transformation into binary images, the values of SegColorScore will affect the clearness of the remaining images of captions. As shown in Figures 5 and 6, captions are clearly seen when SegColorScore is set to 140, but more background parts remain. The situation is reversed when it is set to 180.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-Level Binary Image",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "Here, we propose a new method, called 2-level binary image transformation, which employs two different SegColorScore values to keep captions clear and to remove backgrounds at the same time. The method is described in the following. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "2-Level Binary Image",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "Given a picture, we overlap two binary images obtained using two different SegColorScore values (let HiSegColorScore be the higher one, and LowSegColorScore the lower one). Consider the example shown in Figure 7 . ' ' denotes a black point in both binary images, and '\u00cd' a black point only in the binary image obtained using a lower SegColorScore value. We keep only those '\u00cd' areas adjacent to a ' ', because those areas are regarded as black points, and change the other areas into white points. The resulting image is shown on the right side of Figure 7 . Figure 8 shows the 2-level binary image result obtained from Figures 5 and 6, which is a clearer caption image.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 203,
+                        "end": 211,
+                        "text": "Figure 7",
+                        "ref_id": "FIGREF6"
+                    },
+                    {
+                        "start": 548,
+                        "end": 556,
+                        "text": "Figure 7",
+                        "ref_id": "FIGREF6"
+                    },
+                    {
+                        "start": 559,
+                        "end": 567,
+                        "text": "Figure 8",
+                        "ref_id": "FIGREF7"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "A Simple Method for Chinese Video OCR and Its Application to Question Answering 17",
+                "sec_num": null
+            },
+            {
+                "text": "Consider the image sho wn in Figure 9 , which contains large black areas. It is not easy to remove a background area with a high brightness value using the above method. Thus, another method shown below is proposed to clean such an area if it is large and wide. We will try to deal with small fragments in the next section by using multiple images of the same caption texts. THEN clear all the points in the area adjacent to this point END",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 29,
+                        "end": 37,
+                        "text": "Figure 9",
+                        "ref_id": "FIGREF8"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Removing Large Black Areas",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "We employ another characteristic of captions to remove small and bright backgrounds; i.e., the positions of the images of captions will not change with the camera, but the background will. We overlap all the images with the same caption texts. Those black points which appear in almost all the images are considered as caption texts. In the next two subsections, we will introduce the method we use to detect the changes of caption texts and the method we use to remove backgrounds by means of multiple images.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Removing Backgrounds by Means of Multiple Images",
+                "sec_num": "5."
+            },
+            {
+                "text": "The first task in removing backgrounds with multiple images is to decide which images have the same caption texts. Refer to the example shown in Figure 10 . We record the border information of all the black areas. After reading the next image, we compare the border information with that of the previous one. If the difference is larger than a threshold, say, SceneChangeScore, we postulate that the caption texts are different. In the experiment, the value of SceneChangeScore was set to 0.6.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 145,
+                        "end": 154,
+                        "text": "Figure 10",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Detecting Changes of Captions",
+                "sec_num": "5.1"
+            },
+            {
+                "text": "The same three films used to evaluate the method used to determine caption regions were also used to evaluate this method. Table 2 shows that the performance was quite good. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 123,
+                        "end": 130,
+                        "text": "Table 2",
+                        "ref_id": "TABREF3"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Detecting Changes of Captions",
+                "sec_num": "5.1"
+            },
+            {
+                "text": "After detecting a sequence of images with the same caption texts, we use the following method to remove the backgrounds. Let NumFrames be the total number of sequential images. We consider each point in the caption region. If it is black in 90% of the images (i.e., NumFrames \u00d7 0.9), then we set the point as black. Otherwise, it is set as a white point. Figure 11 shows an example. The background is removed more clearly than that is in Figure  9 .",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 355,
+                        "end": 364,
+                        "text": "Figure 11",
+                        "ref_id": "FIGREF0"
+                    },
+                    {
+                        "start": 438,
+                        "end": 447,
+                        "text": "Figure  9",
+                        "ref_id": "FIGREF8"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Removing Backgrounds by Means of Multiple Images",
+                "sec_num": "5.2"
+            },
+            {
+                "text": "At this point, there exists a binary image that has black characters on a white background for each sentence in a caption. We next apply traditional OCR techniques to retrieve caption texts. The first step in performing OCR is to decide the boundaries of each character.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Character Segmentation",
+                "sec_num": "6."
+            },
+            {
+                "text": "We first decide the left and right boundaries. The most popular way to perform character segmentation is to use projection profiles [Lu, 1995] . As shown in Figure 12 , we project every black point onto a horizontal line. Intuitively, the projection for the space between Chinese characters is zero. However, there is also space inside a Chinese character. We employ another cue to resolve this problem. The width of Chinese characters is often approximately equal to their height. Let the height of a caption region be ImageHeight. The gap that is a distance of ImageHeight\u00d70.7 ~ ImageHeight\u00d71.4 from the previous gap will be regarded as a possible segmentation point.",
+                "cite_spans": [
+                    {
+                        "start": 132,
+                        "end": 142,
+                        "text": "[Lu, 1995]",
+                        "ref_id": "BIBREF7"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 157,
+                        "end": 166,
+                        "text": "Figure 12",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Character Segmentation",
+                "sec_num": "6."
+            },
+            {
+                "text": "After deciding the left and right boundaries, we use the same method to decide the upper and lower boundaries of each character.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Character Segmentation",
+                "sec_num": "6."
+            },
+            {
+                "text": "We adopt a statistical model similar to that of Oka [1982] to perform Chinese OCR. Figure  13 shows an example. Each character image is separated into 16 equal parts. Starting from the center of each part, we observe its up, down, left, and right directions. If there is a black point in a given direction, the corresponding signature value is set to 0. Otherwise, it is set to 1. In this way, we will have 64 (16 parts \u00d7 4 directions) values (called a signature) for each character image.",
+                "cite_spans": [
+                    {
+                        "start": 48,
+                        "end": 58,
+                        "text": "Oka [1982]",
+                        "ref_id": "BIBREF9"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 83,
+                        "end": 93,
+                        "text": "Figure  13",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Optical Character Recognition",
+                "sec_num": "7."
+            },
+            {
+                "text": "A set of character images that were retrieved from the Discovery Channel films formed a corpus for collecting the signatures of a standard character corpus. When recognizing a new image, we first extract its signature and then compare it with the ones in the standard character corpus. The similarity is measured by counting how many values are matched. Therefore, the similarity score will be between 0 and 64. The higher the score is, the more similar the two patterns are. If the highest score of a new image is less than 16, it is regarded as non-character image.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Optical Character Recognition",
+                "sec_num": "7."
+            },
+            {
+                "text": "is compared with ' ' and ' ' in the standard character corpus. The corresponding signatures are as follows: The similarity of the image with ' ' and ' ' is 60 and 50, respectively, so ' ' is ranked as the first candidate of this image. Figure 14 illustrates the first ten candidates of each character image in after OCR is performed. The correct rate of the top one is very high, and most of correct answers appear in the top ten. Table 3 shows the top one performance. The film \"Genetics\" was used in inside test, and \"King of the Pyramids \" and \"The Real Cleopatra\" were used in the outside tests. The results show that the correct rates in the inside test was 91.5%, and that the performance of the outside tests was 78.5% and 81.5% for the two films, respectively. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 236,
+                        "end": 245,
+                        "text": "Figure 14",
+                        "ref_id": "FIGREF0"
+                    },
+                    {
+                        "start": 431,
+                        "end": 438,
+                        "text": "Table 3",
+                        "ref_id": "TABREF4"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "The following is an example. A new image",
+                "sec_num": null
+            },
+            {
+                "text": "We found that nearly 95% of the correct answers were in the top ten candidates, and Table 3 shows that the top one achieved 91.5% performance in the inside test. This section will touch on how to promote the correct answer which is not ranked first initially to the first position to improve the overall performance. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 84,
+                        "end": 91,
+                        "text": "Table 3",
+                        "ref_id": "TABREF4"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "OCR Post-Processing",
+                "sec_num": "8."
+            },
+            {
+                "text": "In Figure 14 , a value enclosed in parentheses before a candidate denotes its similarity score. First, we filter out those candidates whose scores are lower than the score of the top one candidate by a threshold. The filtered characters are shadowed in Figure 14 . Only the characters with larger scores are retained. This will reduce the number of possible candidates. Then, we perform the following steps. Consider three characters denoted ABC sequentially. Generate all the possible candidate pairs for ABC, e.g., A i B j or B m C n . Check if a candidate pair is in a dictionary (i.e., a two -character word), or is a part of a three-character word. If it is, we multiply the OCR similarity scores of these two candidates. Otherwise, their score is set to zero. Next, we find the pair with the highest score. If it is A i B j , then A i and B j are selected, and we start the next iteration from C (i.e., CDE). If it is B m C n , then A 1 , i.e., the top one candidate of A, is selected, and we start the next iteration from B (i.e., BCD).",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 3,
+                        "end": 12,
+                        "text": "Figure 14",
+                        "ref_id": "FIGREF0"
+                    },
+                    {
+                        "start": 253,
+                        "end": 262,
+                        "text": "Figure 14",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Basic Model",
+                "sec_num": "8.1"
+            },
+            {
+                "text": "For the above algorithm, several issues had to be evaluated in the experiments. For example, should we consider all the combinations of characters? Is the top one candidate more important than the others? Are longer words in the dictionary more helpful? We applied 3 strategies to the basic model to examine these factors. The experimental results were compared with those obtained using the Select-First and Longest-First models.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Strategies Used in Experiments",
+                "sec_num": "8.2"
+            },
+            {
+                "text": "[Strategy 1] All pairs of candidates are considered.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Strategies Used in Experiments",
+                "sec_num": "8.2"
+            },
+            {
+                "text": "[Strategy 2] Only pairs consisting of at least one ranked first candidate are proposed.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Strategies Used in Experiments",
+                "sec_num": "8.2"
+            },
+            {
+                "text": "In other words, when AB are recognized, only",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Strategies Used in Experiments",
+                "sec_num": "8.2"
+            },
+            {
+                "text": "A 1 B 1 , A 1 B 2 , \u2026, A 2 B 1 , A 3 B 1 , \u2026 are considered.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Strategies Used in Experiments",
+                "sec_num": "8.2"
+            },
+            {
+                "text": "[Strategy 3] 4-or 3-character words in the dictionary are proposed first. Then, Strategy 2 is considered.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Strategies Used in Experiments",
+                "sec_num": "8.2"
+            },
+            {
+                "text": "The standard character corpus was collected from six Discovery films (i.e., \"Natural Born Winners,\" \"Snakes,\" \"Genetics,\" \"The Southern Rockies,\" \"Great Quakes: Kobe, Japan,\" and \"Galapagos: Beyond Darwin\"). There were in total 7,818 character images, and only 2,256 signatures of distinct characters were recorded. Tables 4 to 6 show the experimental results for the three different films. Among them, \"Genetics\" was used in the inside test; \"King of the Pyramids \" and \"The Real Cleopatra\" were used in the outside tests. The first 700 images of each film were extracted as experiment data. The notations used in the tables are defined Longest-First: select the longest candidate combination which is collected in the dictionary. Tables 4, 5, and 6 show that Strategy 3 was the best one. The correct rates were 82.3% and 85.9% in the outside tests, and 94.2% in the inside test. 5.4% and 7.1% of the characters could not be found in the dictionary in the outside tests, respectively.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Evaluation",
+                "sec_num": "8.3"
+            },
+            {
+                "text": "We further compare the experimental results obtained using Strategy 3 and the Select-First Model in Table 7 , where \"T F\" is the number of characters recognized correctly using Select-First but incorrectly using Strategy 3, and \"F T\" is the number of characters recognized correctly using Strategy 3 but incorrectly using Select-First. From Table 7 , we can find that \"T F\" case was only 0.7%, but that 3.0% to 5.2% of more characters could be recognized correctly. This leads us to the conclusion that post-processing is helpful. Table 8 shows the experimental results for the three whole films. The main error in the outside test was that about 7~10% of the characters were not collected in the standard character corpus. The signatures of the standard character corpus were collected from the real images of the six films, and only those of 2,256 distinct characters were included. To solve this problem, we tried to collect the signatures from the existing font types. We experimented on and . The experimental results are listed in Table 9 . The first experiment was the same the experiment reported in Table 8 . In the second and the third experiments, we used 5,401 frequently used Chinese characters as the standard character corpus in and , respectively. Comparatively speaking, the results were worse, and using was better than using .",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 100,
+                        "end": 107,
+                        "text": "Table 7",
+                        "ref_id": "TABREF7"
+                    },
+                    {
+                        "start": 341,
+                        "end": 348,
+                        "text": "Table 7",
+                        "ref_id": "TABREF7"
+                    },
+                    {
+                        "start": 531,
+                        "end": 538,
+                        "text": "Table 8",
+                        "ref_id": "TABREF8"
+                    },
+                    {
+                        "start": 1037,
+                        "end": 1044,
+                        "text": "Table 9",
+                        "ref_id": "TABREF9"
+                    },
+                    {
+                        "start": 1108,
+                        "end": 1115,
+                        "text": "Table 8",
+                        "ref_id": "TABREF8"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Evaluation",
+                "sec_num": "8.3"
+            },
+            {
+                "text": "In addition, we prepared another standard character corpus for the fourth and the fifth experiments, in which 2,256 signatures came from the original corpus, and the other Chinese characters came from the images. The performance was improved, but it was still not as good as that obtained in the inside test. Meanwhile, the whole character set (13,060) did not perform better than the set of frequently used characters. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Evaluation",
+                "sec_num": "8.3"
+            },
+            {
+                "text": "Since caption texts in video can be extracted successfully using the procedures proposed in the previous sections, we tried to integrate the IR and QA techniques to develop a video question answering system in the next step. Figure 15 shows the interface of the Video QA System. Users issue questions in the submission window. The system finds answers in a film corpus and shows them in the answer window with several indicative pictures extracted from the video for each answer. If the user wants to watch the original film for an answer, he can click on that picture, and the system will play the film starting from the answer fragment.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 225,
+                        "end": 234,
+                        "text": "Figure 15",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Question Answering (QA) System",
+                "sec_num": "9."
+            },
+            {
+                "text": "The technique used for QA was proposed by Lin et al. [2001] . It implements a question answering system on heterogeneous collections including video. The correctness of Video OCR is not 100% yet (82.3% or better is shown in Tables 5 and 6), so pattern matching in traditional techniques (i.e., matching keywords or synonyms, or searching in other semantic trees) has to take OCR similarity into account. The score for extracting answers can be calculated as follows: where |qw i | denotes the number of characters in qw i , and qc k is the k th character in qw i (the same convention is used for pw j ). Ocr(qc k , pc k ) is the OCR similarity of characters qc k and pc k .",
+                "cite_spans": [
+                    {
+                        "start": 42,
+                        "end": 59,
+                        "text": "Lin et al. [2001]",
+                        "ref_id": "BIBREF6"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Video QA System",
+                "sec_num": "9.1"
+            },
+            {
+                "text": "Testing questions were collected from \"Assignment Discovery\" at the web site of Discovery, traditional Chinese version ( http://chinese.discovery.com/sch/index.html).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Questions",
+                "sec_num": "9.2.1"
+            },
+            {
+                "text": "\"Assignment Discovery\" is a project that provides many learning lessons from Discovery programs. This project provides lesson plans, activities, and comprehension questions and answers for teachers to use in designing study programs for students.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Questions",
+                "sec_num": "9.2.1"
+            },
+            {
+                "text": "Top 5 Answer List Query Submission Window",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Answer Film Player",
+                "sec_num": null
+            },
+            {
+                "text": "This paper has introduced a Chinese video OCR system, including image capturing, caption regions deciding, background removal, character segmentation, OCR, and NLP post-processing. The correctness achieved is above 90% for the inside test, and above 80% for the outside test. Its application to v ideo retrieval and a QA system have also been discussed.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "10."
+            },
+            {
+                "text": "There are mainly four kinds of OCR errors: (1) the standard character corpus is not complete; (2) the background is not clear enough;",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "10."
+            },
+            {
+                "text": "(3) character segmentation errors; and (4) errors in OCR post-processing. In our standard character corpus, there are only 2,256 characters. But there are 5,401 frequently used Chinese characters, not to mention 7,659 less frequently used characters. This is why many characters could not be recognized. In our experiments, most of the backgrounds could be cleared successfully. But if the objects do not move, or if small fragments appear behind the captions, it is not easy to remove them using our method. This will affect the performance of character segmentation and OCR. The OCR errors may also propagate to the post-processing module. For example, a character image that is not in the standard character corpus will not have a correct answer among its candidates, and these ten candidates will affect the choice of other characters.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "10."
+            }
+        ],
+        "back_matter": [
+            {
+                "text": "We selected the comprehension questions for six films as our testing questions to do the evaluation. We collected questions from this website in order to avoid bias. The films were \"Elephants,\" \"On Jupiter,\" \"Hubble: Secrets from Space,\" \"Eye of the Serpent,\" \"Whales,\" and \"Lightning.\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "A Simple Method for Chinese Video OCR and Its Application to Question Answering 27",
+                "sec_num": null
+            },
+            {
+                "text": "The performance of the QA system was measured in MRR (Mean Reciprocal Rank), which was used in the QA evaluation of TREC QA-Track [Voorhees, 2000] .There were 43 questions in total for these six films. The experiment results are listed in Table 10 . The MRR result was 0.1848 (=(4+5/2+3/3+1/4+1/5)/43). 32.6% (14/43) of the questions were answered correctly. From our investigation, the main sources of errors were as follows:(1) Characters in keywords were not collected in the standard character corpus, for example, \" \" in the question \" \"(2) Paraphrase problem.",
+                "cite_spans": [
+                    {
+                        "start": 130,
+                        "end": 146,
+                        "text": "[Voorhees, 2000]",
+                        "ref_id": "BIBREF12"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 239,
+                        "end": 247,
+                        "text": "Table 10",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Performance",
+                "sec_num": "9.2.2"
+            },
+            {
+                "text": "\", the answer text is \" \" The two phrases \" \" and \" \" are paraphrases.(3) More precise rules for deciding question focus are required.Consider the question \" \" It is classified as \"QUANTITY,\" so all quantity expressions become possible candidates. But we should only look for temperature expressions as answers.(4) World knowledge is needed.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "For the question \"",
+                "sec_num": null
+            },
+            {
+                "text": "\" The correct answer mentions that Franklin did an experiment in 1752, but \"the first\" is not mentioned.Therefore, it is hard to decide whether he was the first experimenter.We only employ information consisting of question foci, question keywords, and Named Entities in our Chinese QA system. From the above observations, world knowledge and semantic analysis are needed to answer these questions, especially \"How\" and \"Why\" questions. This is a challenging problem.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Consider the question \"",
+                "sec_num": null
+            }
+        ],
+        "bib_entries": {
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Analysis of Error Count Distribution for Improving the Postprocessing Performance of OCCR",
+                "authors": [
+                    {
+                        "first": "Yue-Shi",
+                        "middle": [],
+                        "last": "Lee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Hsin-Hsi",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "81--86",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lee, Yue-Shi and Hsin-Hsi Chen, \"Analysis of Error Count Distribution for Improving the Postprocessing Performance of OCCR,\" Communication of Chinese and Oriental Languages Information Processing Society, 1996, pp. 81-86.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Text Enhancement in Digital Video Using Multiple Frame Integration",
+                "authors": [
+                    {
+                        "first": "Huiping",
+                        "middle": [],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "David",
+                        "middle": [],
+                        "last": "Doermann",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Proceedings of SPIE, Document Recognition IV",
+                "volume": "",
+                "issue": "",
+                "pages": "1--8",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Li, Huiping and David Doermann, \"Text Enhancement in Digital Video Using Multiple Frame Integration,\" Proceedings of SPIE, Document Recognition IV, 1999, pp. 1-8.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Automatic Text Detection and Tracking in Digital Video",
+                "authors": [
+                    {
+                        "first": "Huiping; David",
+                        "middle": [],
+                        "last": "Li",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Omid",
+                        "middle": [],
+                        "last": "Doermann",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Kia",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "IEEE Transactions on Image Processing",
+                "volume": "9",
+                "issue": "1",
+                "pages": "147--156",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Li, Huiping; David Doermann and Omid Kia, \"Automatic Text Detection and Tracking in Digital Video,\" IEEE Transactions on Image Processing, 9(1) 2000, pp. 147-156.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "On the Segmentation of Text in Videos",
+                "authors": [
+                    {
+                        "first": "Rainer",
+                        "middle": [],
+                        "last": "Lienhart",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Axel",
+                        "middle": [],
+                        "last": "Wernicke",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Proceedings of IEEE International Conference on Multimedia and Expo (ICME2000)",
+                "volume": "",
+                "issue": "",
+                "pages": "1511--1514",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lienhart, Rainer and Axel Wernicke, \"On the Segmentation of Text in Videos,\" Proceedings of IEEE International Conference on Multimedia and Expo (ICME2000), 3 2000, pp. 1511-1514.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Automatic Text Segmentation and Text Recognition for Video Indexing",
+                "authors": [
+                    {
+                        "first": "Rainer",
+                        "middle": [],
+                        "last": "Lienhart",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Effelsberg",
+                        "middle": [],
+                        "last": "Wolfgang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lienhart, Rainer and Effelsberg Wolfgang, \"Automatic Text Segmentation and Text Recognition for Video Indexing,\" Technical Report TR-98-009, Praktische Informatik IV, 1998.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Open-Domain Question Answering on Heterogeneous Data",
+                "authors": [
+                    {
+                        "first": "Chuan-Jie",
+                        "middle": [],
+                        "last": "Lin",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Hsin-Hsi",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Che-Chia",
+                        "middle": [],
+                        "last": "Liu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Jin-He",
+                        "middle": [],
+                        "last": "Tsai",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Hong-Jia",
+                        "middle": [],
+                        "last": "Wong",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2001,
+                "venue": "Proceedings of Workshop on Human Language Technology and Knowledge Management, ACL",
+                "volume": "",
+                "issue": "",
+                "pages": "79--85",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lin, Chuan-Jie, Hsin-Hsi Chen, Che-Chia Liu, Jin-He Tsai and Hong-Jia Wong, \"Open-Domain Question Answering on Heterogeneous Data,\" Proceedings of Workshop on Human Language Technology and Knowledge Management, ACL, 2001, pp. 79-85.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Machine Printed Character Segmentation -An Overview",
+                "authors": [
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Lu",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Pattern Recognition",
+                "volume": "28",
+                "issue": "",
+                "pages": "67--80",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lu, Y., \"Machine Printed Character Segmentation -An Overview,\" Pattern Recognition, 28, 1995, pp. 67-80.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "A Simple Method for Chinese Video OCR and Its Application to Question Answering 29",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "A Simple Method for Chinese Video OCR and Its Application to Question Answering 29",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Handwritten Chinese-Japanese Characters Recognition by Using Cellular Feature",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [
+                            "I"
+                        ],
+                        "last": "Oka",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1982,
+                "venue": "Proceedings 6th International Joint Conference on Pattern Recognition",
+                "volume": "",
+                "issue": "",
+                "pages": "783--785",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Oka, R. I., \"Handwritten Chinese-Japanese Characters Recognition by Using Cellular Feature.\" Proceedings 6th International Joint Conference on Pattern Recognition, 1982, pp. 783-785.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Video OCR: Indexing Digital News Libraries by Recognition of Superimposed Caption",
+                "authors": [
+                    {
+                        "first": "Toshio",
+                        "middle": [],
+                        "last": "Sato",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Takeo",
+                        "middle": [],
+                        "last": "Kanade",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Ellen",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Hughes",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Michael",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Smith",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Shin'ichi Satoh",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "ACM Multimedia Systems",
+                "volume": "7",
+                "issue": "5",
+                "pages": "385--395",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Sato, Toshio, Takeo Kanade, Ellen K. Hughes, Michael A. Smith and Shin'ichi Satoh, \"Video OCR: Indexing Digital News Libraries by Recognition of Superimposed Caption,\" ACM Multimedia Systems, 7(5) 1999, pp. 385-395.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Video Skimming and Characterization Through the Combination of Image and Language Understanding Technique",
+                "authors": [
+                    {
+                        "first": "Michael",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Smith",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Takeo",
+                        "middle": [],
+                        "last": "Kande",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of IEEE Conference on Computer Vision and Pattern Recognition",
+                "volume": "",
+                "issue": "",
+                "pages": "775--781",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Smith, Michael A. and Takeo Kande, \"Video Skimming and Characterization Through the Combination of Image and Language Understanding Technique,\" Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, 1997, pp. 775-781.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "Overview of the TREC-9 Question Answering Track",
+                "authors": [
+                    {
+                        "first": "Ellen",
+                        "middle": [],
+                        "last": "Voorhees",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Proceedings of the Ninth Text Retrieval Conference (TREC-9",
+                "volume": "",
+                "issue": "",
+                "pages": "71--80",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Voorhees, Ellen, \"Overview of the TREC-9 Question Answering Track,\" Proceedings of the Ninth Text Retrieval Conference (TREC-9), 2000, pp. 71-80.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Informedia -Search and Summarization in the Video Medium",
+                "authors": [
+                    {
+                        "first": "Howard",
+                        "middle": [],
+                        "last": "Wactlar",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Proceedings of Imagina 2000 Conference",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Wactlar, Howard, \"Informedia -Search and Summarization in the Video Medium,\" Proceedings of Imagina 2000 Conference, 2000.",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "Text Finder: An Automatic System to Detect and Recognize Text in Images",
+                "authors": [
+                    {
+                        "first": "Victor",
+                        "middle": [],
+                        "last": "Wu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Edward",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Riseman",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "IEEE Transactions",
+                "volume": "21",
+                "issue": "11",
+                "pages": "1224--1229",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Wu, Victor and Edward M. Riseman, \"Text Finder: An Automatic System to Detect and Recognize Text in Images,\" IEEE Transactions on pattern analysis and machine intelligence, 21(11) 1998, pp. 1224-1229.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "Finding Text in Images",
+                "authors": [
+                    {
+                        "first": "Victor",
+                        "middle": [
+                            ";"
+                        ],
+                        "last": "Wu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Manmatha",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Edward",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Riseman",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the 2nd intl. conf. on Digital Libraries",
+                "volume": "",
+                "issue": "",
+                "pages": "1--10",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Wu, Victor; Manmatha, R. and Riseman, Edward. M., \"Finding Text in Images,\" Proceedings of the 2nd intl. conf. on Digital Libraries, 1997, pp. 1-10.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "uris": null,
+                "text": "The Architecture of the Video OCR System.",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF2": {
+                "uris": null,
+                "text": "Examples of Deciding Caption Regions (1).",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF3": {
+                "uris": null,
+                "text": "Examples of Deciding Caption Regions (2).",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF4": {
+                "uris": null,
+                "text": "Binary Image of SegColorScore=180.",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF5": {
+                "uris": null,
+                "text": "Binary Image of SegColorScore=140.",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF6": {
+                "uris": null,
+                "text": "Illustration of 2-Level Binary Image Transformation.",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF7": {
+                "uris": null,
+                "text": "2-Level Binary Image of Figure 5 and Figure 6.",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF8": {
+                "uris": null,
+                "text": "An Example of Removing Large Black Areas.Large Black AreasImage Height Range = (height of the caption region) 4; Total = Range \u00d7 Range \u00d7 0.9; CHECK each black point in the caption region Look at a square with edge of Range and with an upper-left corner at this point IF the number of black points in this square >= Total (i.e., 90% of the points are black)",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF9": {
+                "uris": null,
+                "text": "An Example of Detecting the Change of Captions.Images Borders A Simple Method for Chinese Video OCR and Its Application to Question Answering 19",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF10": {
+                "uris": null,
+                "text": "An Example of Removing Backgrounds by Multiple Images.",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF11": {
+                "uris": null,
+                "text": "Signature of Image \" \".",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF12": {
+                "uris": null,
+                "text": "An Example of Character Segmentation. A Simple Method for Chinese Video OCR and Its Application to Question Answering 21 1010100010001101110100100000101111101010010001011111111111001111 : 1010100000001101010101100000100111101010010001011111111111001111: 1110100000011111010001000100100111101100010001011111111111101101",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF13": {
+                "uris": null,
+                "text": "Simple Method for Chinese Video OCR and Its Application to Question Answering 23 : (baseline) select the top one candidate;",
+                "num": null,
+                "type_str": "figure"
+            },
+            "FIGREF14": {
+                "uris": null,
+                "text": "The Interface of Video QA System.",
+                "num": null,
+                "type_str": "figure"
+            },
+            "TABREF2": {
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "text": "Evaluation of Caption Region Deciding.",
+                "content": "<table><tr><td>Films</td><td colspan=\"2\">Actual System Decided</td><td colspan=\"2\">Correct Precision</td><td>Recall</td></tr><tr><td>Lightening</td><td>69</td><td>90</td><td>69</td><td>76.7%</td><td>100.0%</td></tr><tr><td>Animals in the Wild</td><td>66</td><td>161</td><td>64</td><td>39.8%</td><td>97.0%</td></tr><tr><td>Whales</td><td>41</td><td>50</td><td>41</td><td>82.0%</td><td>100.0%</td></tr></table>"
+            },
+            "TABREF3": {
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "text": "Evaluation of Detecting Changes of Subtitles.",
+                "content": "<table><tr><td>Film</td><td colspan=\"3\">Number of Changes Number of False Alarms Correctness</td></tr><tr><td>Lightening</td><td>69</td><td>0</td><td>100.0%</td></tr><tr><td>Animals in the Wild</td><td>66</td><td>3</td><td>95.5%</td></tr><tr><td>Whales</td><td>41</td><td>0</td><td>100.0%</td></tr></table>"
+            },
+            "TABREF4": {
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "text": "Experiment Results of OCR.",
+                "content": "<table><tr><td>Films</td><td>TOTAL</td><td>CORRECT</td><td>ERROR</td><td>MISS</td></tr><tr><td>Genetics</td><td>809</td><td>739 (91.5%)</td><td>69 (8.5%)</td><td>0</td></tr><tr><td>King of the Pyramids</td><td>684</td><td colspan=\"2\">537 (78.5%) 110 (16.1%)</td><td>37 (5.4%)</td></tr><tr><td>The Real Cleopatra</td><td>750</td><td colspan=\"2\">611 (81.5%) 86 (11.5%)</td><td>53 (7.1%)</td></tr></table>"
+            },
+            "TABREF6": {
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "text": "Experimental Results of Post-Processing for the Film \"Genetics\".",
+                "content": "<table><tr><td/><td>TOTAL</td><td>CORRECT</td><td>ERROR</td><td>MISS</td><td>Improve</td></tr><tr><td>Select-First</td><td>809</td><td colspan=\"2\">739 (91.5%) 69 (8.5%)</td><td>0</td><td>------</td></tr><tr><td>Longest-First</td><td>809</td><td colspan=\"2\">753 (93.1%) 56 (6.9%)</td><td>0</td><td>1.6%</td></tr><tr><td>Strategy 1</td><td>809</td><td colspan=\"2\">751 (92.8%) 58 (7.2%)</td><td>0</td><td>1.3%</td></tr><tr><td>Strategy 2</td><td>809</td><td colspan=\"2\">759 (93.8%) 50 (6.2%)</td><td>0</td><td>2.3%</td></tr><tr><td>Strategy 3</td><td>809</td><td colspan=\"2\">762 (94.2%) 47 (5.8%)</td><td>0</td><td>2.7%</td></tr><tr><td colspan=\"6\">Table 5. Experimental Results of Post-Processing for the Film \"King of the Pyramids\"</td></tr><tr><td/><td>TOTAL</td><td>CORRECT</td><td>ERROR</td><td>MISS</td><td>Improve</td></tr><tr><td>Select-First</td><td>684</td><td colspan=\"3\">537 (78.5%) 110 (16.1%) 37 (5.4%)</td><td>-------</td></tr><tr><td>Longest-First</td><td>684</td><td colspan=\"3\">544 (79.5%) 103 (15.1%) 37 (5.4%)</td><td>1.0%</td></tr><tr><td>Strategy 1</td><td>684</td><td colspan=\"3\">546 (79.8%) 101 (14.8%) 37 (5.4%)</td><td>1.3%</td></tr><tr><td>Strategy 2</td><td>684</td><td colspan=\"3\">559 (81.7%) 88 (12.9%) 37 (5.4%)</td><td>3.2%</td></tr><tr><td>Strategy 3</td><td>684</td><td colspan=\"3\">563 (82.3%) 84 (12.3%) 37 (5.4%)</td><td>3.8%</td></tr><tr><td colspan=\"6\">Table 6. Experimental Results of Post-Processing for the Film \"The Real Cleopatra\".</td></tr><tr><td/><td>TOTAL</td><td>CORRECT</td><td>ERROR</td><td>MISS</td><td>Improve</td></tr><tr><td>Select-First</td><td>750</td><td colspan=\"3\">611 (81.5%) 86 (11.5%) 53 (7.1%)</td><td>------</td></tr><tr><td>Longest-First</td><td>750</td><td colspan=\"3\">614 (81.9%) 83 (11.1%) 53 (7.1%)</td><td>0.4%</td></tr><tr><td>Strategy 1</td><td>750</td><td colspan=\"2\">635 (84.5%) 62 ( 8.3%)</td><td>53 (7.1%)</td><td>3.0%</td></tr><tr><td>Strategy 2</td><td>750</td><td colspan=\"2\">640 (85.3%) 57 ( 7.6%)</td><td>53 (7.1%)</td><td>3.8%</td></tr><tr><td>Strategy 3</td><td>750</td><td colspan=\"2\">644 (85.9%) 53 ( 7.1%)</td><td>53 (7.1%)</td><td>4.4%</td></tr></table>"
+            },
+            "TABREF7": {
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "text": "Comparison of Strategy 3 and Select-First.",
+                "content": "<table><tr><td>Film</td><td>Total Result</td><td>T T</td><td>T F</td><td>F T</td><td>F F</td></tr><tr><td>Genetics</td><td colspan=\"5\">809 94.2% 738 (91.2%) 6 (0.7%) 24 (3.0%) 41 (5.1%)</td></tr><tr><td colspan=\"6\">King of the Pyramids 647 87.0% 532 (82.2%) 5 (0.8%) 31 (4.8%) 79 (11.2%)</td></tr><tr><td>The Real Cleopatra</td><td colspan=\"5\">697 92.4% 608 (87.2%) 3 (0.4%) 36 (5.2%) 50 (7.2%)</td></tr></table>"
+            },
+            "TABREF8": {
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "text": "Experimental Results for the Entire Films Obtained Using Strategy 3.",
+                "content": "<table><tr><td>Film</td><td>Real Answers</td><td>Reported by System</td><td>Correct (Recall)</td><td>Error</td><td>Miss</td></tr><tr><td>Genetics</td><td>9189</td><td>8834</td><td colspan=\"3\">8105 (88.2%) 1481(16.1%) 26(0.3%)</td></tr><tr><td colspan=\"2\">King of the Pyramids 7976</td><td>7878</td><td colspan=\"3\">6582 (82.5%) 851(10.7%) 543(6.8%)</td></tr><tr><td>The Real Cleopatra</td><td>8862</td><td>8874</td><td colspan=\"3\">7365 (83.1%) 636(7.18%) 861(9.7%)</td></tr></table>"
+            },
+            "TABREF9": {
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "text": "Experimental Results on Different Standard Character Corpora (\"King of the Pyramids\").",
+                "content": "<table><tr><td/><td>Real Answers</td><td>Reported by System</td><td>Correct (Recall)</td><td>Error</td><td>Miss</td></tr><tr><td>2,256, Original</td><td>7976</td><td>7878</td><td>6582 (82.5%)</td><td>851 (10.7%)</td><td>543(6.8%)</td></tr><tr><td>5,401,</td><td>7976</td><td>7092</td><td>2648 (33.2%)</td><td>5325 (66.8%)</td><td>3(0.0%)</td></tr><tr><td>5,401,</td><td>7976</td><td>7380</td><td>3265 (40.9%)</td><td>4708 (59.0%)</td><td>3(0.0%)</td></tr><tr><td>5,401, Original+</td><td>7976</td><td>7885</td><td>6701 (84.0%)</td><td>1272 (15.9%)</td><td>3(0.0%)</td></tr><tr><td>13060, Original+</td><td>7976</td><td>7885</td><td>6612 (82.9%)</td><td>1272 (15.9%)</td><td>0(0.0%)</td></tr></table>"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O01/O01-3003.json

@@ -0,0 +1,1279 @@
+{
+    "paper_id": "O01-3003",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:08:50.414005Z"
+    },
+    "title": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method",
+    "authors": [
+        {
+            "first": "Jau-Hung",
+            "middle": [],
+            "last": "Chen",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "Advanced Techonlogy Center, Computer and Communication Research Laboratories",
+                "institution": "Industrial Technology Research Institute",
+                "location": {
+                    "addrLine": "Chutung 310",
+                    "country": "Taiwan"
+                }
+            },
+            "email": "chenjh@itri.org.tw"
+        },
+        {
+            "first": "Yung-An",
+            "middle": [],
+            "last": "Kao",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "Advanced Techonlogy Center, Computer and Communication Research Laboratories",
+                "institution": "Industrial Technology Research Institute",
+                "location": {
+                    "addrLine": "Chutung 310",
+                    "country": "Taiwan"
+                }
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "In a text-to-speech (TTS) conversion system based on the time-domain pitch-synchronous overlap-add (TD-PSOLA) method, accurate estimation of pitch periods and pitch marks is necessary for pitch modification to assure optimal quality of synthetic speech. In general, there are two major tasks in pitch marking: pitch detection and location determination. In this paper, an adaptable filter, which serves as a bandpass filter, is proposed for use in pitch detection to transform voiced speech into a sine-like wave. The pass band of the adaptable filter can be adapted based on the fundamental frequency. Based on the sine-like wave, a peak-valley decision method is proposed to determine the appropriate parts (positive part and negative part) of voiced speech for use in pitch mark estimation. In each pitch period, two possible peaks/valleys are searched, and dynamic programming is performed to obtain pitch marks. Experimental results indicate that our proposed method performs very well if correct pitch information is estimated.",
+    "pdf_parse": {
+        "paper_id": "O01-3003",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "In a text-to-speech (TTS) conversion system based on the time-domain pitch-synchronous overlap-add (TD-PSOLA) method, accurate estimation of pitch periods and pitch marks is necessary for pitch modification to assure optimal quality of synthetic speech. In general, there are two major tasks in pitch marking: pitch detection and location determination. In this paper, an adaptable filter, which serves as a bandpass filter, is proposed for use in pitch detection to transform voiced speech into a sine-like wave. The pass band of the adaptable filter can be adapted based on the fundamental frequency. Based on the sine-like wave, a peak-valley decision method is proposed to determine the appropriate parts (positive part and negative part) of voiced speech for use in pitch mark estimation. In each pitch period, two possible peaks/valleys are searched, and dynamic programming is performed to obtain pitch marks. Experimental results indicate that our proposed method performs very well if correct pitch information is estimated.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "In past years, the concatenative synthesis approach has been adopted for use in many text-to-speech (TTS) systems [Hamon et al. 1989] [Iwahashi et al. 1995] [Shih et al. 1996] [Chen et al. 1998 ] [Chou et al. 1998 ] [Charpentier et al. 1986] . Concatenative synthesis uses real recorded speech segments as synthesis units and concatenates them together during synthesis. In addition, the time-domain pitch-synchronous overlap-add (TD-PSOLA) [Charpentier et al. 1986 ] method has been employed to perform prosody modification. This method modifies the prosodic features of a synthesis unit according to the target prosodic information. Generally, the prosodic information of a speech unit includes its pitch (the fundamental frequency, f 0 ), intensity, duration, etc. For a synthesis scheme based on the TD-PSOLA method, it is necessary to obtain a pitch mark for each pitch period in order to assure optimal quality of synthetic speech. The pitch mark is a reference point for the overlap between speech signals.",
+                "cite_spans": [
+                    {
+                        "start": 114,
+                        "end": 133,
+                        "text": "[Hamon et al. 1989]",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 134,
+                        "end": 156,
+                        "text": "[Iwahashi et al. 1995]",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 157,
+                        "end": 175,
+                        "text": "[Shih et al. 1996]",
+                        "ref_id": "BIBREF2"
+                    },
+                    {
+                        "start": 176,
+                        "end": 193,
+                        "text": "[Chen et al. 1998",
+                        "ref_id": "BIBREF3"
+                    },
+                    {
+                        "start": 196,
+                        "end": 213,
+                        "text": "[Chou et al. 1998",
+                        "ref_id": "BIBREF4"
+                    },
+                    {
+                        "start": 216,
+                        "end": 241,
+                        "text": "[Charpentier et al. 1986]",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 441,
+                        "end": 465,
+                        "text": "[Charpentier et al. 1986",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "A speech synthesizer with various voices is useful for speech synthesis. Sometimes, it is also important for a service-providing company to have a synthesizer with the voice of its own employee or its favorite speaker. For conventional TTS systems, however, it is a demanding and tedious job to create a new voice. Recently, corpus-based TTS systems have been developed which use a large number of speech segments. Some approaches select speech segments as candidates for synthesis units. Establishing synthesis units involves speech segmentation, pitch estimation, pitch marking, and so on. Moreover, pitch marking is very labor-intensive task if no automatic mechanism is available.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In general, there are two major tasks in pitch marking: pitch detection and location determination. Compared to the literature on pitch detection [Rabiner et al. 1976] [Rabiner 1977 ] [Noll 1967 ] [Markel 1972 ] [Barnard et al. 1991] [Kadambe et al. 1991] [Barner 2000 ] [Huang et al. 2000] , few papers have focused on pitch marking [Moulines et al. 1990 ] [Kobayashi et al. 1998 ], which is also a difficult problem because of the great variability of speech signals. Moulines et al. [Moulines et al. 1990 ] proposed a pitch-marking algorithm based on the detection of abrupt changes at glottal closure instants. In each period, they assumed that the speech waveform could be represented by the concatenation of the response of two all-pole systems. On the other hand, Kobayashi et al. [Kobayashi et al. 1998 ] used dyadic wavelets for pitch marking. The glottal closure instant was detected by searching for a local peak in the wavelet transform of the speech waveform.",
+                "cite_spans": [
+                    {
+                        "start": 146,
+                        "end": 167,
+                        "text": "[Rabiner et al. 1976]",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 168,
+                        "end": 181,
+                        "text": "[Rabiner 1977",
+                        "ref_id": "BIBREF7"
+                    },
+                    {
+                        "start": 184,
+                        "end": 194,
+                        "text": "[Noll 1967",
+                        "ref_id": "BIBREF8"
+                    },
+                    {
+                        "start": 197,
+                        "end": 209,
+                        "text": "[Markel 1972",
+                        "ref_id": "BIBREF9"
+                    },
+                    {
+                        "start": 212,
+                        "end": 233,
+                        "text": "[Barnard et al. 1991]",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 234,
+                        "end": 255,
+                        "text": "[Kadambe et al. 1991]",
+                        "ref_id": "BIBREF11"
+                    },
+                    {
+                        "start": 256,
+                        "end": 268,
+                        "text": "[Barner 2000",
+                        "ref_id": "BIBREF12"
+                    },
+                    {
+                        "start": 271,
+                        "end": 290,
+                        "text": "[Huang et al. 2000]",
+                        "ref_id": "BIBREF13"
+                    },
+                    {
+                        "start": 334,
+                        "end": 355,
+                        "text": "[Moulines et al. 1990",
+                        "ref_id": "BIBREF14"
+                    },
+                    {
+                        "start": 358,
+                        "end": 380,
+                        "text": "[Kobayashi et al. 1998",
+                        "ref_id": "BIBREF15"
+                    },
+                    {
+                        "start": 470,
+                        "end": 507,
+                        "text": "Moulines et al. [Moulines et al. 1990",
+                        "ref_id": "BIBREF14"
+                    },
+                    {
+                        "start": 771,
+                        "end": 810,
+                        "text": "Kobayashi et al. [Kobayashi et al. 1998",
+                        "ref_id": "BIBREF15"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we propose a pitch-marking method based on an adaptable filter and a peak-valley estimation method. The block diagram of our method is shown in Fig. 1 . The input signals are limited to voiced speech because only the periodic parts are of interest. We introduce an adaptable filter, which serves as a bandpass filter, to transform voiced speech into a sine-like wave. FFT (Fast Fourier Transform) is used to transform voice to the frequency domain, and the filter's pass band is determined by finding the spectral peak of the fundamental frequency. Consequently, the pass band can be adapted based on the fundamental frequency. The autocorrelation method is then used to estimate the pitch periods on the sine-like wave. In addition, a peak-valley decision method is employed to determine which part of the voiced speech is suitable for pitch mark estimation. The positive part (the speech with positive amplitude) and the negative part (the speech with negative amplitude) are investigated in this method. This is demonstrated by Fig. 3(a) , which shows an example of a waveform having a negative part that reveals explicit periodicity. In general, it is possible to achieve better speech quality if the pitch marks are labeled at the positions of the extreme",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 159,
+                        "end": 165,
+                        "text": "Fig. 1",
+                        "ref_id": "FIGREF0"
+                    },
+                    {
+                        "start": 1046,
+                        "end": 1055,
+                        "text": "Fig. 3(a)",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "points (peaks and valleys) of speech. In each pitch period, two possible peaks/valleys are searched. Finally, the pitch marks are obtained through dynamic programming by calculating the degree of pitch distortion. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 33",
+                "sec_num": null
+            },
+            {
+                "text": "The proposed adaptable filter serves as a bandpass filter in which the pass band extends from 50 Hz to the detected fundamental frequency, up to 500 Hz, of the voiced speech. First, we will define the following symbols, which are used in this algorithm: The algorithm of the adaptable filter is described as follows:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "N:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "Step 1. Use FFT to transform the signal s m [n] to obtain the frequency response SF m [k].",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "Step 2. Find the position k p of the spectral peak of the fundamental frequency for SF m [k] by searching the first forty points of \u23d0SF m [k]\u23d0.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "Step 3. Decide on the filter's pass band.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "Let YF m [k]=SF m [k] if 3\u2264k\u2264k p +2 or 3\u2264N-k\u2264k p +2; otherwise, let YF m [k]=0.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "Step 4. Normalize YF m [k] by multiplying a scale of",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "Max k (\u23d0YF m [k]\u23d0)/\u23d0YF m [k p ]\u23d0.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "Step 5. Use IFFT (Inverse FFT) to transform the normalized YF m [k] to the time domain. Let o m [n] be the real part of the time domain signal.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "Finally, the refined pitch periods are obtained by analyzing the filtered speech o[n] using the conventional autocorrelation method. The waveform of o m [n] after IFFT may be discontinuous at the frame boundaries. A typical example is shown in Fig. 2 . However, such waveform discontinuity is not very significant and does not significantly affect the results of pitch period estimation.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 244,
+                        "end": 250,
+                        "text": "Fig. 2",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Pitch Detection Using an Adaptable Filter Followed by Application of the Autocorrelation Method",
+                "sec_num": "2."
+            },
+            {
+                "text": "An example of an adaptable filter is displayed in Fig. 3 . Panels (a) and (b) show the waveforms of the original speech and the filtered speech, respectively. It can be seen that the",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 50,
+                        "end": 56,
+                        "text": "Fig. 3",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Figure 2 A typical example of waveform discontinuity after IFFT.",
+                "sec_num": null
+            },
+            {
+                "text": "filtered speech is generally a sine-like wave with clear periodicity than the original speech waveform. For a frame in the middle of the voiced speech, the spectral contour is depicted in panel (d). Note that the frequency axis is not linearly plotted to allow inspection of the first spectral peak. The first peak was found at 168 Hz, which was the fundamental frequency. Finally, the pitch periods were obtained by analyzing the filtered speech using the conventional autocorrelation method. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 35",
+                "sec_num": null
+            },
+            {
+                "text": "Dynamic Programming",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Mark Determination Using a Peak-Valley Decision Method and",
+                "sec_num": "3."
+            },
+            {
+                "text": "From observations, we have found that voiced speech, s [\u2022] , is synchronous with filtered speech, o [\u2022] , either at peaks or at valleys. The cases illustrated in Figs. 3 (a) and 2 (b) are synchronous at valleys having explicit periodicity instead of at peaks. As a result, the pitch marks can be more easily determined in the negative part than in the positive part. In the following, the peak-valley decision method is used to calculate two costs by summing the amplitudes of s [q] , where q represents the position of the local extreme point of o [\u2022] over each pitch period:",
+                "cite_spans": [
+                    {
+                        "start": 55,
+                        "end": 58,
+                        "text": "[\u2022]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 100,
+                        "end": 103,
+                        "text": "[\u2022]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 479,
+                        "end": 482,
+                        "text": "[q]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 549,
+                        "end": 552,
+                        "text": "[\u2022]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 162,
+                        "end": 173,
+                        "text": "Figs. 3 (a)",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Peak-Valley Decision",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "\u2211 = \u22c5 = peak n peak peak peak N n Pos s N C 1 ]] [ [ 1 ,",
+                        "eq_num": "(1)"
+                    }
+                ],
+                "section": "Peak-Valley Decision",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "\u2211 = \u22c5 \u2212 = valley n valley valley valley N n Pos s N C 1 ]] [ [ 1 ,",
+                        "eq_num": "(2)"
+                    }
+                ],
+                "section": "Peak-Valley Decision",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "where the symbols are defined as follows: ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Peak-Valley Decision",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "Once the peak or valley, say the peak, has been adopted, the positions of the pitch marks are determined by picking the peaks of s [\u2022] . For a speech segment with a length of one pitch period, the PSOLA method can be used to synthesize good quality speech if the pitch mark is denoted at the signal with the largest amplitude. However, the largest peak may not correspond to the largest one in the next period (as shown in Fig. 4) . This inconsistency will result in unpleasant speech after the PSOLA method is used. Therefore, the two highest peaks in each period are searched in pitch mark determination. We do not use three peaks or more because this would improve the performance very little. In this paper, we consider that a peak is located at the signal with the largest amplitude among consecutive positive signals. Among",
+                "cite_spans": [
+                    {
+                        "start": 131,
+                        "end": 134,
+                        "text": "[\u2022]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 423,
+                        "end": 430,
+                        "text": "Fig. 4)",
+                        "ref_id": "FIGREF4"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Pitch Mark Determination Based on Dynamic Programming",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "peaks, the highest peak is the one with the largest amplitude. The second highest peak is the highest of the two peaks, the left-side and the right-side peaks, neighboring the highest peak.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 37",
+                "sec_num": null
+            },
+            {
+                "text": "For the i-th pitch period, P i , suppose the highest and the second highest peaks are located at L i1 and L i2 , respectively. It might occur that the second one is absent. In this case, we let L i2 = L i1 . For all the detected peaks, pitch mark determination is then performed based on dynamic programming. The distortion of the pitch period, d i (j,k), and its accumulation, A i (j), are defined as follows:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 37",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": ") , ( ) , ( ) 1 ( k j g P L L k j d i k i ij i + \u2212 \u2212 = \u2212 , for i=2,\u2026,PN,",
+                        "eq_num": "(3)"
+                    }
+                ],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 37",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "\u23ad \u23ac \u23ab \u23a9 \u23a8 \u23a7 + + = \u2212 \u2212 (2) ) 2 , ( (1), ) 1 , ( min ) ( 1 1 i i i i i A j d A j d j A , for i=2,3,\u2026,PN,",
+                        "eq_num": "(4)"
+                    }
+                ],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 37",
+                "sec_num": null
+            },
+            {
+                "text": "where PN is the total number of pitch period and j, k=1,2. In Equation 3 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 37",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "= = = otherwise , PN 1 1 or 1 if , 0 ) , ( k j k j g .",
+                        "eq_num": "(5)"
+                    }
+                ],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 37",
+                "sec_num": null
+            },
+            {
+                "text": "The penalty function is introduced here due to the preference for the highest peak.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 37",
+                "sec_num": null
+            },
+            {
+                "text": "The search path of the dynamic programming is illustrated in Fig. 5 . The peak locations (pitch marks) can be obtained by back tracing the peak sequence corresponding to the smallest values of A i (1) and A i (2). An example of the results of pitch marking is shown in Fig. 3(c) . A procedure similar to that described above can be applied for the case of a \"valley.\" ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 61,
+                        "end": 67,
+                        "text": "Fig. 5",
+                        "ref_id": "FIGREF5"
+                    },
+                    {
+                        "start": 269,
+                        "end": 278,
+                        "text": "Fig. 3(c)",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 37",
+                "sec_num": null
+            },
+            {
+                "text": "A continuous speech database was established which provides the basic synthesis units of our Mandarin Chinese TTS system. This database is composed of 70 phrases, and their lengths are from 4 to 6 Chinese characters. It includes a total of 436 tonal syllables comprising the required 413 basic synthesis units. A native female speaker read them in normal speaking style. The speech signals were then digitized by a 16-bit A/D converter at a 44.1k Hz sampling rate. Syllable segmentation was done manually in order to obtain the precise boundaries of the voiced speech and unvoiced speech. The total duration of the 436 voiced speech segments was about 2.1 minutes. For each syllable, the voiced speech was used to test the proposed methods. The frame size used in the adaptable filter was set to 4096 speech samples (92.8 ms). We used large frame size so that we could deal with signals with very low f 0 values.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experimental Environment",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "For the voiced speech, the waveforms along with the pitch marks obtained using our pitch-marking program were visually displayed. The pitch marks were then checked and corrected by an experienced person through a friendly interface. For evaluation of the experiments, we obtained 436 sets of human-labeled pitch marks, denoted as H, which comprises 23,868 pitch marks.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experimental Environment",
+                "sec_num": "4.1"
+            },
+            {
+                "text": "The peak-valley decision results were verified by human judgment based on visual displays. A success rate of 99.1% was obtained (4 of the 436 results disagreed). For the female speaker, we found that 97.2% of the voiced segments revealed clear periodicity in the negative parts.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Performance of the Pitch Marking Method",
+                "sec_num": "4.2"
+            },
+            {
+                "text": "The proposed method generated 23,860 pitch marks, denoted as I, without any duplication. The success rate of the pitch marking method is calculated as follows: ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 39",
+                "sec_num": null
+            },
+            {
+                "text": "As shown in Table 1 , a success rate of 97.2% was obtained (baseline), in contrast with 95% and 97% success rates obtained using the methods proposed in [Moulines et al. 1990] and [Kobayashi et al. 1998 ], respectively. Moreover, we found that most of the errors resulted from incorrect pitch detection results. Most of the pitch errors were due to large changes of pitch located at the boundaries of the voiced speech. With correct pitch information provided, our method achieved a success rate of 99.5%.",
+                "cite_spans": [
+                    {
+                        "start": 153,
+                        "end": 175,
+                        "text": "[Moulines et al. 1990]",
+                        "ref_id": "BIBREF14"
+                    },
+                    {
+                        "start": 180,
+                        "end": 202,
+                        "text": "[Kobayashi et al. 1998",
+                        "ref_id": "BIBREF15"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 12,
+                        "end": 19,
+                        "text": "Table 1",
+                        "ref_id": "TABREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 39",
+                "sec_num": null
+            },
+            {
+                "text": "The tone type of voice significantly affects the results of the detection of f0. The main reason for error detection of f 0 is dependent on the tone types of voice. There are five tones in Mandarin speech, including a high-level tone (Tone 1), a mid-rising tone (Tone2), a mid-falling-rising tone (Tone 3), a high-falling tone (Tone 4), a neutral tone (Tone 5). In our system, it is easy to detect f 0 for tone 1 and tone 2 since the spectral peak of f 0 is prominent ( Fig. 3 (d) ). For tone 3, tone 4 and tone 5, f 0 may be erroneously detected at the end of the voice segment if the consecutive pitch periods change abruptly (Fig. 6 (a) ). For this case, the spectral peak of f 0 is unclear (Fig. 6 (b) ), which may result in error detection. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 470,
+                        "end": 480,
+                        "text": "Fig. 3 (d)",
+                        "ref_id": "FIGREF1"
+                    },
+                    {
+                        "start": 628,
+                        "end": 639,
+                        "text": "(Fig. 6 (a)",
+                        "ref_id": "FIGREF6"
+                    },
+                    {
+                        "start": 694,
+                        "end": 705,
+                        "text": "(Fig. 6 (b)",
+                        "ref_id": "FIGREF6"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method 39",
+                "sec_num": null
+            },
+            {
+                "text": "Baseline Using correct pitch Success rate 97.2% 99.5% ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Condition",
+                "sec_num": null
+            },
+            {
+                "text": "In this paper, a preliminary work on pitch marking has been proposed. We have presented an adaptable filter which can be combined with the autocorrelation method to perform pitch detection. On the other hand, a peak-valley decision method has been proposed to select either the positive or the negative part for pitch mark evaluation. Also, a dynamic-programming-based pitch mark determination method has been demonstrated, where two peaks/valleys are searched in each period. In the experiments, our pitch-marking method achieved a 97.2% success rate. Furthermore, a high success rate of 99.5% was obtained when correct pitch information was provided.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions",
+                "sec_num": "5"
+            }
+        ],
+        "back_matter": [
+            {
+                "text": "This paper is a partial result of Project 3XS1B11 conducted by ITRI under sponsorship of the Ministry of Economic Affairs, Taiwan, R.O.C.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Acknowledgement",
+                "sec_num": null
+            }
+        ],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "A diphone synthesis based on time-domain prosodic modifications of speech",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Hamon",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Moulines",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Charpentier",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1989,
+                "venue": "International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "238--241",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Hamon, C., E. Moulines, and F. Charpentier, \"A diphone synthesis based on time-domain prosodic modifications of speech,\" International Conference on Acoustics, Speech, and Signal Processing, 1989, pp.238-241.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Speech segment network approach for optimization of synthesis unit set",
+                "authors": [
+                    {
+                        "first": "N",
+                        "middle": [],
+                        "last": "Iwahashi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Sagisaka",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Computer Speech and Language",
+                "volume": "",
+                "issue": "",
+                "pages": "335--352",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Iwahashi, N. and Y. Sagisaka, \"Speech segment network approach for optimization of synthesis unit set,\" Computer Speech and Language, 1995, pp.335-352.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Issues in text-to-speech conversion for Mandarin",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Shih",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Sproat",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Computational Linguistics and Chinese Language Processing",
+                "volume": "1",
+                "issue": "",
+                "pages": "37--86",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Shih, C. L. and R. Sproat, \"Issues in text-to-speech conversion for Mandarin,\" Computational Linguistics and Chinese Language Processing, vol.1, 1996, pp.37-86.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "An RNN-based prosodic information Synthesizer for Mandarin text-to-speech",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Hwang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Wang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "IEEE Transactions on Speech and Audio Processing",
+                "volume": "6",
+                "issue": "3",
+                "pages": "226--239",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, S. H., S. H. Hwang and Y. R. Wang, \"An RNN-based prosodic information Synthesizer for Mandarin text-to-speech,\" IEEE Transactions on Speech and Audio Processing, 6(3), 1998, pp. 226-239.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Corpus-based Mandarin speech synthesis with contextual syllabic units based on phonetic properties",
+                "authors": [
+                    {
+                        "first": "F",
+                        "middle": [
+                            "C"
+                        ],
+                        "last": "Chou",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [
+                            "Y"
+                        ],
+                        "last": "Tseng",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "893--896",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chou, F. C. and C. Y. Tseng, \"Corpus-based Mandarin speech synthesis with contextual syllabic units based on phonetic properties,\" International Conference on Acoustics, Speech, and Signal Processing, 1998, pp.893-896.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Diphone synthesis using an overlap-add technique for speech waveforms concatenation",
+                "authors": [
+                    {
+                        "first": "F",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Charpentier",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "G"
+                        ],
+                        "last": "Stella",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1986,
+                "venue": "International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "2015--2020",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Charpentier, F. J. and M. G. Stella, \"Diphone synthesis using an overlap-add technique for speech waveforms concatenation,\" International Conference on Acoustics, Speech, and Signal Processing, 1986, pp. 2015-2020.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "A Comparative performance study of several pitch detection algorithms",
+                "authors": [
+                    {
+                        "first": "L",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Rabiner",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Cheng",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "A",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Rosenberg",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1976,
+                "venue": "IEEE Transactions on Acoustics., Speech and Signal Processing",
+                "volume": "24",
+                "issue": "",
+                "pages": "399--417",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Rabiner, L. R., M. J. Cheng, A. E. Rosenberg, and C. A. McGonegal, \"A Comparative performance study of several pitch detection algorithms,\" IEEE Transactions on Acoustics., Speech and Signal Processing, 24, 1976, pp. 399-417.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "On the use of autocorrelation analysis for pitch detection",
+                "authors": [
+                    {
+                        "first": "L",
+                        "middle": [
+                            "R"
+                        ],
+                        "last": "Rabiner",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1977,
+                "venue": "IEEE Transactions on Acoustics., Speech and Signal Processing",
+                "volume": "25",
+                "issue": "",
+                "pages": "24--33",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Rabiner, L. R., \"On the use of autocorrelation analysis for pitch detection,\" IEEE Transactions on Acoustics., Speech and Signal Processing, 25, 1977, pp. 24-33.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Cepstrum pitch determination",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Noll",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1967,
+                "venue": "Journal of the Acoustical Society of America",
+                "volume": "47",
+                "issue": "",
+                "pages": "293--309",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Noll, A. M., \"Cepstrum pitch determination,\" Journal of the Acoustical Society of America, 47, 1967, pp. 293-309.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "The SIFT algorithm for fundamental frequency estimation",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Markel",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1972,
+                "venue": "IEEE Transactions on Audio Electroacoustics",
+                "volume": "",
+                "issue": "",
+                "pages": "367--377",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Markel, J. D., \"The SIFT algorithm for fundamental frequency estimation,\" IEEE Transactions on Audio Electroacoustics, Au-20, 1972, pp. 367-377.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Pitch detection with a neural-net classifier",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Barnard",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "R",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Cole",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "P"
+                        ],
+                        "last": "Vea",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Alleva",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "IEEE Transactions on Signal Processing",
+                "volume": "39",
+                "issue": "2",
+                "pages": "298--307",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Barnard, E., R. A. Cole, M. P. Vea, and F. A. Alleva, \"Pitch detection with a neural-net classifier,\" IEEE Transactions on Signal Processing, 39(2), 1991, pp. 298-307.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "A comparison of a wavelet functions for pitch detection of speech signals",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Kadambe",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Boudreaux-Bartels",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1991,
+                "venue": "International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "449--452",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Kadambe, S., G. F. Boudreaux-Bartels, \"A comparison of a wavelet functions for pitch detection of speech signals,\" International Conference on Acoustics, Speech, and Signal Processing, 1991, pp. 449-452.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "Colored L-l filters and their application in speech pitch detection",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Barner",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "IEEE Transactions on Signal Processing",
+                "volume": "48",
+                "issue": "9",
+                "pages": "2601--2606",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Barner, K. E., \"Colored L-l filters and their application in speech pitch detection,\" IEEE Transactions on Signal Processing, 48(9), 2000, pp. 2601-2606.",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "Pitch tracking and tone features for Mandarin speech recognition",
+                "authors": [
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Seide",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "1523--1526",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Huang, H. and F. Seide, \"Pitch tracking and tone features for Mandarin speech recognition,\" International Conference on Acoustics, Speech, and Signal Processing, 2000, pp.1523-1526.",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "A real-time French text-to-speech system generating high-quality synthetic speech",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Moulines",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Emerard",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Larreur",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "L"
+                        ],
+                        "last": "Le Saint Milon",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "L",
+                        "middle": [
+                            "Le"
+                        ],
+                        "last": "Faucheur",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Marty",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Charpentier",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Sorin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1990,
+                "venue": "International Conference on Acoustics, Speech, and Signal Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "309--312",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Moulines, E., F. Emerard, D. Larreur, J. L. Le Saint Milon, L. Le Faucheur, F. Marty, F. Charpentier, and C. Sorin, \"A real-time French text-to-speech system generating high-quality synthetic speech,\" International Conference on Acoustics, Speech, and Signal Processing, 1990, pp.309-312.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "Wavelet analysis used in text-to-speech synthesis",
+                "authors": [
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Kobayashi",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Sakamoto",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Saito",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Hashimoto",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [],
+                        "last": "Nishimura",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Suzuki",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "IEEE Transactions on Circuits and Systems-II",
+                "volume": "45",
+                "issue": "8",
+                "pages": "1125--1129",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Kobayashi, M., M. Sakamoto, T. Saito, Y. Hashimoto, M. Nishimura, and K. Suzuki, \"Wavelet analysis used in text-to-speech synthesis,\" IEEE Transactions on Circuits and Systems-II, Analog and Digital Signal Processing, 45(8), 1998, pp. 1125-1129.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "Block diagram of the proposed pitch-marking method."
+            },
+            "FIGREF1": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "Results obtained using the adaptable filter and pitch mark determination. (a) Waveform of the voiced speech with explicit periodicity in the negative part. (b) Waveform of the filtered speech. (c) Detected pitch marks. (d) Spectral contour (note that the frequency axis is not linearly plotted)."
+            },
+            "FIGREF2": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "Pos peak : position of the n-th peak of o[\u2022].][n Pos valley : position of the n-th valley of o[\u2022].The peak-valley decision is made as follows:If peak C > valley C, then the positive part (peak) of s[\u2022]  is adopted for evaluation of the pitch marks. Otherwise, the negative part (valley) of s[\u2022]   is adopted."
+            },
+            "FIGREF4": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "An example of a waveform (syllable /a/ of tone 3), in which the largest peak does not correspond to the largest one in the next period (indicated by the circles)."
+            },
+            "FIGREF5": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "Illustration of the peak-picking search path of the dynamic programming."
+            },
+            "FIGREF6": {
+                "num": null,
+                "uris": null,
+                "type_str": "figure",
+                "text": "An example of unclear spectral peaks. (a) Waveform of the syllable /a/ of tone 3. (b) Spectral contour corresponding to the end part of the waveform (note that the frequency axis is not linearly plotted)."
+            },
+            "TABREF2": {
+                "num": null,
+                "text": "Success rate of the pitch-marking method.",
+                "content": "<table/>",
+                "type_str": "table",
+                "html": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O02/O02-1002.json

@@ -0,0 +1,1022 @@
+{
+    "paper_id": "O02-1002",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:05:51.583801Z"
+    },
+    "title": "Word Sense Disambiguation and Sense-Based NV Event Frame Identifier",
+    "authors": [
+        {
+            "first": "Jia-Lin",
+            "middle": [],
+            "last": "Tsai",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Academia Sinica",
+                "location": {
+                    "settlement": "Nankang",
+                    "region": "Taipei",
+                    "country": "Taiwan, R.O.C"
+                }
+            },
+            "email": "tsaijl@iis.sinica.edu.tw"
+        },
+        {
+            "first": "Wen-Lian",
+            "middle": [],
+            "last": "Hsu",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Academia Sinica",
+                "location": {
+                    "settlement": "Nankang",
+                    "region": "Taipei",
+                    "country": "Taiwan, R.O.C"
+                }
+            },
+            "email": "hsu@iis.sinica.edu.tw"
+        },
+        {
+            "first": "Jeng-Woei",
+            "middle": [],
+            "last": "Su",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Academia Sinica",
+                "location": {
+                    "settlement": "Nankang",
+                    "region": "Taipei",
+                    "country": "Taiwan, R.O.C"
+                }
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "Word sense is ambiguous in natural language processing (NLP). This phenomenon is particularly keen in cases involving noun-verb (NV) w ord-pairs. This paper describes a sense-based noun-verb event frame (NVEF) identifier that can be used to disambiguate word sense in Chinese sentences effectively. A knowledge representation system (the NVEF-KR tree) for the NVEF sense-pair identifier is also proposed. We use the word sense of Hownet, which is a Chinese-English bilingual knowledge-base dictionary. Our experiment showed that the NVEF identifier was able to achieve 74.8% accuracy for the test sentences studied based only on NVEF sense-pair knowledge. By applying the techniques of longest syllabic NVEF-word-pair first and exclusion word checking, the sense accuracy for the same test sentences could be further improved to 93.7%. There were four major reasons for the incorrect cases: (1) lack of a bottom-up tagger, (2) lack of non-NVEF knowledge, (3) inadequate word segmentation, and (4) lack of a multi-NVEF analyzer. If these four problems could be resolved, the accuracy would reach 98.9%. The results of this study indicate that NVEF sense-pair knowledge is effective for word sense disambiguation and is likely to be important for general NLP.",
+    "pdf_parse": {
+        "paper_id": "O02-1002",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "Word sense is ambiguous in natural language processing (NLP). This phenomenon is particularly keen in cases involving noun-verb (NV) w ord-pairs. This paper describes a sense-based noun-verb event frame (NVEF) identifier that can be used to disambiguate word sense in Chinese sentences effectively. A knowledge representation system (the NVEF-KR tree) for the NVEF sense-pair identifier is also proposed. We use the word sense of Hownet, which is a Chinese-English bilingual knowledge-base dictionary. Our experiment showed that the NVEF identifier was able to achieve 74.8% accuracy for the test sentences studied based only on NVEF sense-pair knowledge. By applying the techniques of longest syllabic NVEF-word-pair first and exclusion word checking, the sense accuracy for the same test sentences could be further improved to 93.7%. There were four major reasons for the incorrect cases: (1) lack of a bottom-up tagger, (2) lack of non-NVEF knowledge, (3) inadequate word segmentation, and (4) lack of a multi-NVEF analyzer. If these four problems could be resolved, the accuracy would reach 98.9%. The results of this study indicate that NVEF sense-pair knowledge is effective for word sense disambiguation and is likely to be important for general NLP.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "Word sense disambiguation (WSD) has been a pervasive problem in natural language processing (NLP) since 1949 [Weaver 1949] . Word sense ambiguity (or lexical ambiguity), is generally classified into two types: syntactic and semantic ambiguity [Small et al. 1988 , Krovetz et al. 1992 . Syntactic ambiguity is caused by differences in syntactic categories (e.g.",
+                "cite_spans": [
+                    {
+                        "start": 109,
+                        "end": 122,
+                        "text": "[Weaver 1949]",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 243,
+                        "end": 261,
+                        "text": "[Small et al. 1988",
+                        "ref_id": "BIBREF8"
+                    },
+                    {
+                        "start": 262,
+                        "end": 283,
+                        "text": ", Krovetz et al. 1992",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "\"play\" can occur as a noun or verb). Semantic ambiguity is caused by homonymy (e.g. \"bank\" in \"to put money in a bank,\" \"the bank of a river\") or polysemy (e.g. \"face\" in \"human face,\" \"face of a clock\"). Although many approaches have been adopted to disambiguate word sense, algorithms for word sense determination still are not reliable [Krovetz et al. 1992 , Resnik et al. 2000 . Human beings usually can disambiguate word sense by using additional information from the speaker, the writer or the context. When out-of-context (or out-of-sentence) information is not symbolized and processed in the computer, WSD either becomes very difficult or, sometimes, impossible. Therefore, it is crucial to investigate what kind of knowledge is useful for WSD [Krovetz et al. 1992 ].",
+                "cite_spans": [
+                    {
+                        "start": 339,
+                        "end": 359,
+                        "text": "[Krovetz et al. 1992",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 360,
+                        "end": 380,
+                        "text": ", Resnik et al. 2000",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 753,
+                        "end": 773,
+                        "text": "[Krovetz et al. 1992",
+                        "ref_id": "BIBREF5"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "According to a study in cognitive science [Choueka et al. 1983] , people often disambiguate word sense using only a few other words in a given context (frequently only one additional word). Thus, the relationships between one word and others can be effectively used to resolve ambiguity. Furthermore, from [Small et al. 1988 , Krovetz et al. 1992 , Resnik et al. 2000 , most ambiguities occur with nouns and verbs, and the object-event (i.e. noun-verb) distinction is a major ontological division for humans [Carey 1992 ]. However, no clear data has been collected to support these claims. These observations motivated us to demonstrate through an experiment, how noun-verb (NV) relationships can be used to disambiguate word sense in Chinese sentences.",
+                "cite_spans": [
+                    {
+                        "start": 42,
+                        "end": 63,
+                        "text": "[Choueka et al. 1983]",
+                        "ref_id": "BIBREF2"
+                    },
+                    {
+                        "start": 306,
+                        "end": 324,
+                        "text": "[Small et al. 1988",
+                        "ref_id": "BIBREF8"
+                    },
+                    {
+                        "start": 325,
+                        "end": 346,
+                        "text": ", Krovetz et al. 1992",
+                        "ref_id": "BIBREF5"
+                    },
+                    {
+                        "start": 347,
+                        "end": 367,
+                        "text": ", Resnik et al. 2000",
+                        "ref_id": "BIBREF6"
+                    },
+                    {
+                        "start": 508,
+                        "end": 519,
+                        "text": "[Carey 1992",
+                        "ref_id": "BIBREF0"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we shall focus on word sense disambiguation involving NV word-pairs since these are most troublesome. Consider the following sentence: \" (This car moves well).\" In this sentence, we have two possible NV word-pairs, \" -(car, move)\" and \" -(auto-shop, move).\" It is clear that the permissible NV word-pair is \" -(car, move).\" We shall call such a permissible NV word-pair an NV-event frame (NVEF) word-pair. Using a collection of pre-learned NVEF word-pairs, we can identify the NVEF word-pair \" -\" from the sentence \" .\" The word \" \" in a dictionary can have three possible senses: 'surname' (noun), 'car' (noun) and 'turn' (ve rb). To resolve this ambiguity, we can use the pre-defined sense of the NVEF word-pair \" -(car, move)\" to determine that the correct sense of the Chinese word \" \" is \"car\" in the above Chinese sentence.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we shall show that knowledge of NVEF sense-pairs (to be defined in Section 2) can be effectively used to resolve word sense ambiguity. In the next section, we will propose an NVEF sense-pair identifier, which is based on pre-stored knowledge of NVEF sense-pairs. We use this NVEF sense-pair identifier to identify NVEF word-pairs in an input sentence and to determine the corresponding word senses. In Section 3, we will present and analyze the results of a WSD experiment on a set of test sentences using the NVEF sense-pair identifier. Finally, we will give conclusions and directions for future research in Section 4.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "We use Hownet [Dong] as our system's Chinese machine-readable dictionary (MRD). Hownet is a Chinese-English bilingual knowledge-base dictionary, which provides knowledge of the Chinese lexicon, parts-of-speech (POS) and word senses.",
+                "cite_spans": [
+                    {
+                        "start": 14,
+                        "end": 20,
+                        "text": "[Dong]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Development of an NVEF Sense-Pair Identifier",
+                "sec_num": "2."
+            },
+            {
+                "text": "The sense of a word is defined as its DEF (concept definition) in Hownet. Table 1 lists three different senses of the Chinese word \" (Che/car/turn).\" In Hownet, the DEF of a word consists of its main feature and secondary features. For example, in the DEF \"character| ,surname| ,human| ,ProperName| \" of the word \" (Che),\" the first item \"character| \" is the main feature, and the remaining three items, \"surname| ,\" \"human| ,\" and \"ProperName| ,\" are its secondary features. The main feature in Hownet can inherit features in the hypernym-hyponym hierarchy. There are approximately 1,500 features in Hownet. Each of these features is called a sememe, which refers to the smallest semantic unit that cannot be further reduced. The Hownet dictionary used in this study contains 50,121 Chinese words, among which there are 29,719 nouns, 16,652 verbs and 16,242 senses (including 9,893 noun-senses and 4,440 verb-senses). Table 2 gives the statistics of the number of senses per Chinese word and the number of Chinese words per sense used in Hownet. Now, take the NV word-pair \" -(car, move)\" for example. According to the sense of the Chinese word \" (Che/car/turn)\" and the sense of the Chinese word \" (move),\" the only permissible NVEF sense-pair for the NV word-pair \" -(car, move)\" is \"LandVehicle| \"-\"VehicleGo| .\" We call such a permissible NV sense-pair an NVEF sense-pair in this paper. Note that an NVEF sense-pair is a class that includes the permissible word-pair instance \" -(car, move).\"",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 74,
+                        "end": 81,
+                        "text": "Table 1",
+                        "ref_id": "TABREF0"
+                    },
+                    {
+                        "start": 919,
+                        "end": 926,
+                        "text": "Table 2",
+                        "ref_id": "TABREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Definition of an NVEF Sense-Pair",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "A knowledge representation tree (KR-tree) of NVEF sense-pairs is shown in Fig.1 . There are two types of nodes in the KR-tree, namely, function nodes and concept nodes. Concept nodes refer to words and features in Hownet. Function nodes are used to define the relationships between their parent and children concept nodes. If a concept node A is the child of another concept node B, then A is a subclass of B. Following this convention, we can omit the function node \"subclass\" (which should exist) between A and B. We can classify the noun-sense class ( ) into 15 subclasses according to their main features. They are \" (bacteria),\" \" (animal),\" \" (human),\" \" (plant),\" \" (artifact),\" \" (natural),\" \" (event),\" \" (mental),\" \" (phenomena),\" \" (shape),\" \" (place),\" \" (location),\" \" (time),\" \" (abstract)\" and \" (quantity).\" Appendix A gives a sample table of 15 main features of nouns in each noun-sense subclass. subclasses, we have designed three subclass sense-symbols, in which \"=\" means \"exact,\" \"&\" means \"like,\" and \"%\" means \"inclusive.\" An example using these symbols is given below.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 74,
+                        "end": 79,
+                        "text": "Fig.1",
+                        "ref_id": "FIGREF0"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Knowledge Representation Tree of NVEF Sense-Pairs",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "Given three senses S 1 , S 2 and S 3 defined by a main feature A and three secondary features B, C and D, let",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Knowledge Representation Tree of NVEF Sense-Pairs",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "S 1 = A, B, C, D,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Knowledge Representation Tree of NVEF Sense-Pairs",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "S 2 = A, B, and",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Knowledge Representation Tree of NVEF Sense-Pairs",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "S 3 = A, C, D.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Knowledge Representation Tree of NVEF Sense-Pairs",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "Then, we have that sense S 2 is in the \"=A,B\" exact-subclass; senses S 1 and S 2 are in the \"&A,B\" like-subclass; and senses S 1 S 2 , and S 3 are in the \"%A\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Knowledge Representation Tree of NVEF Sense-Pairs",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "inclusive-subclass.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Knowledge Representation Tree of NVEF Sense-Pairs",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "(2) Word-Instance ( ): The content of its children are the words belonging to the sense subclass of its parent node. These words are self-learned by the NVEF sense-pair identifier according to the sentences under the Test-Sentence nodes.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Knowledge Representation Tree of NVEF Sense-Pairs",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "(3) Test-Sentence ( ): The content of its children is several selected test sentences in support of its corresponding NVEF subclass sense-pair.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Knowledge Representation Tree of NVEF Sense-Pairs",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "To speedup the creation of the KR-tree, an example-based algorithm is proposed to generate the KR-tree semi-automatically. This algorithm is described below.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Generation of NVEF Sense-Pairs",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "Step 1. Select a noun-sense, such as \"disease| ,\" in Hownet.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Generation of NVEF Sense-Pairs",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "Step 2. Collect all Chinese polysyllabic words of the selected noun-sense. (Monosyllabic words are not considered at this stage.)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Generation of NVEF Sense-Pairs",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "Step 3. Select those Chinese un-segmented sentences that include at least one word collected in Step 2 from the Sinica corpus (which is a Chinese corpus of two millions words [CKIP 1995]) or other domain specific collections. For example, the Chinese sentence \" (A doctor's job is to prevent a disease and to cure the patient)\" is a candidate sentence that includes the Chinese word \" (disease).\"",
+                "cite_spans": [
+                    {
+                        "start": 175,
+                        "end": 187,
+                        "text": "[CKIP 1995])",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Generation of NVEF Sense-Pairs",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "Step 4. Find all possible verb-senses from the sentences selected in Step 3 to form all possible verb-senses for the selected noun-sense. Calculate the frequency for each verb-sense.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Generation of NVEF Sense-Pairs",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "Step 5. Sort all possible different verb-senses according to their corresponding frequencies from large to small. (See Fig. 2 ) Determine a cut-off frequency in the list. Among all verb-senses above the cut-off frequency, manually pick the permissible ones for the selected noun-sense. Meanwhile, determine their subclass sense-symbols (i.e. \"&,\" \"%\" and \"=\".)",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 119,
+                        "end": 125,
+                        "text": "Fig. 2",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Generation of NVEF Sense-Pairs",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "Step 6. Add these permissible NVEF subclass sense-pairs to the KR-tree.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Generation of NVEF Sense-Pairs",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "Note that among the above steps, only step 5 requires human intervention. This step is quite laborious, but through learning, human involvement can be greatly reduced. Fig. 2 shows the top 5 possible verb-senses picked by the above algorithm for the noun-sense \"disease| \" collected from 302 sentences in the Sinica corpus. In Fig. 2 , the permissible verb-senses for the noun-sense \"disease| \" are \"cure| \" with a frequency of 24, \"Cause Affect| , medical| \" with one of 23, \"Result In| \" with one of 19 and \"obstruct| \" with one of 14. It is observed that, if the number of sentences collected in Step 3 is greater than 300, then the top 5 verb-senses will almost always form NVEF sense-pairs with the selected noun-sense. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 168,
+                        "end": 174,
+                        "text": "Fig. 2",
+                        "ref_id": "FIGREF1"
+                    },
+                    {
+                        "start": 324,
+                        "end": 333,
+                        "text": "In Fig. 2",
+                        "ref_id": "FIGREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Generation of NVEF Sense-Pairs",
+                "sec_num": "2.3"
+            },
+            {
+                "text": "Based on the KR-tree, we shall develop a primitive NVEF sense-pair identifier as follows. For a given sentence, the algorithm will first identify all NVEF sense-pairs in the KR-tree that have corresponding NVEF word-pairs in the sentence. It will then arrange these NVEF sense-pairs and their corresponding NVEF word-pairs into a tree, called a sentence-NVEF tree, as shown in Fig. 3 . A more formal description of the primitive NVEF sense-pair identifier is given below:",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 377,
+                        "end": 383,
+                        "text": "Fig. 3",
+                        "ref_id": "FIGREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "A Primitive NVEF Sense-Pair Identifier",
+                "sec_num": "2.4"
+            },
+            {
+                "text": "Step 1. Input a sentence.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "A Primitive NVEF Sense-Pair Identifier",
+                "sec_num": "2.4"
+            },
+            {
+                "text": "Step 2. Generate all possible NV word-pairs of the input sentence.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "A Primitive NVEF Sense-Pair Identifier",
+                "sec_num": "2.4"
+            },
+            {
+                "text": "Step 3. Check each NV word-pair got in step 2 to see if its corresponding NV sense-pairs can be matched to an NVEF subclass sense-pair in the KR-tree. If matches are found, then use the corresponding noun-senses and verb-senses to form the permissible NVEF sense-pairs, respectively, for this sentence.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "A Primitive NVEF Sense-Pair Identifier",
+                "sec_num": "2.4"
+            },
+            {
+                "text": "Step 4. Arrange all permissible NVEF sense-pairs and their corresponding NVEF word-pairs in a sentence-NVEF tree.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "A Primitive NVEF Sense-Pair Identifier",
+                "sec_num": "2.4"
+            },
+            {
+                "text": "A system overview of the primitive NVEF sense-pair identifier is given in Fig. 4 . ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 74,
+                        "end": 80,
+                        "text": "Fig. 4",
+                        "ref_id": "FIGREF3"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "A Primitive NVEF Sense-Pair Identifier",
+                "sec_num": "2.4"
+            },
+            {
+                "text": "In Fig. 3 , the correct segmented results of the two Chinese sentences are \" / / / / \" and \" / / / ,\" respectively. The upper part of Fig. 3 is a sentence-NVEF tree with a single NVEF sense-pair, \"LandVehicle| \"-\"VehicleGo| ,\" which has two corresponding NV word-pairs, i.e. \" -\" and \" -.\" If we further apply the \"longest syllabic NVEF-word-pair first\" strategy (LS-NVWF), the incorrect NVEF word-pair \" -\" will be successfully dropped. Note that the \"longest syllabic word first strategy\" is an effective technique for Chinese word segmentation [Chen et al. 1986] . The lower part of Fig. 3 is a sentence-NVEF tree with two NVEF sense-pairs including \"expel| \"-\"livestock| \" (NV word-pair is \" -\") and \"facilities| , space| , @foster| , #livestock| \"-\"GoInto| \" (NV word-pair is \" -\").",
+                "cite_spans": [
+                    {
+                        "start": 547,
+                        "end": 565,
+                        "text": "[Chen et al. 1986]",
+                        "ref_id": "BIBREF1"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 3,
+                        "end": 9,
+                        "text": "Fig. 3",
+                        "ref_id": "FIGREF2"
+                    },
+                    {
+                        "start": 134,
+                        "end": 140,
+                        "text": "Fig. 3",
+                        "ref_id": "FIGREF2"
+                    },
+                    {
+                        "start": 586,
+                        "end": 592,
+                        "text": "Fig. 3",
+                        "ref_id": "FIGREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "An NVEF Sense-Pair Identifier",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "Another useful technique is to exclude certain nouns or verbs from the sentence-NVEF tree. A word with very low frequency as a noun or a verb is treated as a word of exclusion for the NVEF sense-pair identifier. Take the Chinese word \" (of/target)\" as an example. Its frequency as a noun or a verb is only 0.004% (computed according to the Sinica corpus). Thus, \" \" becomes a word of exclusion. In our experiment, the exclusion word list (EWL) consists of those words whose frequencies as nouns or verbs are no greater than 5%. When an NVEF word-pair includes at least one exclusion word, its corresponding NVEF sense-pair is excluded from the sentence-NVEF tree. This process is called EWL checking. Appendix B lists all of the exclusion words used in this experiment.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An NVEF Sense-Pair Identifier",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "Thus, our final NVEF sense-pair identifier can be described as follows.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An NVEF Sense-Pair Identifier",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "Step 1. Input a sentence.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An NVEF Sense-Pair Identifier",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "Step 2. Generate all possible NV word-pairs of the input sentence. Exclude certain word-pairs based on EWL checking.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An NVEF Sense-Pair Identifier",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "Step 3. Check each NV word-pair to see if its corresponding NV sense-pairs can be matched to an NVEF subclass sense-pair in the KR-tree. For each NV sense-pair that matches an NVEF subclass sense-pair in the KR-tree, use it to the set of permissible NVEF sense-pairs, respectively, for this sentence. Resolve conflicts using the LS-NVWF strategy.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An NVEF Sense-Pair Identifier",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "Step 4. Arrange all permissible NVEF sense-pairs and their corresponding NVEF word-pairs in a sentence-NVEF tree.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An NVEF Sense-Pair Identifier",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "A system overview of the NVEF sense-pair identifier is given in Fig. 5 . To evaluate the WSD performance of the NVEF sense-pair identifier, we will consider a WSD experiment in the next section.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 64,
+                        "end": 70,
+                        "text": "Fig. 5",
+                        "ref_id": "FIGREF4"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "An NVEF Sense-Pair Identifier",
+                "sec_num": "2.5"
+            },
+            {
+                "text": "Within a sentence, the number of available NVEF sense-pairs is finite. Consider the Chinese sentence \"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The WSD experiment",
+                "sec_num": "3."
+            },
+            {
+                "text": "(This car moves well).\" Table 3 gives eight possible pairs of NVEF senses found in this sentence, but there is only one permissible NVEF sense-pair, \"LandVehicle| \"-\"VehicleGo| .\" characters per sentence (of the 445 Chinese test sentences) were 4, 24 and 11.5, respectively. In addition, the numbers of single-NVEF sentences and multi-NVEF sentences among the test sentences were 96 and 349, respectively.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 24,
+                        "end": 31,
+                        "text": "Table 3",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "The WSD experiment",
+                "sec_num": "3."
+            },
+            {
+                "text": "We conducted the experiment in a progressive manner. The NVEF sense accuracy of the test sentences determined using the NVEF sense-pair identifier with only the knowledge of the KR-tree was 74.8% (see Table 4 ). When the strategy of adopting the longest syllabic NVEF-word-pair first (LS-NVWF) was used together with the NVEF sense-pair identifier, the NVEF sense accuracy reached 87.6%. When the exclusion word list (EWL checking) was adopted together with the NVEF sense-pair identifier, the NVEF sense accuracy reached 89.2%. When the techniques of both LS-NVWF and EWL checking were adopted with the NVEF sense-pair identifier (see Table 4 ), the NVEF sense accuracy improved to 93.7%. Meanwhile, along with the NVEF sense-pair identifier, the word-segmentation accuracy (for those ambiguous NVEF word-pairs) for these sentences was 99.6% (443/445). This result also supports the aforementioned claim that the NVEF word-segmentation accuracy was better than the NVEF sense accuracy. Appendix C presents two successful and one unsuccessful sentence-NVEF trees obtained in this experiment. a \"Using LS-NVWF\" represents NVEF sense accuracy using LS-NVWF with the NVEF sense-pair identifier. b \"Using EWL\" represents NVEF sense accuracy using EWL checking with the NVEF sense-pair identifier. c \"Using Both\" represents NVEF sense accuracy using both LS-NVWF and EWL checking with the NVEF sense-pair identifier.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 201,
+                        "end": 208,
+                        "text": "Table 4",
+                        "ref_id": "TABREF2"
+                    },
+                    {
+                        "start": 636,
+                        "end": 643,
+                        "text": "Table 4",
+                        "ref_id": "TABREF2"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "The WSD experiment",
+                "sec_num": "3."
+            },
+            {
+                "text": "Although the NVEF sense accuracy could reach 93.7% when the techniques of both LS-NVWF and EWL checking were adopted with the NVEF sense-pair identifier, there was still a room for improvement. Below, we have classified the reasons behind the unsuccessful cases into four major types:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An Analysis of the Unsuccessful Cases",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "(1) Lack of a bottom-up tagger: There are many specific linguistic units, such as names, addresses, determinative -measure compounds, etc. in sentences which need to be recognized in order to supplement the NVEF sense-pair identifier (which works in a top-down fashion).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An Analysis of the Unsuccessful Cases",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "In this study, 6 sentences were unsuccessful for this reason. Although the techniques of LS-NVWF and EWL checking inadvertently resolved these cases, this is still a potential problem.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An Analysis of the Unsuccessful Cases",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "include the correct NVEF word-pair for word segmentation. However, the converse is not true. That is, a correct NVEF word-pair cannot guarantee that the corresponding NVEF sense-pair is permissible. Thus, the NVEF word-segmentation accuracy is normally better than the NVEF sense accuracy.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "An Analysis of the Unsuccessful Cases",
+                "sec_num": "3.2"
+            },
+            {
+                "text": "In this paper, we have described an NVFE sense-pair identifier which we attempted to use to disambiguate word sense in Chinese sentences. A WSD experiment was conducted using the NVEF sense-pair identifier with the KR-tree. The knowledge in the KR-tree was created with the help of a semi-automatic NVEF generation tool.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions and Directions for Future Research",
+                "sec_num": "4."
+            },
+            {
+                "text": "Based on current techniques, our experiment showed that the NVEF sense accuracy reached 93.7% and the NVEF word-segmentation accuracy 99.6%. We have indicated, in Section 3, several ways to further improve the performance of our system, some of which are currently being studied.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions and Directions for Future Research",
+                "sec_num": "4."
+            },
+            {
+                "text": "Our experiment indicated that NVEF sense-pair knowledge can be used effectively to achieve NVEF word-sense disambiguation in Chinese sentences. It also supports the claim in [Choueka et al. 1983 ] that people usually disambiguate word sense using only a few words (frequently only one word) in the given context. We are particularly pleased to note that the NVEF knowledge can achieve high accuracy in NVEF word-segmentation since correct word-segmentation is one key to a successful Chinese NLP [Slocum et al. 1985] .",
+                "cite_spans": [
+                    {
+                        "start": 174,
+                        "end": 194,
+                        "text": "[Choueka et al. 1983",
+                        "ref_id": "BIBREF2"
+                    },
+                    {
+                        "start": 496,
+                        "end": 516,
+                        "text": "[Slocum et al. 1985]",
+                        "ref_id": "BIBREF7"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions and Directions for Future Research",
+                "sec_num": "4."
+            },
+            {
+                "text": "Although we have a semi-automatic NVEF generation tool, it was still a laborious task to create our current level of NVEF knowledge, which constitutes only 7.7% of the entire NVEF knowledge. Hence, a systematic method for fully automatic NVEF knowledge generation is highly desired. Furthermore, we will try to develop a combined top-down and bottom-up NVEF sense-pair identifier that can address the issues involved in the four unsuccessful cases described in Section 3.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions and Directions for Future Research",
+                "sec_num": "4."
+            },
+            {
+                "text": "We plan to create a full fledged KR-tree so that we can investigate the robustness of the sense-based approach for monolingual and bilingual (e.g. English-Chinese) WSD. The study of NVEF will also be extended to noun-noun pairs, noun-adjective pairs and verb-adverb pairs. Another related research goal is to apply the NVEF sense-pair identifier to other fields of NLP, in particular, document classification, information retrieval, question answering and speech understanding.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusions and Directions for Future Research",
+                "sec_num": "4."
+            }
+        ],
+        "back_matter": [
+            {
+                "text": "We are grateful to the our colleagues in the Intelligent Agent Systems Lab., Li-Yeng Chiu, Mark Shia, Gladys Hsieh, Masia Yu and Yi -Fan Chang, who helped us create all the necessary NVEF knowledge for this study. We would also like to thank Prof. Zhen-Dong",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Acknowledgements",
+                "sec_num": null
+            },
+            {
+                "text": "To evaluate the performance of WSD by using the NVEF sense-pair identifier with the KR-tee, we define the NVEF sense accuracy for a set of test sentences to be NVEF sense accuracy = # of successful sentences / # of test sentences,(1) where a sentence is successful if all NVEF sense-pairs and their corresponding NVEF word-pairs obtained from the NVEF sense-pair identifier are correct for this sentence. With the KR-tree, the WSD performance for the test sentences can be evaluated by computing the NVEF sense accuracy. This equation is designed from the viewpoint of natural language understanding. Since NVEF sense-pairs often represent a key feature in the meaning of a sentence, any incorrect NVEF sense-pair identification could result in misunderstanding this sentence. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "annex",
+                "sec_num": null
+            },
+            {
+                "text": "The framework of WSD evaluation for the NVEF sense-pair identifier is as follows.1.Select a set of Chinese test sentences from the Sinica Corpus [CKIP 1995] randomly.2.Use the tool of example-based possible NVEF generation to search and create all permissible NVEF subclass sense-pairs found in these test sentences in the KR-tree.3.Apply the NVEF sense-pair identifier to these test sentences and obtain their corresponding sentence-NVEF trees.4. Compute the NVEF sense accuracy for the test sentences using Equation 1.In this study, we analyzed 7.7% (=764/9,893) of the noun-senses in Hownet and created 4,028 NVEF subclass sense-pairs in the KR-tree. The minimum, maximum and mean of (2) Lack of Non-NVEF knowledge: Consider the Chinese sentence, \" (A wife wants to take her husband's wallet into her hands).\" There were three different noun-senses of the Chinese word, \"(teacher/doctor/husband),\" which could form an NVEF sense-pair with the verb-sense \" (take\u2026into one's hands).\" To get the correct noun-sense \" (husband)\" for this sentence, we need the knowledge of a noun-noun (NN) sense-pair, such as \" (wife)\"-to-\" (husband),\" or other contextual information.This knowledge is not available from the KR-tree and needs to be collected separately. In this study, 15 sentences were unsuccessful for this reason, and this problem could not be resolved using the technique of LS-NVWF or EWL checking.(3) Inadequate word segmentation: Consider the Chinese sentence, \" (He obtained the championship with a full mark).\" There were two possible verbs with the same verb-sense \" (obtain)\" and \" (obtain)\" that could form NVEF sense-pairs with the noun-sense \" (champ).\" In this case, we have two conflicting NVEF sense-pairs and need a better segmentation algorithm to determine that the correct verb are \" (obtain)\"for this sentence (the correct segmented result of this sentence is \" / / / / \").In this study, 3 sentences were unsuccessful for this reason, and this problem could not be resolved using the technique of LS-NVWF or EWL checking.(4) Lack of a multi-NVEF analyzer: Consider the Chinese sentence \" (Take airplane to leave Taipei).\" The NVEF sense-pair identifier detected that there were three NVEF sense-pairs: multi-NVEF sense-pairs. In this study, 5 sentences were unsuccessful for this reason, and this problem could not be resolved using the technique of LS-NVWF or EWL checking.If these four problems could be resolved, the NVEF sense accuracy could be improved to (417+15+3+5) / (445) = 98.9%.Based on this experiment, we find that our NVEF sense-pair identifier has the potential to provide the following information for a given sentence: (1) main verbs, (2) nouns, (3) NVEF word-pairs, (4) NVEF sense-pairs, (5) NVEF phrase-boundaries, and (6) the initial relationship among multi-NVEF sense/word-pairs. A correct NVEF sense-pair will naturally ",
+                "cite_spans": [
+                    {
+                        "start": 145,
+                        "end": 156,
+                        "text": "[CKIP 1995]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "WSD Evaluation",
+                "sec_num": "3.1"
+            }
+        ],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "The origin and evolution of everyday concepts",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Carey",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "Cognitive Models of Science",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Carey, S., \"The origin and evolution of everyday concepts (In R. N. Giere, ed.),\" Cognitive Models of Science, Minneapolis: University of Minnesota Press, 1992.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "A model for Lexical Analysis and Parsing of Chinese Sentences",
+                "authors": [
+                    {
+                        "first": "C",
+                        "middle": [
+                            "G"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "J"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "L",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Lee",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1986,
+                "venue": "Proceedings of 1986 International Conference on Chinese Computing",
+                "volume": "",
+                "issue": "",
+                "pages": "33--40",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, C. G., K. J. Chen and L. S. Lee, \"A model for Lexical Analysis and Parsing of Chinese Sentences,\" Proceedings of 1986 International Conference on Chinese Computing, Singapore, 1986, pp.33-40.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "A Connectionist Scheme for Modeling Word Sense Disambiguation",
+                "authors": [
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Choueka",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Lusignan",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1983,
+                "venue": "Cognition and Brain Theory",
+                "volume": "6",
+                "issue": "1",
+                "pages": "89--120",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Choueka, Y. and S. Lusignan, \"A Connectionist Scheme for Modeling Word Sense Disambiguation,\" Cognition and Brain Theory, 6 (1) 1983, pp.89-120.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "the content and illustration of Sinica corpus of Academia Sinica",
+                "authors": [],
+                "year": 1995,
+                "venue": "CKIP",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "CKIP. Technical Report no. 95-02, the content and illustration of Sinica corpus of Academia Sinica. Institute of Information Science, Academia Sinica, 1995. http://godel.iis.sinica.edu.tw/CKIP/r_content.html",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Lexical Ambiguity and Information Retrieval",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Krovetz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "B"
+                        ],
+                        "last": "Croft",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1992,
+                "venue": "ACM Transactions on Information Systems",
+                "volume": "10",
+                "issue": "2",
+                "pages": "115--141",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Krovetz, R. and W. B. Croft, \"Lexical Ambiguity and Information Retrieval,\" ACM Transactions on Information Systems, 10 (2), 1992, pp.115-141.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Distinguishing Systems and Distinguishing Senses: New Evaluation Methods for Word Sense Disambiguation",
+                "authors": [
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Resnik",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "Yarowsky",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Natural Language Engineering",
+                "volume": "5",
+                "issue": "3",
+                "pages": "113--133",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Resnik, P. and D. Yarowsky, \"Distinguishing Systems and Distinguishing Senses: New Evaluation Methods for Word Sense Disambiguation,\" Natural Language Engineering, 5 (3), 2000, pp.113-133.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "Transportability to Other Language: The Natural Language Processing Project in the AI Program at MCC",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Slocum",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [
+                            "F"
+                        ],
+                        "last": "Justus",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1985,
+                "venue": "ACM Transactions on Office Information Systems",
+                "volume": "3",
+                "issue": "2",
+                "pages": "204--230",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Slocum, J. and C. F. Justus, \"Transportability to Other Language: The Natural Language Processing Project in the AI Program at MCC,\" ACM Transactions on Office Information Systems, 3(2) 1985, pp.204-230.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Lexical Ambiguity Resolution",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Small",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [],
+                        "last": "Cottrell",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "E"
+                        ],
+                        "last": "Tannenhaus",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1988,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Small, S., and G. Cottrell, and M. E. Tannenhaus, Lexical Ambiguity Resolution, Morgan Kaufmann, Palo Alto, Calif., 1988.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Using WordNet to Disambiguate Word Senses for Text Retrieval",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Voorhees",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "ACM-SIGIR",
+                "volume": "",
+                "issue": "",
+                "pages": "171--180",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Voorhees, E. M., \"Using WordNet to Disambiguate Word Senses for Text Retrieval,\" ACM-SIGIR, 1993, pp. 171-180.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "An illustration of the KR-tree using \" (artifact)\" as an example noun-sense subclass. (The English words in parentheses are there for explanatory purposes only.)Three function nodes are used in the KR-tree as shown inFig. 1:(1) Major-Event ( ): The content of its parent node represents a noun-sense subclass, and the content of its child node represents a verb-sense subclass. A noun-sense subclass and a verb-sense subclass linked by a Major-Event function node is an NVEF subclass sense-pair, such as \"&LandVehicle| \" and \"=VehcileGo| \" inFig. 1. To describe various relationships between noun-sense and verb-sense",
+                "num": null
+            },
+            "FIGREF1": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "Top 5 possible verb-senses for creating permissible NVEF sense subclasses for the noun-sense \"disease| .\" Word Sense Disambiguation and Sense-Based NV Event Frame Identifier 35",
+                "num": null
+            },
+            "FIGREF2": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "Two sentence-NVEF trees for the input Chinese sentences (a) \" \" (a single-NVEF sentence) and (b) \" \" (a multi-NVEF sentence), respectively.",
+                "num": null
+            },
+            "FIGREF3": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "System overview of the primitive NVEF sense-pair identifier.",
+                "num": null
+            },
+            "FIGREF4": {
+                "uris": null,
+                "type_str": "figure",
+                "text": "A system overview of the NVEF sense-pair identifier.",
+                "num": null
+            },
+            "TABREF0": {
+                "num": null,
+                "html": null,
+                "type_str": "table",
+                "content": "<table><tr><td>Che</td><td>Noun</td><td>character|</td><td>, surname| , human| , ProperName|</td></tr><tr><td>car</td><td>Noun</td><td>LandVehicle|</td><td/></tr><tr><td>turn</td><td>Verb</td><td>cut|</td><td/></tr><tr><td colspan=\"4\">a C.Word refers to a Chinese word; E.Word refers to an English word</td></tr></table>",
+                "text": "Three different senses of the Chinese word \" (Che/car/turn).\" C.Word a E.Word a Part-of-speech Sense (i.e. DEF in Hownet)"
+            },
+            "TABREF1": {
+                "num": null,
+                "html": null,
+                "type_str": "table",
+                "content": "<table><tr><td>Item a</td><td>Total</td><td>Noun</td><td>Verb</td></tr><tr><td>Maximum number of senses per Chinese word</td><td>27</td><td>14</td><td>24</td></tr><tr><td>Mean number of senses per Chinese word</td><td>1.24</td><td>1.14</td><td>1.23</td></tr><tr><td>Maximum number of Chinese words per sense</td><td>374</td><td>372</td><td>129</td></tr><tr><td>Mean number of Chinese words per sense</td><td>3.8</td><td>3.0</td><td>4.6</td></tr></table>",
+                "text": "Statistics of the number of senses per Chinese word and the number of Chinese words per sense used in Hownet."
+            },
+            "TABREF2": {
+                "num": null,
+                "html": null,
+                "type_str": "table",
+                "content": "<table><tr><td colspan=\"4\"># of NVEF NVEF sense accuracy Using LS-NVWF a Using EWL b</td><td>Using Both c</td></tr><tr><td>4,028</td><td>74.8%(333/445)</td><td>87.6%(390/445)</td><td colspan=\"2\">89.2%(397/445) 93.7%(417/445)</td></tr></table>",
+                "text": "Results of the WSD experiment for 445 Chinese un-segmented test sentences."
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O02/O02-1003.json

@@ -0,0 +1,163 @@
+{
+    "paper_id": "O02-1003",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:05:50.030063Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "This thesis presents a description of a semantic disambiguation model applied in the syntax parsing process of the machine translation system. The model uses Hownet as its main semantic resource, which is a common-sense knowledge base unveiling inter-conceptual relations and inter-attribute relations of concepts as connoting in lexicons of the Chinese and their English equivalents. It can provide rich semantic information for our disambiguation. The model makes the word sense and structure disambiguation in the way of \"preferring\". \"preferring\" is applied in the results produced by the parsing process. It combines the rule-based method and statistic based method. First we extract from a large the co-occurrence information of each sense-atom. The corpus is untagged so the extracting process is unguided. We can construct restricted rules from the co-occurrence information according to certain transfer template. The semantic entry of a word in the Hownet is made of sense-atoms, so we can make out the restricted rules for each entry of any word.",
+    "pdf_parse": {
+        "paper_id": "O02-1003",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "This thesis presents a description of a semantic disambiguation model applied in the syntax parsing process of the machine translation system. The model uses Hownet as its main semantic resource, which is a common-sense knowledge base unveiling inter-conceptual relations and inter-attribute relations of concepts as connoting in lexicons of the Chinese and their English equivalents. It can provide rich semantic information for our disambiguation. The model makes the word sense and structure disambiguation in the way of \"preferring\". \"preferring\" is applied in the results produced by the parsing process. It combines the rule-based method and statistic based method. First we extract from a large the co-occurrence information of each sense-atom. The corpus is untagged so the extracting process is unguided. We can construct restricted rules from the co-occurrence information according to certain transfer template. The semantic entry of a word in the Hownet is made of sense-atoms, so we can make out the restricted rules for each entry of any word.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "During the course of disambiguation, the model constructs the context-related words set for each notational word in the input sentence. The semantic collocation relations between notional words can play a very important role in the syntax structure disambiguation. Our evaluation of some candidates is based on the degree of tightness of match between notional words in the structure. We compare the context-related words set of the word in the current structure with all the restricted rules of the word in the lexicon, and find the best match. Then the entry with the best match is taken as the word's explanation. And the degree of similarity shows how the word in the structure matches with other notional words in it, so it can be taken as the reference of the notional words. Because the discrepancy of different candidate parses of a structure, the same word has different content-related words set, and so will get different scores. We can calculate the best match according to \u4e00\u7a2e\u57fa\u65bc\u77e5\u7db2\u7684\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b\u7814\u7a76 49 the score of all the notional words of the sentence. In this way we can solve the most of word sense disambiguation and structural disambiguation at the same time.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "The semantic disambiguation model proposed in this thesis has been implemented in MTG system. Our experiment shows that the model is very effective for this purpose. And it is obviously more tolerant and much better than traditional YES or NO clear cut method.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "In this thesis we first put forward the general idea of the method and give a brief introduce to the Hownet Dictionary. Then we give the methods of extracting co-occurrence information for each sense-atom from the corpus and transferring this information to restricted rules. Then the algorithm of disambiguation is proposed with detail, which includes constructing context-related words set, the calculation of the similarity between atom-senses, and between restricted-rules and the context-related sets. The experiment result given in the end of the paper shows that the method is effective. (1) \u8a5e\u7684\u591a\u7fa9\uff0c\u53c8\u7a31\u8fad\u5f59\u6b67\u7fa9\uff0c\u5373\u540c\u4e00\u8a5e\u8a9e\u53ef\u80fd\u5177\u6709\u591a\u500b\uf967\u540c\u7684\u7fa9\u9805\uff1b\u5982\"\u6253\uff02\u4e00\u8a5e \u5728\"\u6253\u5b57\uff02\u3001\"\u6253\u9152\uff02\u3001\"\u6253\u7403\uff02\u3001\"\u6253\u5730\u57fa\uff02\u3001\"\u6253\u4eba\uff02\u4e2d\u5c31\u6709\uf967\u540c\u7684\u610f\u7fa9\uff1b (2) \u77ed\u8a9e\u7684\u540c\u5f62\uf962\u69cb\uff0c\u53c8\u7a31\u7d50\u69cb\u6b67\u7fa9\uff0c\u5373\u540c\u7a2e\u7d44\u5408\u537b\u542b\u6709\uf967\u540c\u7684\uf906\u6cd5\u529f\u80fd\u7d50\u69cb\u3002\u5982 \"VP+\u7684+\u662f+NP\uff02\u5c31\u662f\u4e00\u500b\u6709\u6b67\u7fa9\u7684\u7d50\u69cb\uff1a \"\u626e\u6f14\u7684\u662f\u4e00\u500b\u6f14\u54e1\uff02 50 \u694a\u66c9\u5cf0\u3001\uf9e1\u5802\u79cb \u9019\uf906\u8a71\u53ef\u4ee5\uf9e4\u89e3\u7232\"\u4e00\u500b\u6f14\u54e1\u626e\u6f14\uf9ba\u5287\u4e2d\u67d0\u500b\u89d2\u8272\uff02(\"\u626e\u6f14\u7684\uff02\u662f\u65bd\u4e8b)\uff0c\u4e5f\u53ef\u4ee5 \uf9e4\u89e3\u7232\"\u88ab\u626e\u6f14\u6210\u4e00\u500b\u6f14\u54e1\uff02(\"\u626e\u6f14\u7684\uff02\u7684\u662f\u53d7\u4e8b)\u3002 \u9084\u6709\"N1+N2+N3\uff02\uff0c\u53ef\u4ee5\u88ab\uf9e4\u89e3\u7232((N1+N2)+N3)\uff0c\u4e5f\u53ef\u4ee5\uf9e4\u89e3\u7232(N1+(N2+N3))\u3002 \u9019\uf9d0\u7684\u6b67\u7fa9\u7d50\u69cb\u5728\u6f22\u8a9e\u4e2d\u6709\u5f88\u591a\uff0c\u5b83\u4e00\u76f4\u662f\u8a9e\u6cd5\u5b78\u5bb6\u7814\u7a76\u7684\u71b1\u9ede\u554f\u984c\u3002 \u5728\u6a5f\u5668\u7ffb\u8b6f\u4e2d\uff0c\u8fad\u5f59\u6b67\u7fa9\u8868\u73fe\u7232\u8b6f\u6587\u6703\u6709\u591a\u7a2e\u7684\u9078\u64c7\uff0c\u800c\u7d50\u69cb\u6b67\u7fa9\u8868\u73fe\u7232\uf906\u6cd5\u5206\u6790 \u4e2d\uff0c\u4e00\u500b\u8a5e\u8a9e\u6216\u7247\u8a9e\u53ef\u80fd\u6703\u7523\u751f\u4e00\u500b\u4ee5\u4e0a\u7d50\u69cb\uf967\u540c\u7684\u5206\u6790\u7d50\u679c\u3002 \u6b67\u7fa9\u5728\u7279\u5b9a\u7684\u8a9e\u7fa9\u53ca\u5e38\uf9fc\u4e0b\uff0c\u4e26\uf967\u4e00\u5b9a\u90fd\u80fd\u5920\u6210\uf9f7\uff0c\uf9b5\u5982\u5728\"\u6253\u7403\uff02\u4e2d\uff0c\u6839\u64da\"\u6253\uff02 \u7684\u53d7\u4e8b\u7269\u4ef6\u6211\u5011\u53ef\u4ee5\u77e5\u9053\"\u6253\uff02\u53ea\u80fd\u9078\u64c7\"Play\uff02\u7684\u8b6f\u6587\uff1b\u800c\u5728\"\u53cd\u5c0d\u7684\u662f\u6230\u722d\uff02\u4e2d\uff0c \u6211\u5011\u4e5f\u53ef\u4ee5\u77e5\u9053\"\u6230\u722d\uff02\u662f\u53cd\u5c0d\u7684\u53d7\u4e8b\u9ad4\u800c\uf967\u53ef\u80fd\u662f\u65bd\u4e8b\u9ad4\u3002\u6d88\u9664\u9019\uf9d0\uf967\u7b26\u5408\u8a9e\u7fa9\u77e5\uf9fc \u7684\u6b67\u7fa9\u904e\u7a0b\u7a31\u7232\u6b67\u7fa9\u7684\u6d88\u89e3\uff0c\u4e5f\u7a31\u6392\u6b67\u3002\u6211\u5011\u5728\u6a5f\u5668\u7ffb\u8b6f\u7684\uf906\u6cd5\u5206\u6790\u904e\u7a0b\u4e2d\uff0c\u5fc5\u9808\u8981\u5f15 \u5165\u8a9e\u7fa9\u7684\u77e5\uf9fc\u624d\u80fd\u5920\uf901\u597d\u5730\u5b8c\u6210\u6b67\u7fa9\u7684\u81ea\u52d5\u6d88\u89e3\u3002 (1) \u5c0d\u65bc\u5c6c\u6027\u503c\u3001\uf969\uf97e\u503c\uf9d0\u7fa9\u539f\uff0c\u524d\u540c\u73fe\u96c6\u5408\u4e2d\u5be6\u9ad4\uf9d0\u7684\u7fa9\u539f\u4f5c\u7232\"EXPERIENCE\uff02\u683c\u3002 (2) \u5c0d\u65bc\u5c6c\u6027\u503c\u3001\uf969\uf97e\u503c\uf9d0\u7fa9\u539f\uff0c\u5f8c\u540c\u73fe\u96c6\u5408\u4e2d\u5be6\u9ad4\uf9d0\u7684\u7fa9\u539f\u4f5c\u7232\"THEME\uff02\u683c\uff1b (3) \u5f9e\u52d5\u4f5c\uf9d0\u7fa9\u539f\u7684\u524d\u5f8c\u540c\u73fe\u96c6\u5408\u4e2d\u53d6\u51fa\"implement|\u5668\u5177\uff02\u53ca\u5176\u6240\u6709\u7684\u4e0b\u4f4d\u7fa9\u539f\uff0c\u5c07 \u5b83\u5011\u4f5c\u7232\"INSTRUMENT\uff02\u683c\uff1b (4) \u5f9e\u52d5\u4f5c\uf9d0\u7fa9\u539f\u7684\u524d\u5f8c\u540c\u73fe\u96c6\u5408\u4e2d\u53d6\u51fa\"earth|\u5927\u5730\uff02\u3001\"place|\u5730\u65b9\uff02\u3001\"space| \u7a7a\u9593\uff02\u53ca\u5176\u6240\u6709\u7684\u4e0b\u4f4d\u7fa9\u539f\uff0c\u5c07\u5b83\u5011\u4f5c\u7232\"LOCATION\uff02\u683c\uff1b (5) \u5f9e\u52d5\u4f5c\uf9d0\u7fa9\u539f\u7684\u524d\u5f8c\u540c\u73fe\u96c6\u5408\u4e2d\u53d6\u51fa\"time|\u6642\u9593\uff02\u53ca\u5176\u6240\u6709\u7684\u4e0b\u4f4d\u7fa9\u539f\uff0c\u5c07\u5b83\u5011\u4f5c \u7232\"TIME\uff02\u683c\uff1b (6) \u5f9e\u52d5\u4f5c\uf9d0\u7fa9\u539f\u7684\u524d\u5f8c\u540c\u73fe\u96c6\u5408\u4e2d\u53d6\u51fa\"degree|\u7a0b\ufa01\uff02\u3001 \"range|\u5e45\ufa01\uff02\u3001 \" frequency|\u983b\uf961\uff02\u53ca\u5176\u6240\u6709\u7684\u4e0b\u4f4d\u7fa9\u539f\uff0c\u5c07\u5b83\u5011\u5206\u5225\u4f5c\u7232\"DEGREE\uff02\u3001 \"RANGE\uff02\u3001\"FREQUENCEY\uff02\u683c\uff1b (7) \u5f9e\u52d5\u4f5c\uf9d0\u7fa9\u539f\u7684\u5f8c\u540c\u73fe\u96c6\u5408\u4e2d\u53d6\u51fa\u7684\u5c6c\u6027\u503c\u3001\uf969\uf97e\u503c\uf9d0\u7fa9\u539f\uff0c\u5c07\u5b83\u5011\u4f5c\u7232\"RESULT\uff02 \u683c\uff1b (8) \u5c07\u52d5\u4f5c\uf9d0\u7fa9\u539f\u7684\u524d\u540c\u73fe\u96c6\u5408\u9918\u4e0b\u7684\u5be6\u9ad4\uf9d0\u7fa9\u539f\u4f5c\u7232\"AGENT\uff02\u683c\uff1b (9) \u5c07\u52d5\u4f5c\uf9d0\u7fa9\u539f\u7684\u5f8c\u540c\u73fe\u96c6\u5408\u4e2d\u7684\u9918\u4e0b\u7684\u5be6\u9ad4\uf9d0\u7fa9\u539f\u4f5c\u7232 THEME \u683c\uff1b \uf9b5\u5982\u5c0d\u65bc\u52d5\u4f5c\uf9d0\u7fa9\u539f\"eat|\u5403\uff02\uff0c\uf9dd\u7528\u8f49\u5316\u6a21\u677f\u5f97\u5230\u7684\u7fa9\u539f\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u7232\uff1a \u91cd\u65b0\u8207\"\u53e0 1\uff02\u7684\u898f\u5247\u9032\ufa08\u6bd4\u8f03\u8a08\u7b97\uff0c\u8abf\u6574\u5f8c\"\u88ab\u5b50\uff02\u5728\"\u53e0\uff02\u4e0b\u7684\u8a55\u50f9\u5206\u503c\u9054\u5230 \uf9ba 100 \u5206\uff0c\u9ad8\u65bc\u8abf\u6574\u95be\u503c\uff0c\u6211\u5011\u5c31\u53ef\u4ee5\u8a8d\u7232\u5728\u6b64\u7d50\u679c\u6846\u67b6\u4e2d\u7684 AGENT \u61c9\u8abf\u6574\u7232 THEME\u3002 3.5.2 \u78ba\u5b9a\u5206\u6790\u7d50\u679c\u7684\u8a9e\u7fa9\u95dc\u4fc2 \u6f22\u8a9e\u4e2d\u5b58\u5728\u67d0\u4e9b\u7279\u6b8a\u7684\uf906\u578b\uff0c\u5982\"VP+NP\uff02\u3001\"VP+\u7684+\u662f+NP\uff02\u6216\"VP+\u7684+NP\uff02\uff0c\u6216 \u662f\"NP+VP\uff02\u3001\"NP+\u7684+VP\uff02 \u7b49\u504f\u6b63\u7d50\u69cb\uff0c\u5176\u4e2d\u7684 NP \u5728 VP \u7684\u8a9e\u7fa9\u74b0\u5883\u4e2d\u6240\u5145\u7576\u7684 \u8a9e\u7fa9\u95dc\u4fc2\u683c\uf967\u662f\u56fa\u5b9a\u7684\u3002\u6211\u5011\u628a NP \u7576\u4f5c VP \u8a9e\u7fa9\u74b0\u5883\u7684 Parent \u683c\u8a5e\u8a9e\u3002NP \u53ef\u4ee5\u4f5c VP \u7684\u4efb\u4f55\u4e00\u500b\uf967\u5b58\u5728\u7684\u8a9e\u7fa9\u6210\u5206\u3002\uf9b5\u5982\uff1a 1\uff1a\u88dd\u4fee\u5716\u66f8\u9928\u7684\u5de5\u4eba\u6574\u6574\u5fd9\uf9ba\u4e00\u5929\u3002(\"\u5de5\u4eba\uff02\u505a\"\u88dd\u4fee\uff02\u7684 AGENT) 2\uff1a\u5929\u4e0b\u54ea\u6709\u767d\u5403\u7684\u5348\u9910\uff1f(\"\u5348\u9910\uff02\u4f5c\"\u5403\uff02\u7684 THEME) 3\uff1a\u6211\u5011\u4ee5\u524d\u4f4f\u7684\u5730\u65b9\u73fe\u5728\u662f\u7247\u5ee3\u5834\uf9ba\u3002(\"\u5730\u65b9\uff02\u4f5c\"\u4f4f\uff02\u7684 LOCATION) 4\uff1a\u5979\u7684\u6253\u64fe\u6703\u4f7f\u4ed6\u611f\u5230\u53ad\u7169\u3002(\"\u5979\uff02\u4f5c\"\u6253\u64fe\uff02\u7684 AGENT) 5\uff1a\u611f\u8b1d\u4f60\u5c0d\u85dd\u8853\u4e8b\u696d\u7684\u5927\uf98a\u652f\u6301\uff01(\"\u4e8b\u696d\uff02\u4f5c\"\u652f\u63f4\uff02\u7684 THEME) \u5f9e\u4e0a\u9762\u7684\uf9b5\u5b50\u6211\u5011\u53ef\u4ee5\u770b\u5230\uff0cNP \u5728 VP \u4e2d\u53ef\u4ee5\u5145\u7576\u7684\u6210\u5206\u662f\u5f88\u8c50\u5bcc\u7684 [1] \u3002\u5728\u9032\ufa08\u8a5e \u7fa9\u9078\u64c7\u6642\uff0c\u6211\u5011\u6839\u64da\u8a5e\u8a9e\u7684\u8a9e\u7fa9\u74b0\u5883\u8207\u7fa9\u9805\u9650\u5236\u898f\u5247\u9032\ufa08\u76f8\u4f3c\ufa01\u8a08\u7b97\uff0c\u9700\u8981\u78ba\u5b9a\u9019\u500b NP \u5728 VP \u8a9e\u7fa9\u74b0\u5883\u4e2d\u7684\u8a9e\u7fa9\u683c\u3002 \u8207 3.5.1 \u7684\u601d\uf937\u76f8\u540c\uff0c\u6211\u5011\u4e5f\u53ef\u4ee5\u63a1\u7528\u6e2c\u8a66\u6bd4\u8f03\u6cd5\uf92d\u89e3\u6c7a NP \u7684\u6210\u5206\u78ba\u5b9a\u554f\u984c\u3002\u4e0b\u9762 \u4e3b\u8981\u91dd\u5c0d\"VP+DE+NP\uff02\u6216\"VP+NP\uff02\uf9d0\u7684\u5b9a\u4e2d\u7d50\u69cb\u9032\ufa08\u8a0e\uf941\uff0c\u5c0d\u65bc\"NP+VP\uff02\u6216 \"NP+\u7684+VP\uff02\u578b\u7684\u7d50\u69cb\uff0c\u4e5f\u53ef\u540c\u6a23\u9032\ufa08\u8655\uf9e4\u3002 \u8a2d VP+DE+NP \u6216 VP+NP \u7d50\u69cb\u4e2d VP \u7684\u4e2d\u5fc3\u52d5\u8a5e Word \u7684\u8a9e\u7fa9\u74b0\u5883\u7232 W\uff1d((PARENT PWORD)(CASE 1 word 1 )\u2026\u2026(CASE n word n )) ",
+                "cite_spans": [
+                    {
+                        "start": 2180,
+                        "end": 2183,
+                        "text": "[1]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u7684 \u76f8 \u4f3c \ufa01 \u8a08 \u7b97 \uf92d \u9078 \u64c7 \u6700 \u4f73 \u7684 \u5339 \u914d \u3002 \u9019 \uf9d0 \u65b9 \u6cd5 \u4e2d \u6bd4 \u8f03 \u6210 \u529f \u7684 \u7cfb \u7d71 \u6709 \u65b0 \u52a0 \u5761 \u7684 LEXAS(",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Atoms(k) ) ( \u222a W F Window k \u2212 \u2208 COUNT-F[a,b]=COUNT-F[a,b]+1\uff1b } for each a\u2208Atoms(W)\uff0c\u8655\uf9e4\u5f8c\u540c\u73fe\u96c6\u5408\uff1a { for each b\u2208 Atoms(k) ) ( \u222a W F Window k \u2212 \u2208 \u90fd\u505a COUNT-B[a,b]=COUNT-B[a,b]+1\uff1b } } \u5728\u5c0d\u8a9e\uf9be\u5eab\u4e2d\u7684\uf906\u5b50\u8655\uf9e4\u5b8c\u7562\u5f8c\uff0c\u53ef\u4ee5\u8a08\u7b97\u51fa\u6bcf\u500b\u7fa9\u539f\u7684\u540c\u73fe\u7fa9\u539f\u96c6 S-(A)= {(b freq(a,b)) | freq(a,b)=\u03b31*count-F[a,b]/Total, b\u2208ATOMSSET } S+(A)= {(b freq(a,b)) | freq(a,b)=\u03b32*count-B[a,b]/Total, b\u2208ATOMSSET } \u694a\u66c9\u5cf0\u3001\uf9e1\u5802\u79cb \u5176\u4e2d \u03b3 1 , \u03b3 2 \u5206\u5225\u662f\u524d\u5f8c\u540c\u73fe\u96c6\u5408\u4e2d\u540c\u73fe\u6982\uf961\u7684\u653e\u5927\u4fc2\uf969\u3002\u592a\u5c0f\u7684\u6982\uf961\u96e3\u4ee5\u8868\u73fe\u540c\u73fe\u7fa9",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "( (CROOT \u6015) (CAT V) (AGENT ((HUMAN +) (CAT PRON)(AGREE SG) (PERSON FIRST) (CROOT \u6211))) (THEME ( (CROOT \uf943) (AGENT ( (HUMAN +)(CAT PRON) (AGREE SG) (PERSON SECOND) (CROOT \u4f60))) (THEME ((CAT N) (CROOT \u7b46\u5c16))) (RESULT ((CROOT \u65b7) (CAT V))) )) ) \u5728\u683c\u7d50\u69cb\u4e2d\u52d5\u8a5e\u6216\u5f62\u5bb9\u8a5e\u7684\u8a9e\u7fa9\u74b0\u5883\u662f\u7531 Agent\u3001Theme\u3001Result\u3001Clause \u7b49\u683c\u8cc7\u8a0a \uf92d\u63cf\u8ff0\u3002\u7531\u65bc\u7f3a\u5c11\u8a9e\u7fa9\u8cc7\u8a0a\uff0c\u5728\u5206\u6790\u968e\u6bb5\u7d66\u51fa\u7684\u8a9e\u7fa9\u683c\u8cc7\u8a0a\u4e26\uf967\u4e00\u5b9a\u6b63\u78ba\uff0c\u6211\u5011\u53ef\u4ee5\u5728 \u8a5e\u7fa9\u6392\u6b67\u968e\u6bb5\u9032\ufa08\u81ea\u52d5\u7684\u8a9e\u7fa9\u683c\u8abf\u6574\u3002 \u5177\u9ad4\u7684\u7fa9\u539f\u7684\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u7684\u6a21\u578b\u8868\u793a\u7232\uff1a Rule = (SenseAtom Rule-Items ) Rule-Items=(Logic-Op {(Rule-Items)}+) | {(Sem-Case Logic-Item)}+ Logic-Item = ( Logic-Op {Logic-Item}+ ) | { Sense-Item}+ Sem-Case = Agent | Theme | CO-THEME | Result |",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u8f49\u63db\u6a21\u677f\u7232\u6211\u5011\u69cb\u9020\uf9ba\u4e00\u500b\u521d\u59cb\u898f\u5247\u5eab\uff0c\u898f\u5247\u5eab\u4e2d\u5b9a\u7fa9\uf9ba\u7fa9\u539f Agent\u3001Theme\u3001 Result\u3001Instrument \u7b49\u683c\u7684\u9650\u5236\u63cf\u8ff0\u3002\u5c0d\u65bc\u8a9e\u6cd5\u6027\u8cea\u7c21\u55ae\u7684\u7fa9\u539f\uff0c\u9019\u4e9b\u683c\u63cf\u8ff0\u5df2\u7d93\u8db3\u5920\u3002 \u4f46\u662f\u5c0d\u65bc\u5927\u591a\uf969\u7fa9\u539f\u800c\u8a00\uff0c\u9019\u4e9b\u81ea\u52d5\u751f\u6210\u7684\u898f\u5247\u5c31\u904e\u65bc\u7c21\u55ae\uf9ba\uff0c\u56e0\u6b64\u6211\u5011\u9700\u8981\u5728\u521d\u59cb\u898f \u5247\u7684\u57fa\u790e\u4e0a\u624b\u5de5\u5c0d\u5176\u9032\ufa08\u4fee\u6539\u8207\u8abf\u6574\uff0c\u52a0\u4e0a\u5fc5\u8981\u7684\u683c\u63cf\u8ff0\u3001\u5254\u9664\u932f\u8aa4\u7684\u7279\u5fb5\u7fa9\u539f\u3002\uf9b5\u5982\uff0c \u5c0d\u65bc\u5982\"URGE|\u4fc3\u4f7f\uff02\uff0c\u7531\u8a72\u7fa9\u539f\u5b9a\u7fa9\u7684\u8a5e\u8a9e\u5982\"\u4fc3\u4f7f\uff02\u3001\"\u63a8\u52d5\uff02\u3001\"\u9f13\uf97f\uff02\u7b49\u8a5e\u8a9e\uff0c \u4e00\u822c\u90fd\u5e36\u6709\u517c\u8a9e\uff0c\u56e0\u6b64\u6211\u5011\u8981\u7232\u5176\u589e\u52a0\u517c\u8a9e\u8868\u793a\u7684\u683c(EVENT)\u7684\u9650\u5236\u63cf\u8ff0\uff1b\u9084\u6709\u5982\u7531 \" GIVE| \u7d66 \uff02 \u5b9a \u7fa9 \u7684 \u8a5e \u8a9e \uff0c \u4e00 \u822c \u90fd \u5e36 \u6709 \u96d9 \u8cd3 \u8a9e \uff0c \u6211 \u5011 \u4e5f \u8981 \u7232 \" GIVE| \u7d66 \uff02 \u5b9a \u7fa9 CO-THEME(\u9593\u63a5\u8cd3\u8a9e)\u683c\u7684\u9650\u5236\u3002\u53e6\u8655\u6211\u5011\u4e5f\u9700\u7232\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Weight(i) = 2*(Depth -i) / (Depth*(Depth+1)) (3.2) \u800c\u7fa9\u539f A\uff0cB \u7684\u8a9e\u7fa9\u76f8\u4f3c\ufa01\u7232\uff1a (3.3) ; 100 * B)) ATOM(A, - DISTANCE - (1 ; 0 B) ATOMS(A, SIM B A, \u23a9 \u23a8 \u23a7 = \u2212 \u5426\u5219 \uf967\u5728\u540c\u4e00\u68f5\u5206\u7c7b\u6811 \u5982\u679c\u4e49\u539f \u5176\u4e2d MAXDISTANCE \u662f\u7fa9\u539f A \u6240\u5728\u5206\uf9d0\u6a39\u7684\u6700\u9577\u7684\u8a9e\u7fa9\u8ddd\uf9ea\u3002 \u4ee5\u4e0a\u5b9a\u7fa9\u7684\u8a9e\u7fa9\u8ddd\uf9ea\u8207\u8a9e\u7fa9\u76f8\u4f3c\ufa01\u4e2d\u7684 A\uff0cB \u662f\u53ef\u4ea4\u63db\u7684\uff0c\u5373 A \u8207 B \u7684\u8a9e\u7fa9\u76f8\u4f3c\ufa01 \u7b49\u65bc B \u8207 A \u7684\u8a9e\u7fa9\u76f8\u4f3c\ufa01\u3002\u4f46\u662f\u6211\u5011\u5c07\u6703\u770b\u5230\uff0cSIM-ATOMS \u4e2d\u6bd4\u8f03\u7684\u662f\u6a21\u5f0f\u7fa9\u539f A \u8207 \u5be6\u969b\u7fa9\u539f B \u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\u3002\u5982\u679c\u5be6\u969b\u7fa9\u539f B \u662f\u6a21\u5f0f\u7fa9\u539f A \u7684\u4e0b\u4f4d\uff0c\u5247\u5b83\u5011\u7684\u8a9e\u7fa9\u8ddd\uf9ea\u61c9 \u6bd4\u8f03\u5c0f\uff0c\u5982\u679c\u898f\u5247\u7fa9\u539f\u5728\u5be6\u969b\u7fa9\u539f\u7684\u4e0b\u4f4d\u6216\u662f\u5b83\u5011\u53ea\u662f\u64c1\u6709\u67d0\u500b\u76f8\u540c\u7956\u5148\u7684\uf978\u500b\u7d50\u9ede\uff0c \u5247\u8a9e\u7fa9\u8ddd\uf9ea\u61c9\u8f03\u5927\u3002\u9019\u6a23\u6211\u5011\u61c9\u5c0d\u8a9e\u7fa9\u8ddd\uf9ea\u51fd\uf969\u4fee\u6539\u5982\u4e0b\uff1a ) (3.1' 2 / )) ( ) ( * ( MAXVALUE B) ATOM(A, - DISTANCE ; ; 1 1 \u23aa \u23a9 \u23aa \u23a8 \u23a7 + = \u2211 \u2211 + = + = \u5426\u5219 \uf967\u5728\u540c\u4e00\u68f5\u5206\u7c7b\u6811\u4e0a B \u5982\u679c\u4e49\u539fA, a c i b c j j Weight i Weight m \u5728\u5be6\u969b\u8a08\u7b97\u4e2d\uff0c\u6211\u5011\u53ef\u4ee5\u8b93 m \u7684\u503c\u53d6\u5f97\u8db3\u5920\u5927\uff0c\u4f7f\u5f97\u7576\u7fa9\u539f B \u662f\u7fa9\u539f A \u7684\u5b50\u5b6b\u7d50 \u9ede\u6642\u8a9e\u7fa9\u8ddd\uf9ea\u8f03\u5c0f\uff0c\u5426\u5247\u5c07\u5f97\u5230\u4e00\u500b\u8f03\u5927\u7684\u8a9e\u7fa9\u8a9e\uf9ea\u3002\u6839\u64da\u5be6\u9a57\uff0cM \u503c\u8a2d\u5b9a\u7232 5 \u80fd\u5920\u53d6 \u5f97\u6bd4\u8f03\u597d\u7684\u6392\u6b67\u6548\u679c\u3002 2)\u8a5e\u8a9e\u7fa9\u9805\u8207\u898f\u5247\u7684\u683c\u9650\u5236\u63cf\u8ff0\u7684\u76f8\u4f3c\ufa01 \u7fa9\u9805\u7684\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u4e2d\u5b9a\u7fa9\uf9ba\u67d0\u4e00\u8a9e\u7fa9\u683c\u53ef\u80fd\u51fa\u73fe\u7684\u7279\u5fb5\u7fa9\u539f\u7684\uf913\u8f2f\u7d44\u5408\uff0c\u800c\u8a5e\u8a9e\u7684\u7fa9 \u9805\u662f\u7531\u7fa9\u539f\u53ca\u8a9e\u7fa9\u95dc\u4fc2\u69cb\u6210\u7684\u3002\u6211\u5011\u5e0c\u671b\u80fd\u80fd\u5920\u5224\u5b9a\u4e00\u500b\u7fa9\u9805\u6eff\u8db3\u898f\u5247\u7684\u683c\u9650\u5236\u63cf\u8ff0\u7684 \u694a\u66c9\u5cf0\u3001\uf9e1\u5802\u79cb \u7a0b\ufa01\uff0c\u5373\u7fa9\u9805\u8207\u898f\u5247\u7684\u683c\u63cf\u8ff0\u7684\u76f8\u4f3c\ufa01\u3002 \u8a2d\u67d0\u7fa9\u9805\u8a9e\u7fa9\u898f\u5247\u4e2d\u898f\u5b9a\uf9ba\u67d0\u683c\u7684\u9650\u5236\u63cf\u8ff0\u7232 C\uff0c C=(OP (R 1 S 1 ) (R 2 S 2 )\u2026(R m S m ) ,CR 1 ,CR 2 ,CR z )\u3002 \u5176\u4e2d\u5c0d\u65bc 1\u2264i\u2264m,R' i \u2208RelationSet\uff0cS i \u662f\u63cf\u8ff0\u4e2d\u5b57\u9996\u8a9e\u7fa9\u95dc\u4fc2\u7232 R' i \u7684\u7279\u5fb5\u7fa9\u539f\u3002\u63cf\u8ff0\u4e2d\u53ef \u80fd\u542b\u6709\u7121\u5b57\u9996\u8a9e\u7fa9\u95dc\u4fc2\u7684\u7fa9\u539f\uff0c\u5b83\u5011\u672c\u8eab\u5c31\u662f\u683c\u63cf\u8ff0\u7684\u7279\u5fb5\u8a9e\u7fa9\u5c6c\u6027\uff0c\u7232\uf9ba\u4fdd\u6301\u8a08\u7b97\u7684 \u7d71\u4e00\u6027\uff0c\uf967\u59a8\u5047\u5b9a\u9019\u4e9b\u7fa9\u539f\u7684\u5b57\u9996\u8a9e\u7fa9\u95dc\u4fc2\u7232 Property\u3002OP \u662f\uf913\u8f2f\u7d44\u5408\u904b\u7b97\u7b26\uff0c\u5305\u62ec *AND*,*OR*,*NOT*\uff0c\u5b83\u5011\u7684\u529f\u80fd\u5728\u524d\u9762\u5df2\uf96f\u660e\u904e\u3002CR i (1\u2264i\u2264z)\u662f\u5e36\u6709\uf913\u8f2f\u7d44\u5408\u904b\u7b97\u7b26 \u7684\u5d4c\u5957\u8a9e\u7fa9\u63cf\u8ff0\u96c6\u3002 \u73fe\u6709\u4e00\u8a5e\u8a9e\u7fa9\u9805 Entry=((R' 1 E 1 ) (R' 2 E 2 )\u2026(R' n E n )) \u5c0d\u65bc 1\u2264i\u2264n \u6709 E i \u2282 AtomSets\uff0c R i \u2208RelationSet\u3002\u540c\uf9e4\uff0c\u5c0d\u65bc\u7fa9\u9805\u4e2d\u7121\u5b57\u9996\u8a9e\u7fa9\u95dc\u4fc2\u7684\u7fa9\u539f\uff0c\u6211\u5011\u4e5f\u5047\u5b9a\u5176\u5b57\u9996\u8a9e\u7fa9\u95dc\u4fc2 \u7232 Property\u3002 \u8a2d(R S)\u662f C \u4e2d\u7684\u4e00\u500b\u5143\u7d20\uff0c\u5176\u4e2d S=( (atom 1 prob 1 ) (Atom 2 Prob 2 )\u2026(Atom m Prob m ) )\uff0c \u5b9a\u7fa9\u51fd\uf969 (3.4) \u23aa \u23a9 \u23aa \u23a8 \u23a7 = \u2264 \u2264 \u00d7 \u2208 \u2208 = ; prob a) ATOMS(atom - SIM E' a S, prob) (atom MAX Entry) , S) , EM((R RELATIONIT - ENTRY - SIM 0 R R' , 1 , ; i , i \u5426\u5219 \u6709 \u5982\u679c\u5b58\u5728 n i i \u8981\u5f97\u5230 C \u8207 Entry \u7684\u76f8\u4f3c\ufa01\uff0c\u53ef\u4ee5\u5c07 OP \u7684\u904b\u7b97\u5143\u4f9d\u6b21\u8207 Entry \u9032\ufa08\u76f8\u4f3c\ufa01\u7684\u8a08\u7b97\uff0c \u4e26\u6839\u64da OP \u7684\u503c\u5f9e\u8a08\u7b97\u7d50\u679c\u4e2d\u6311\u9078\u51fa\u5408\u9069\u7684\u7d50\u679c\u3002\u6ce8\u610f\u5c0d\u65bc CR i (1\u2264i\u2264z)\uff0c\u9700\u8981\u905e\u8ff4\u5730\u9032\ufa08 \u8a08\u7b97\u3002 \u6211\u5011\u5b9a\u7fa9\u96c6\u5408 RS \u662f OP \u7684\u6bcf\u4e00\u500b\u904b\u7b97\u5143\u8207 Entry \u6bd4\u8f03\u7684\u76f8\u4f3c\ufa01\u96c6\u5408\uff0c\u5373 RS={SIM-ENTRY-RELATIONITEM( (R i , S i ), Entry) | 1\u2264i\u2264m} \u222a { SIM-Entry-SC(Entry , CRj) |",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "V) (CROOT \u4fee) (AGENT ((CAT N) (CROOT \u7238\u7238))) (THEME ((CAT N) (CROOT \u96fb\u8996) (ATTRIBUTE ( (CAT ADJ) (CROOT \u820a))) ))) (1) \u5c07\u4e2d\u9593\u7d50\u679c\u8868\u793a\u7232\u4f9d\u5b58\u95dc\u4fc2\u6a39\u7684\u5f62\u5f0f\uff1a (2) \u78ba\u5b9a\u52d5\u8a5e\u7684\u4e0a\u4e0b\u6587\u8a9e\u7fa9\u74b0\u5883\uff1a (\u4fee ((AGENT \u7238\u7238 N) (THEME \u96fb\u8996 N))) (\u820a ((THEME \u96fb\u8996 N))) (3) \u78ba\u5b9a\u4e2d\u9593\u8a9e\u8a00\u7d50\u69cb\u4e2d\u5404\u8a9e\u7fa9\u7684\u7fa9\u9805\u898f\u5247 Theme \u7238\u7238(N) \u96fb\u8996(N) \u4fee(V",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u5176\u4e2d PWORD \u662f VP \u77ed\u8a9e\u7684\u4fee\u98fe\u8a5e\u8a9e\u3002 74 \u694a\u66c9\u5cf0\u3001\uf9e1\u5802\u79cb \u73fe\u6709 W \u7684\u4e00\u500b\u7fa9\u9805\u898f\u5247 R\uff1a R\uff1d((CASE' 1 CASESC 1 )(CASE' 2 CASESC 2 ) \u2026(CASE' m CASESC m )) \u6211\u5011\u53ef\u4ee5\u5c07 W \u4e2d\u7684 PARENT \u683c\u4f9d\u6b21\u66ff\u63db\u7232\u5728 R \u4e2d\u5b58\u5728\u800c\u5728 W \u4e2d\uf967\u5b58\u5728\u7684\u683c\uff0c\u7576\u7136 \u66ff\u63db\u7684\u683c\u5fc5\u9808\uf967\u9055\u53cd W \u4e2d\u7684\u67d0\u4e9b\u5c0d\u6210\u5206\u8981\u6c42\u7684\u8a9e\u6cd5\u9650\u5236\uff0c\uf9b5\u5982\u7576 W \u662f\uf967\u53ca\u7269\u52d5\u8a5e\u6642\uff0c Parent \u5c31\uf967\u80fd\u66ff\u63db\u6210\u7232 THEME \u683c\u3002Parent \u66ff\u63db\u5f8c\u5f97\u5230\u7684\u65b0\u8a9e\u7fa9\u74b0\u5883 W'\u8207 R \u9032\ufa08\u76f8\u4f3c\ufa01 \u7684\u8a08\u7b97\u3002\u8a2d\u7576 Parent \u63db\u6210 Case i \u6642\u53d6\u5f97\u6700\u5927\u7684\u76f8\u4f3c\ufa01\u503c V\uff0c\u4e14 V \u5927\u65bc\u4e00\u500b\u9810\u5b9a\u7fa9\u7684\u95be\u503c\uff0c \u5247\u6211\u5011\u53ef\u4ee5\u78ba\u5b9a Parent \u5728 VP \u4e2d\u5145\u7576 Case i \u7684\u8a9e\u7fa9\u6210\u5206\u3002 \u5982\u679c{CASE' 1 , Case' 2 , \u2026,Case' n }-{Case 1 ,Case 2 ,\u2026,Case m }= \u03a6 \uff0c\u6216\u8005 V",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [
+            {
+                "text": "\u8f38\u5165\uff1a\u8a5e\u8a9e Word \u53ca Word \u7684\u4e0a\u4e0b\u6587\u8a9e\u5883 ENV\uff0c\u5176\u4e2d ENV\uff1d((CASE' 1 WORD 1 CAT 1 )(CASE' 2 WORD 2 CAT 2 )\u2026(CASE' n WORD n CAT n )) ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "annex",
+                "sec_num": null
+            }
+        ],
+        "bib_entries": {},
+        "ref_entries": {
+            "FIGREF0": {
+                "uris": null,
+                "num": null,
+                "type_str": "figure",
+                "text": "*OR* (MEDICINE|\u85e5\u7269 0.52)(PART|\u90e8\u4ef6 0.38) (HUMAN|\u4eba 0.20) (FOOD|\u98df\u54c1 0.15) \u2026 ) (RESULT (*OR* ((ATTRIBUTE|\u5c6c\u6027 0.26) (DESIRED|\uf97c 0.20) ) )"
+            },
+            "TABREF0": {
+                "html": null,
+                "num": null,
+                "type_str": "table",
+                "content": "<table><tr><td>\u6458</td><td>\u8981</td></tr><tr><td colspan=\"2\">\u672c\u6587\u63d0\u51fa\uf9ba\u6a5f\u5668\u7ffb\u8b6f\u4e2d\uf906\u6cd5\u5206\u6790\u7684\u4e00\u7a2e\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b\uff0c\u8a72\u6a21\u578b\u4ee5\u300a\u77e5\u7db2\u300b\u7232\u4e3b</td></tr><tr><td colspan=\"2\">\u8981\u8a9e\u7fa9\u77e5\uf9fc\u6e90\u3002 \u300a\u77e5\u7db2\u300b \u662f\u4e00\u500b\u4ee5\u6f22\u8a9e\u548c\u82f1\u8a9e\u7684\u8a5e\u8a9e\u6240\u4ee3\u8868\u7684\u6982\uf9a3\u7232\u63cf\u8ff0\u7269\u4ef6\uff0c</td></tr><tr><td colspan=\"2\">\u4ee5\u63ed\u793a\u6982\uf9a3\u8207\u6982\uf9a3\u4e4b\u9593\u4ee5\u53ca\u6982\uf9a3\u6240\u5177\u6709\u7684\u5c6c\u6027\u4e4b\u9593\u7684\u95dc\u4fc2\u7232\u57fa\u672c\u5167\u5bb9\u7684\u5e38\uf9fc</td></tr><tr><td colspan=\"2\">\u77e5\uf9fc\u5eab,\u5b83\u7232\u6211\u5011\u7684\u6392\u6b67\u63d0\u4f9b\uf9ba\u8c50\u5bcc\u7684\u8a9e\u7fa9\u8cc7\u8a0a\u3002\u6392\u6b67\u6a21\u578b\u7d50\u5408\uf9ba\u57fa\u65bc\u898f\u5247\u53ca</td></tr><tr><td colspan=\"2\">\u57fa\u65bc\u7d71\u8a08\u7684\u65b9\u6cd5\uff0c\u61c9\u7528\u65bc\u5206\u6790\u6240\u7523\u751f\u7684\u4e2d\u9593\u7d50\u69cb\u4e2d\uff0c\u5f9e\"\u512a\u9078\uff02\u7684\u89d2\ufa01\u9032\ufa08\u8a5e</td></tr><tr><td>\u7fa9\u53ca\u7d50\u69cb\u7684\u6392\u6b67\u3002</td><td/></tr><tr><td colspan=\"2\">\u6392\u6b67\u6a21\u578b\u9996\u5148\uf9dd\u7528\u5927\u898f\u6a21\u7684\u8a9e\uf9be\u5eab\u7372\u53d6\u7fa9\u539f\u7684\u540c\u73fe\u96c6\u5408\uff0c\u8a72\u8a9e\uf9be\u5eab\u672a\u9032\ufa08\u4efb\u4f55</td></tr><tr><td colspan=\"2\">\u7684\u8a9e\u7fa9\u6a19\u8a8c\uff0c\u56e0\u6b64\u7372\u53d6\u904e\u7a0b\u662f\u7121\u6307\u5c0e\u7684\u3002\u7136\u5f8c\u5b83\u6839\u64da\u8f49\u63db\u6a21\u677f\u69cb\u9020\u51fa\u7fa9\u539f\u7684\u8a9e</td></tr><tr><td colspan=\"2\">\u7fa9\u9650\u5236\u898f\u5247\u3002\u300a\u77e5\u7db2\u300b\u4e2d\u7684\u8a5e\u8a9e\u7fa9\u9805\u7531\u7fa9\u539f\u7d44\u6210\uff0c\u7fa9\u9805\u7684\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u53ef\u4ee5\u7531</td></tr><tr><td>\u5176\u69cb\u6210\u7fa9\u539f\u7684\u8a9e\u7fa9\u898f\u5247\u5f97\u5230\u3002</td><td/></tr><tr><td colspan=\"2\">\u5728\u8a9e\u7fa9\u6392\u6b67\u968e\u6bb5\uff0c\u6211\u5011\u9996\u5148\u78ba\u5b9a\u8f38\u5165\uf906\u7684\u6bcf\u500b\u5be6\u7fa9\u8a5e\u7684\u4e0a\u4e0b\u6587\u76f8\u95dc\u8a5e\u96c6\u3002\u7531\u65bc</td></tr><tr><td colspan=\"2\">\u5be6\u7fa9\u8a5e\u7684\u8a9e\u7fa9\u95dc\u4fc2\u5728\u5c0d\u7576\u524d\uf906\u5b50\u7684\u8a9e\u6cd5\u7d50\u69cb\u78ba\u5b9a\u53ca\u5404\u8a5e\u8a9e\u8a5e\u7fa9\u7684\u9078\u64c7\u8d77\u8457\u76f8</td></tr><tr><td colspan=\"2\">\u7576\u91cd\u8981\u7684\u4f5c\u7528\uff0c\u6211\u5011\u5c0d\u4e00\u500b\uf906\u5b50\u7684\u8a55\u50f9\u5c31\u5efa\uf9f7\u5728\u5c0d\u8a72\uf906\u4e2d\u5be6\u7fa9\u8a5e\u7684\u8a55\u50f9\u57fa\u790e\u4e4b</td></tr><tr><td colspan=\"2\">\u4e0a\u3002\u628a\u8a5e\u8a9e\u7684\u7576\u524d\u4e0a\u4e0b\u6587\u76f8\u95dc\u8a5e\u96c6\u8207\u8a5e\u8a9e\u5404\u7fa9\u9805\u7684\u9650\u5236\u898f\u5247\u6240\u63cf\u8ff0\u8a9e\u7fa9\u7279\u5fb5\u8cc7</td></tr><tr><td colspan=\"2\">\u8a0a\u9032\ufa08\u6bd4\u8f03\uff0c\u6839\u64da\u6bd4\u8f03\u7684\u76f8\u4f3c\ufa01\u9078\u64c7\u6700\u5408\u9069\u7684\u7fa9\u9805\u3002\u540c\u6642\u5c07\u76f8\u4f3c\ufa01\u7684\u6700\u5927\u503c\u4f5c</td></tr><tr><td colspan=\"2\">\u7232\u8a72\u8a5e\u8a9e\u7684\u8a55\u50f9\u503c\u3002\u4e2d\u9593\u5206\u6790\u7d50\u679c\u4e2d\u5404\u5be6\u7fa9\u8a5e\u7684\u8a55\u50f9\u5206\u503c\u53ef\u4ee5\u6210\u7232\u8a55\u50f9\u6b64\u4e2d\u9593</td></tr><tr><td colspan=\"2\">\u7d50\u679c\u7684\u4f9d\u64da\uff0c\u4ee5\u6b64\u5728\u591a\u500b\u4e2d\u9593\u7d50\u69cb\u4e2d\u9078\u51fa\u6700\u4f73\u7684\u7d50\u679c\u3002\u9019\u6a23\uff0c\u6211\u5011\u5728\u89e3\u6c7a\u8a5e\u7fa9</td></tr><tr><td>\u6b67\u7fa9\u7684\u57fa\u790e\u4e0a\u540c\u6642\u4e5f\u89e3\u6c7a\uf9ba\u7d50\u69cb\u6b67\u7fa9\u3002</td><td/></tr><tr><td colspan=\"2\">\u672c\u6587\u6240\u63d0\u51fa\u7684\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b\u5df2\u5728\u6a5f\u5668\u7ffb\u8b6f\u7cfb\u7d71\u4e2d\u5177\u9ad4\u5730\u5be6\u73fe\u3002\u5be6\u9a57\uf9b5\uf906\u7684\u6e2c\u8a66</td></tr><tr><td colspan=\"2\">\u8868\u660e\u8a72\u6392\u6b67\u6a21\u578b\u5c0d\u89e3\u6c7a\uf906\u6cd5\u5206\u6790\u4e2d\u7684\u8fad\u5f59\u6b67\u7fa9\u3001\u7d50\u69cb\u6b67\u7fa9\u662f\u6709\u6548\u7684\uff0c\u4e26\u4e14\u512a\u65bc</td></tr><tr><td>\u50b3\u7d71\u7684 YES/NOT \u7684\u65b9\u6cd5\u3002</td><td/></tr></table>",
+                "text": ""
+            },
+            "TABREF6": {
+                "html": null,
+                "num": null,
+                "type_str": "table",
+                "content": "<table><tr><td/><td>\u4e00\u7a2e\u57fa\u65bc\u77e5\u7db2\u7684\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b\u7814\u7a76</td><td>\u694a\u66c9\u5cf0\u3001\uf9e1\u5802\u79cb 63 \u694a\u66c9\u5cf0\u3001\uf9e1\u5802\u79cb</td></tr><tr><td colspan=\"3\">\u9019\u4e9b\uf9b5\uf906\u4e2d\u90fd\u542b\u6709\u7531\u8a72\u7fa9\u539f\u5b9a\u7fa9\u7684\u8a5e\u8a9e\u3002\u6211\u5011\u6839\u64da\uf96b\u8003\uf9b5\uf906\u5c0d\u521d\u59cb\u898f\u5247\u9032\ufa08\u4fee\u6539\u5b8c\u5584\u3002 \u65bc\u662f\u6211\u5011\u5c31\u53ef\u4ee5\u76f4\u63a5\u628a\"&amp;\uff02\u6a19\u6ce8\u7684\u7fa9\u539f\u4f5c\u7232\u7fa9\u9805\u7684\u9650\u5b9a\u898f\u5247\u3002\u5982\u4e0a\uf9b5\u4e2d\uff0c\u7fa9\u9805\"\u672c 1\uff02 \u5bb9\u8a5e\u77ed\u8a9e\u4e2d ADJ \u662f NP \u7684\u9650\u5b9a\u8a5e\u3002\u5982\u679c\u9650\u5b9a\u8a5e\u8207\u88ab\u9650\u5b9a\u8a5e\u4e4b\u9593\u662f\u504f\u6b63\u7684\u4fee\u98fe\u95dc\u4fc2\uff0c\u5982 \u5728\u5206\uf9d0\u6a39\u5c0d\u61c9\u7d50\u9ede\u9593\u7684\u6700\u77ed\uf937\u5f91\u7684\u908a\uf969\u6709\u95dc\uff1a\u6700\u77ed\uf937\u5f91\u8d8a\u9577\uff0c\u5247\u8868\u793a\uf978\u500b\u7fa9\u539f\u7684\u8a9e\u7fa9\u8ddd</td></tr><tr><td colspan=\"3\">\u7531\u65bc\u7fa9\u539f\u7684\u8a9e\u7fa9\u898f\u5247\u898f\u6a21\u6709\u9650\uff0c\u4e26\u4e14\u4e8b\u5148\u6709\u4e00\u500b\u81ea\u52d5\u751f\u6210\u7684\u521d\u59cb\u898f\u5247\u96c6\uff0c\u56e0\u6b64\u624b\u5de5\u5236\u5b9a \u8207\"\u8f1b 1\uff02\u7684\u898f\u5247\u5206\u5225\u7232 ADJ+NP,NP+NP \u7b49\uff0c\u9650\u5b9a\u8a5e\u5145\u7576 ATTRIBUTE\u3001MANNER \u7b49\u9644\u5c6c\u6210\u5206\uff0c\u5247\u5728\u4f9d\u5b58\u95dc\u4fc2 \uf9ea\u8d8a\u9060\uff0c\u5b83\u5011\u7684\u76f8\u4f3c\u7a0b\ufa01\u8d8a\u5c0f\u3002\u4e26\u4e14\uff0c\u4f4d\u65bc\u5206\uf9d0\u6a39\u4e0b\u90e8\u7684\u4e00\u5c0d\u7236\u5b50\u7d50\u9ede\u5c0d\u61c9\u7684\u7fa9\u539f\u9593\u7684</td></tr><tr><td colspan=\"3\">\u8207\u4fee\u6539\u898f\u5247\u6240\u82b1\u8cbb\u7684\u5de5\u4f5c\uf97e\u4e26\uf967\u5927\u3002\u901a\u904e\u5728\u521d\u59cb\u898f\u5247\u7684\u57fa\u790e\u4e0a\u9032\ufa08\u4eba\u5de5\u8abf\u6574\u7684\u65b9\u6cd5\uff0c\u6211 \u5011\u53ef\u4ee5\u5f97\u5230\u4e00\u500b\u8f03\u7232\u5b8c\u5584\u7684\u8a9e\u7fa9\u898f\u5247\u96c6\u3002 3.3.3 \u7fa9\u9805\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u7684\u78ba\u5b9a \u5728\u7232\u52d5\u4f5c\uf9d0\u7fa9\u539f\u53ca\u5c6c\u6027\u503c\u3001\uf969\uf97e\u503c\uf9d0\u7fa9\u539f\u5b9a\u7fa9\u9650\u5236\u898f\u5247\u5f8c\uff0c\u4f9d\u64da\u300a\u77e5\u7db2\u300b\u4e2d\u5c0d\u7fa9\u9805\u7684\u4e00 \u4e9b\u898f\u5b9a\uff0c\u6211\u5011\u53ef\u4ee5\u81ea\u52d5\u5730\u751f\u6210\u52d5\u8a5e\u6216\u5f62\u5bb9\u8a5e\u7684\u4efb\u610f\u4e00\u500b\u7fa9\u9805\u7684\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u3002 \u5c0d\u65bc\u52d5\u4f5c\uf9d0\u52d5\u8a5e\uff0cDEF \u9805\u7684\u7b2c\u4e00\u4f4d\u7f6e\u53ea\u80fd\u662f\u4e8b\u4ef6\uf9d0\u898f\u5b9a\u7684\u4e3b\u8981\u7279\u5fb5\uff1b\u56e0\u6b64\u53ef\u4ee5\u76f4\u63a5 \u6a39\u4e2d\u9ad4\u73fe\u7232\u9650\u5b9a\u8a5e\u662f\u88ab\u9650\u5b9a\u8a5e\u7684\u5b50\u7d50\u9ede\uff1b\u5982\u679c\u5b83\u5011\u4e4b\u9593\u662f\u4e3b\u8b02\u7d50\u69cb\uff0c\u5982 NP+VP+NP(\u4e3b\u8b02 \u8a9e\u7fa9\u8ddd\uf9ea\u61c9\u7576\u5c0f\u65bc\u4f4d\u65bc\u5176\u4e0a\u7684\u4efb\u610f\u4e00\u5c0d\u7236\u5b50\u3002\u56e0\u6b64\uff0c\u5728\u8a08\u7b97\u8a9e\u7fa9\u8ddd\uf9ea\u6642\u61c9\u5c0d\u5206\uf9d0\u6a39\u4e0a\u7684 \u672c 1\uff1a(THEME publications|\u66f8\u520a) \u8cd3)\u3001NP+ADJP \u7b49\uff0c\u5247\u9650\u5b9a\u8a5e\u5728\u6a39\u4e2d\u5145\u7576\u88ab\u9650\u5b9a\u8a5e\u7684\u7236\u7d50\u9ede\u3002 \u908a\u52a0\u6b0a\u3002\u6b64\u5916\uff0c\u8a08\u7b97\u4e2d\u9084\u61c9\u9ad4\u73fe\u51fa\uf967\u540c\uf9d0\u578b\u7684\u7fa9\u539f(\u5982\u5be6\u9ad4\uf9d0\u8207\u52d5\u4f5c\uf9d0)\u4e4b\u9593\u8a9e\u7fa9\u7684\uf967\u53ef \u8f1b 1\uff1a(THEME LandVehicle|\uf902) \u6bd4\u6027\u3002\u6839\u64da\u4e0a\u8ff0\u5206\u6790\uff0c\u7fa9\u539f\u9593\u8a9e\u7fa9\u8ddd\uf9ea\u548c\u8a9e\u7fa9\u76f8\u4f3c\ufa01\u7684\u5b9a\u7fa9\u5982\u4e0b\uff1a \u6211\u5011\u4e4b\u6240\u4ee5\u8981\u5c0d\u8a5e\u8a9e\u5340\u5206\u9650\u5b9a\u8a5e\u8207\u975e\u9650\u5b9a\u8a5e\u662f\u7531\u65bc\u9019\uf978\uf9d0\u8a5e\u7684\u8a55\u50f9\u6f14\u7b97\u6cd5\uf967\u76f8 \u8207\u6b64\u540c\u6642\uff0c\u6211\u5011\u9084\u53ef\u4ee5\u7232\u67d0\u4e9b\u8a9e\u7fa9\u683c\u5b9a\u7fa9\u901a\u7528\u7684\u898f\u5247\uff0c\u4ee5\u8655\uf9e4\u4e2d\u9593\u7d50\u69cb\u4e2d\uf967\u80fd\u6839\u64da \u5728\u5176\u9650\u5b9a\u8a5e\u7fa9\u9805\u78ba\u5b9a\u4e4b\u5f8c\u9032\ufa08\uff0c\u9078\u53d6\u8207\u9019\u500b\u7fa9\u9805\u7684\u5be6\uf9b5\u7684\u6700\u9ad8\u6bd4\u8f03\u503c\u4f5c\u7232\u8a55\u50f9\u4f9d\u64da\u3002\u5c0d \u683c\u5b9a\u7fa9\u901a\u7528\u7684\u9650\u5b9a\u898f\u5247\u5982\u4e0b\uff1a \u50f9\u662f\u901a\u904e\u5404\u500b\u7fa9\u9805\u7684\u5be6\uf9b5\u96c6\u8207\u5b83\u7684\u88ab\u9650\u5b9a\u8a5e\u8a9e\u96c6\u7684\u6bd4\u8f03\u7372\u5f97\u7684\uff1b\u800c\u975e\u9650\u5b9a\u8a5e\u7684\u8a55\u50f9\u5247\u8981 \u4e2d\u4e00\u822c\u542b\u6709\u5730\u9ede\u7b49\u8cc7\u8a0a\uff0c\u800c TIME \u683c\u4e2d\u4e00\u822c\u542b\u6709\u6642\u9593\u7b49\u8cc7\u8a0a\uff0c\u56e0\u6b64\uff0c\u6211\u5011\u53ef\u4ee5\u7232\u9019\uf978\u500b \u64c7\u53ca\u8a55\u50f9\u5206\uf969\u5f71\u97ff\u8457\u7576\u524d\u5206\u6790\u7d50\u69cb\u7684\u7e3d\u9ad4\u8a55\u50f9\u503c\u3002\u5728\u5177\u9ad4\u7684\u6f14\u7b97\u6cd5\u5be6\u73fe\u4e0a\uff0c\u9650\u5b9a\u8a5e\u7684\u8a55 \u642d\u914d\u95dc\u4fc2\u9032\ufa08\u6392\u6b67\u7684\u8a5e\u8a9e\u3002\u9019\u4e9b\u8a9e\u7fa9\u683c\u901a\u5e38\u542b\u6709\u8f03\u660e\u986f\u7684\u8a9e\u7fa9\u7279\u5fb5\uff0c\u5982 LOCATION \u683c \u540c\u3002\u6839\u64da\u5b9a\u7fa9\uff0c\u9650\u5b9a\u8a5e\u5728\u8a5e\u8a9e\u4e4b\u9593\u7684\u642d\u914d\u95dc\u4fc2\u4e2d\u8d77\u8457\u4e3b\u8981\u7684\u4fee\u98fe\u652f\u914d\u4f5c\u7528\uff0c\u5b83\u7684\u7fa9\u9805\u9078 \u8a2d\u7fa9\u539f A\u3001B\uff0c\u5b83\u5011\u5206\u5225\u4f4d\u65bc\u5206\uf9d0\u6a39\u7684\u7b2c a</td></tr><tr><td colspan=\"3\">\u5c07\u7fa9\u9805\u7b2c\u4e00\u4f4d\u7f6e\u4e0a\u7684\u7fa9\u539f\u7684\u9650\u5236\u898f\u5247\u4f5c\u7232\u8a72\u7fa9\u9805\u7684\u9650\u5236\u898f\u5247\u3002 LOCATION\uff1a(*OR* PLACE \u5730\u65b9) \u4e00\u68f5\u95dc\u4fc2\u6a39\u9032\ufa08\u8a55\u50f9\u5c31\u662f\u5efa\uf9f7\u5728\u5c0d\uf906\u4e2d\u5145\u7576\u9650\u5b9a\u8a5e\u7684\u8a5e\u8a9e\u7d50\u9ede\u7684\u8a55\u50f9\u4e0a\u7684[\u8a79\u885b\u6771</td></tr><tr><td>1996]\u3002</td><td>\u6253 1\uff1abuy|\u8cb7,commercial|\u5546 TIME\uff1a (*OR* \u6642\u9593)</td></tr><tr><td colspan=\"3\">\u6253 2\uff1aexercise|\u935b\uf996,sport|\u9ad4\u80b2 \u5c0d\u65bc\u4e00\u500b\u53ef\u5145\u7576\u9650\u5b9a\u8a5e\u7684\u8a5e\u8a9e\uff0c\u6211\u5011\u5b9a\u7fa9\u5b83\u7684\u4e0a\u4e0b\u6587\u8a9e\u5883\uff0c\u4e5f\u7a31\u7232\u4e0a\u4e0b\u6587\u7a97\u53e3\uff0c\u662f \u5403 1\uff1aeat|\u5403 3.4 \u8a5e\u7fa9\u6392\u6b67\u6f14\u7b97\u6cd5\u7684\u8a73\u7d30\u63cf\u8ff0 \u6240\u6709\u88ab\u5176\u9650\u5b9a\u7684\u8a5e\u8a9e\u4e26\u4e0a\u6240\u5177\u6709\u7684\u683c\u8cc7\u8a0a\u3002\u7d66\u5b9a\uf906\u5b50\u7684\u4e00\u500b\uf906\u6cd5\u7d50\u69cb\uff0c\u901a\u904e\u8a72\u7d50\u69cb\u5c0d\u61c9 \u5403 2\uff1adestroy|\u6d88\u6ec5,military|\u8ecd \u7684\u4f9d\u5b58\u6a39\uff0c\u6211\u5011\u53ef\u4ee5\u5f88\u65b9\uf965\u5730\u5f97\u5230\u5404\u8a5e\u7684\u4e0a\u4e0b\u6587\u8a9e\u5883\u3002\u4e0b\u9762\u662f\u8a5e\u8a9e\u5728\u7576\u524d\u4f9d\u5b58\u6a39\u4e2d\u7684\u8a9e \u4e0a\u4e00\u7bc0\u4ecb\u7d39\uf9ba\u5982\u4f55\u7372\u5f97\u8a5e\u8a9e\u7fa9\u9805\u7684\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u3002\u672c\u7bc0\u5c07\u8a0e\uf941\u5982\u4f55\u904b\u7528\u7fa9\u9805\u7684\u8a9e\u7fa9\u898f\u5247 \u5728\u4e0a\u9762\u5e7e\u500b\u52d5\u8a5e\u7fa9\u9805\uff0c\u6211\u5011\u5206\u5225\u5c07\"buy|\u8cb7\uff02\u3001\"exercise|\u935b\uf996\uff02\u3001\"eat|\u5403\uff02\u3001 \u5883\u9078\u53d6\u539f\u5247\uff1a \u5728\u7d66\u5b9a\u7684\u4e2d\u9593\u7d50\u69cb\u4e2d\u9032\ufa08\u8a5e\u7fa9\u6392\u6b67\u3002 \"destroy|\u6d88\u6ec5\uff02\u7684\u9650\u5b9a\u898f\u5247\u4f5c\u7232\u6253 1\uff0c\u6253 2\uff0c\u5403 1\uff0c\u5403 2 \u7684\u9650\u5b9a\u898f\u5247\u3002 \u800c\u5c0d\u65bc\u5f62\u5bb9\u8a5e\uff0c\u5b83\u5011\u7684\u7fa9\u9805\u4e3b\u8981\u7531\u5c6c\u6027\u503c\uf9d0\u7fa9\u539f\u53ca\uf969\uf97e\uf9d0\u7fa9\u539f\u69cb\u6210\u3002\"\u5c6c\u6027\u503c\uff02\u662f (1) \u5982\u679c\u8a5e\u8a9e\u662f\u52d5\u8a5e\uff0c\u5247\u9078\u53d6\u5176\u5b50\u7d50\u9ede\u7684 AGENT\u3001THEME\u3001MANNER\u3001TENSE\u3001TOOLS\u3001 3.4.1 \u78ba\u5b9a\u8a5e\u8a9e\u7684\u8a9e\u7fa9\u74b0\u5883 LOCATION\u3001TIME \u7b49\u9644\u5c6c\u6210\u5206\uff1b</td></tr><tr><td colspan=\"3\">\u6240\u6709\u5c6c\u65bc\u5c6c\u6027\u503c\u6982\uf9a3\u7684\u552f\u4e00\u7684\u4e3b\u8981\u7279\u5fb5\uff0c\"\uf969\uf97e\u503c\uff02\u662f\u6240\u6709\u5c6c\u65bc\uf969\uf97e\u503c\u6982\uf9a3\u7684\u552f\u4e00\u7684\u4e3b \u672c\u6587\u7b2c\u4e8c\u7ae0\u7d66\u51fa\uf9ba\u6a5f\u8b6f\u7cfb\u7d71\u4e2d\u9593\u8a9e\u8a00\u7684\u8868\u793a\u65b9\u6cd5\uff0c\u7232\uf9ba\u65b9\uf965\u5730\u7372\u5f97\u67d0\u4e00\u5be6\u7fa9\u8a5e\u7684\u4e0a\u4e0b\u6587 (2) \u5982\u679c\u8a5e\u8a9e\u7232\u5176\u7236\u7d50\u9ede\u7684 PROPOSTION(\u5f9e\uf906)\u3001ATTRIBUTE \u7b49\u9644\u5c6c\u6210\u5206\uff0c\u5247\u9700\u4e26\u4e0a\u7236\u7d50</td></tr><tr><td colspan=\"3\">\u8981\u7279\u5fb5\uff0c\u5b83\u5011\u5206\u5225\u662f\u5f62\u5bb9\u8a5e\u7684\u5404\u7fa9\u9805\u7684\u9996\u4f4d\u6a19\uf9fc\uff1b\u5c6c\u6027\u503c\uf9d0\u7fa9\u539f\u548c\uf969\uf97e\u503c\uf9d0\u7fa9\u539f\u9664\u9996\u4f4d \u76f8\u95dc\u8a5e\uff0c\u6211\u5011\u5148\u5c07\u4e2d\u9593\u8a9e\u8a00\u8f49\u653e\u6210\u7232\u4f9d\u5b58\u95dc\u4fc2\u6a39\u7684\u5f62\u5f0f\u3002\u6a39\u7684\u6839\u7d50\u9ede\u662f\uf906\u5b50\u7684\u6838\u5fc3\u8a5e\uff0c \u9ede\u4e26\u5c07\u5176\u683c\u6ce8\u7232 Parent\uff1b</td></tr><tr><td colspan=\"3\">\u6a19\uf9fc\u5916\u5fc5\u9808\u9084\u5305\u542b\u6709\u4e00\u500b\u6b21\u8981\u7279\u5fb5\u3002\u5728\u7b2c\u4e8c\u4f4d\u5143\u4e0a\u4e00\u5b9a\u8981\u6a19\u6ce8\u8a72\u5c6c\u6027\u503c\u6216\uf969\uf97e\u503c\u6240\u6307\u5411 \u5176\u4ed6\u53d7\u6838\u5fc3\u8a5e\u652f\u914d\u7684\u9644\u5c6c\u6210\u5206\u5c31\u4f5c\u7232\u6839\u7d50\u9ede\u7684\u5b50\u6a39\uff0c\u9019\u4e9b\u5b50\u6a39\u5206\u5225\u4e5f\u662f\u4ee5\u5404\u9644\u5c6c\u6210\u5206\u7232 (3) \u5982\u679c\u52d5\u8a5e\u7684\u8a9e\u7fa9\u683c\u7232 EVENT\uff0c\u9019\u6642\u52d5\u8a5e\u505a\u517c\u8a9e\uf906\u88cf\u7684\u517c\u8a9e\uff0c\u6b64\u6642\u52d5\u8a5e\u6240\u5728\u517c\u8a9e\uf906\u4e2d</td></tr><tr><td colspan=\"3\">\u7684\u5c6c\u6027\u6216\uf969\uf97e\u7279\u5fb5\uff1b\u800c\u901a\u5e38\u7d55\u5927\u591a\uf969\u60c5\u6cc1\u4e0b\u5728\u7b2c\u4e09\u4f4d\u7f6e\u4e0a\u6a19\u6ce8\u8a72\u5c6c\u6027\u503c\u6216\uf969\uf97e\u503c\u7684\u5177\u9ad4 \u503c\uff0c\u800c\u9019\u4e9b\u5177\u9ad4\u503c\u6b63\u662f\u6211\u5011\u6240\u611f\u8208\u8da3\u7684\u3002\u6709\u6642\u5728 DEF \u7b2c\u4e09\u4f4d\u7f6e\u5f8c\u9084\u6709\u4e00\u4e9b\u8f14\u52a9\u7279\u5fb5\uff0c\u5b83 \u5011\u53ea\u662f\u9032\u4e00\u6b65\u5c0d\u95dc\u9375\u7fa9\u539f\u9032\ufa08\u88dc\u5145\uf96f\u660e\uff0c\u5c0d\u7fa9\u9805\u7684\u8a9e\u7fa9\u5f71\u97ff\u5f88\u5c0f\uff0c\u56e0\u6b64\u6211\u5011\u5728\u5b9a\u7fa9\u5f62\u5bb9 \u8a5e\u7684\u9650\u5236\u898f\u5247\u4e2d\uff0c\u53ea\u9700\u8003\u616e\u7b2c\u4e09\u4f4d\u7684\u7fa9\u539f\u3002\uf9b5\u5982\u4e0b\u9762\u662f\u5e7e\u500b\u5f62\u5bb9\u8a5e\u7684\u7fa9\u9805\u5b9a\u7fa9\uff1a \u5de8\u5927 1\uff1aDEF=aValue|\u5c6c\u6027\u503c,size|\u5c3a\u5bf8,big|\u5927 \u5de8\u5927 2\uff1aDEF=QValue|\uf969\uf97e\u503c,amount|\u591a\u5c11,many|\u591a \u9999 1\uff1aDEF=aValue|\u5c6c\u6027\u503c,circumstances|\u5883\u6cc1,flourishing|\u8208,desired|\uf97c \u9999 2 \uff1aDEF=aValue|\u5c6c\u6027\u503c,odor|\u6c23\u5473,fragrant|\u9999,desired|\uf97c \u7684\u8cd3\u8a9e\u61c9\u5145\u7576\u5176 AGENT \u683c\uff1b \u6839\u7d50\u9ede\u800c\u5efa\uf9f7\u8d77\uf92d\u7684\u4f9d\u5b58\u95dc\u4fc2\u6a39\u3002\uf9b5\u5982\u8f38\u5165\uf906\"\u7dad\u4fee/\u5716\u66f8\u9928/\u7684/\u7a7a\u8abf\uff02\u7684\uf978\u500b\u53ef\u80fd\u7684\uf906 \u6cd5\u7d50\u69cb\u6a39\u5982\u4e0b\uff1a \u5716 3.1 ((\u7dad\u4fee \u5716\u66f8\u9928)\u7684 \u7a7a\u8abf) Theme \u7dad\u4fee(V) \u5716\u66f8\u9928(N) \u7a7a\u8abf(N) ATTRIBUT Theme \u7dad\u4fee(V) \u7a7a\u8abf(N) ATTRIBUT (4) \u5426\u5247\u5982\u679c\u5f62\u5bb9\u8a5e\u7232 PREDICATE \u683c\uff0c\u5373\u5f62\u5bb9\u8a5e\u4f5c\u8868\u8a9e\uff0c\u5247\u9078\u53d6\u5144\u5f1f\u7d50\u9ede\u4e2d\u7684 \u5716\u66f8\u9928(N) Experiencer \u683c\uff1b \u5716 3.2 (\u7dad\u4fee(\u5716\u66f8\u9928\u7684 \u7a7a\u8abf)) \u7684\u6a39\u7d50\u69cb \u7684\u6a39\u7d50\u69cb (7) \u5982\u679c\u8a5e\u8a9e\u662f\uf97e\u8a5e\u4e14\u7232 Quatity \u683c\uff0c\u5247\u9078\u53d6\u5176\u7236\u7d50\u9ede\u4e26\u5c07\u5176\u8a9e\u6cd5\u683c\u6ce8\u7232 THEME\u3002</td></tr><tr><td colspan=\"3\">\u5728\u4e0a\u9762\u7684\uf9b5\u5b50\u4e2d\uff0c\u6211\u5011\u5c07\u9078\u64c7\"big|\u5927\uff02\u3001\"many|\u591a\uff02\u3001\"flourishing|\u8208\uff02\u3001\"fragrant| 3.4.2 \u8a5e\u8a9e\u7684\u8a55\u50f9\u503c\u8a08\u7b97 \u5728\u78ba\u5b9a\u8a5e\u8a9e\u7684\u4e0a\u4e0b\u6587\u8a9e\u5883\u524d\u6211\u5011\u5b9a\u7fa9\u8a5e\u8a9e\u7684\u9650\u5b9a\u95dc\u4fc2\uff1a\u8a2d\u6709\u8a5e\u8a9e A\uff0cB\uff0c\u5982\u679c\u5728\uf906 \u9999\uff02\u7684\u7fa9\u539f\u9650\u5236\u898f\u5247\u4f5c\u7232\u5c0d\u61c9\u5f62\u5bb9\u8a5e\u7fa9\u9805\u7684\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u3002 \u5b50\u4e2d A \u4fee\u98fe\u3001\u652f\u914d B\uff0c\u5247\u7a31\u5728\u7576\u524d\uf906\u4e2d A \u662f B \u7684\u9650\u5b9a\u8a5e\uff0cB \u662f A \u7684\u88ab\u9650\u5b9a\u8a5e\u3002\u9019\u88cf\u7684 \u5728 3.3.3 \u4e2d\u4ecb\u7d39\uf9ba\u52d5\u8a5e\u3001\u5f62\u5bb9\u8a5e\u53ca\uf97e\u8a5e\u7684\u7fa9\u9805\u7684\u8a9e\u7fa9\u9650\u5b9a\u898f\u5247\u7372\u53d6\u65b9\u6cd5\u3002\u4e0a\u4e00\u7bc0\u63cf\u8ff0\u5982\u4f55</td></tr><tr><td colspan=\"3\">EXPECT|\u671f\u671b\uff02\u7b49\u7fa9\u539f\u898f\u5247\u5b9a\u7fa9 \u5c0d\u65bc\uf97e\u8a5e\u6211\u5011\u4e5f\u53ef\u4ee5\u81ea\u52d5\u751f\u6210\u9650\u5236\u898f\u5247\uff0c\u898f\u5247\u4e2d\u898f\u5b9a\uf97e\u8a5e\u4fee\u98fe\u7684\u8a5e\u7684\u8a9e\u7fa9\u7279\u5fb5\u3002\u5728 \u9650\u5b9a\u95dc\u4fc2\u8207\u50b3\u7d71\u7684\u4f9d\u5b58\u95dc\u4fc2\u6709\u6240\uf967\u540c\uff1a\u4f9d\u5b58\u8a9e\u6cd5\u8a8d\u7232\u767c\u751f\u4f9d\u5b58\u95dc\u4fc2\u7684\u4e00\u5c0d\u8a5e\u4e2d\uff0c\u5982\u679c\u8a5e \u5f97\u5230\u4e00\u68f5\u4e2d\u9593\u8a9e\u8a00\u7684\u7d50\u69cb\u4f9d\u5b58\u6a39\u4e2d\u5404\u652f\u914d\u8a5e\u7684\u4e0a\u4e0b\u6587\u8a9e\u5883\u3002\u73fe\u5728\u53ef\u4ee5\u6839\u64da\u8a5e\u8a9e\u6240\u5728\u8a9e\u7fa9</td></tr><tr><td colspan=\"3\">CLAUSE \u683c\u63cf\u8ff0\u3002 \u300a\u77e5\u7db2\u300b\u4e2d\uff0c\u540d\uf97e\u8a5e\u7684\u5b9a\u7fa9\u88cf\u7528\"&amp;\uff02\u6a19\u6ce8\u5176\u6307\u5411\u7684\u5c6c\u6027\u6216\u4e8b\u7269\u7684\uf9d0\u578b\uff1b\uf9b5\u5982\uff1a A \u4fee\u98fe\u8a5e B\uff0c\u5247 B \u7232\u4e3b\u8a5e\uff0cA \u7232\u5f9e\u5c6c\u8a5e\uff0cA \u662f B \u7684\u9644\u5c6c\u6210\u5206\uff0c\u5728\u4f9d\u5b58\u95dc\u4fc2\u6a39\u4e2d\u9ad4\u73fe\u7232 A \u74b0\u5883\u53ca\u5176\u5404\u7fa9\u9805\u7684\u9650\u5b9a\u898f\u5247\u4e2d\u6240\u63cf\u8ff0\u7684\u8a9e\u7fa9\u74b0\u5883\u9032\ufa08\u76f8\u4f3c\ufa01\u7684\u8a08\u7b97\u3002</td></tr><tr><td colspan=\"3\">\u672c\uff1a DEF=NounUnit|\u540d\uf97e,&amp;publications|\u66f8\u520a \u662f B \u7684\u5b50\u7d50\u9ede\uff1b\u800c\u5728\u6211\u5011\u5b9a\u7fa9\u7684\u9650\u5b9a\u95dc\u4fc2\u4e2d\u88ab\u9650\u5b9a\u8a5e\u5247\u662f\u9650\u5b9a\u8a5e\u7684\u4fee\u98fe\u652f\u914d\u7269\u4ef6\u3002\u5982\u5c0d 1)\u8a5e\u8a9e\u7fa9\u9805\u8207\u898f\u5247\u7684\u683c\u9650\u5236\u63cf\u8ff0\u7684\u76f8\u4f3c\ufa01 \u5728\u4fee\u6539\u7fa9\u539f\u7684\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u6642\uff0c\u672c\u6587\u5f9e\u8a9e\uf9be\u5eab\u4e2d\u7232\u6bcf\u500b\u7fa9\u539f\u9078\u53d6\u4e00\u5b9a\uf969\uf97e\u7684\uf9b5\uf906\uff0c \u8f1b\uff1a DEF=NounUnit|\u540d\uf97e,&amp;LandVehicle|\uf902 \u65bc\u4e3b\u8b02\u8cd3\u7d50\u69cb\u7684\uf906\u5b50\uf92d\uf96f\uff0c\u4e3b\u52d5\u8a5e\u5c31\u662f\uf906\u5b50\u7684\u8a5e\u8a9e\u4e3b\u8a9e\u3001\u8cd3\u8a9e\u7684\u9650\u5b9a\u8a5e\uff1bADJ\uff0bNP \u7684\u5f62 \u300a\u77e5\u7db2\u300b\u4e2d\u63d0\u4f9b\uf9ba\u52d5\u4f5c\uf9d0\uff0c\u5be6\u9ad4\uf9d0\u3001\u5c6c\u6027\u503c\uf9d0\u7b49\u7fa9\u539f\u5206\uf9d0\u6a39\uff0c\u7fa9\u539f\u9593\u7684\u8a9e\u7fa9\u8ddd\uf9ea\u8207\u7fa9\u539f</td></tr></table>",
+                "text": "\u5982\u679c\u52d5\u8a5e\u7684\u8a9e\u7fa9\u683c\u7232 THEME\uff0c\u9019\u6642\u52d5\u8a5e\u505a\u8cd3\u8a9e\u6216\u5c0f\uf906\u8cd3\u7d50\u69cb\u7684\u8cd3\u8a9e\u5f9e\uf906\u3002\u5982\u679c\u8a72\u52d5 \u8a5e\u7684 AGENT \u8a9e\u7fa9\u683c\uf967\u5b58\u5728\uff0c\u5247\u52a0\u5165\u7236\u7d50\u9ede\u8a9e\u5883\u4e2d\u7684 AGENT \u683c\u3002\u5982\u679c\u8a72\u52d5\u8a5e\u662f\u88ab\u52d5\u8a9e \u614b\uff0c\u8981\u5c07 AGENT \u683c\u6539\u7232 THEME \u683c\uff1b (5) \u5982\u679c\u52d5\u8a5e\u7684\u8a9e\u7fa9\u683c\u7232 SUBEVENT\uff0c\u9019\u6642\u52d5\u8a5e\u505a\uf99a\u52d5\uf906\uff0c\u5247\u52a0\u5165\u7236\u7d50\u9ede\u7684\u4e3b\u8a9e\u8a5e\u53ca\u5176\u6240 \u5145\u7576\u7684\u8a9e\u7fa9\u683c\u3002\u5982\u679c\u8a72\u52d5\u8a5e\u662f\u88ab\u52d5\u8a9e\u614b\uff0c\u8981\u5c07 AGENT \u683c\u6539\u7232 THEME \u683c\uff1b (6) \u5982\u679c\u8a5e\u8a9e\u662f\u5f62\u5bb9\u8a5e\u4e14\u7232 ATTRIBUTE \u683c\uff0c\u5247\u9078\u53d6\u5176\u7236\u7d50\u9ede\u4e26\u5c07\u5176\u8a9e\u6cd5\u683c\u6ce8\u7232 THEME\uff1b \u548c\u7b2c b \u5c64(\u898f\u5b9a\u6839\u7d50\u9ede\u7232\u7b2c 0 \u5c64)\uff0c\u5b83\u5011\u7684 \u6700\u8fd1\u5171\u540c\u7956\u5148(\u53ef\u4ee5\u662f A \u6216 B \u672c\u8eab)\u4f4d\u65bc\u7b2c c \u5c64\uff0c\u5247 A \u8207 B \u7684\u8a9e\u7fa9\u8ddd\uf9ea\u7232\uff1a"
+            },
+            "TABREF7": {
+                "html": null,
+                "num": null,
+                "type_str": "table",
+                "content": "<table><tr><td>\u4e00\u7a2e\u57fa\u65bc\u77e5\u7db2\u7684\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b\u7814\u7a76</td><td>69</td></tr><tr><td>\"\u7238\u7238\u6b63\u5728\u4fee\u90a3\u53f0\u820a\u96fb\u8996\u5462\u3002\uff02</td><td/></tr><tr><td>\u5206\u6790\u5f8c\u5f97\u5230\u7684\u4e2d\u9593\u7d50\u679c\u7232\uff1a</td><td/></tr><tr><td>((CAT</td><td/></tr><tr><td>3.4.3 \u8a5e\u8a9e\u7684\u7fa9\u9805\u9078\u64c7</td><td/></tr><tr><td colspan=\"2\">\u5728\u6f14\u7b97\u6cd5 3.3 \u7d66\u51fa\uf9ba\u8a5e\u8a9e\u7684\u8a55\u50f9\u503c\u8a08\u7b97\u65b9\u6cd5\u3002\u8a72\u6f14\u7b97\u6cd5\u53ef\u4ee5\u5f97\u5230\u4e2d\u9593\u7d50\u69cb\u4e2d\u5404\u9650\u5b9a\u8a5e\u8a5e</td></tr><tr><td colspan=\"2\">\u7684\u8a55\u50f9\u503c\u53ca\u6700\u4f73\u7fa9\u9805\uff0c\u540c\u6642\u9084\u53ef\u4ee5\u78ba\u5b9a\u53d7\u8a72\u8a5e\u8a9e\u9650\u5b9a\u7684\u5404\u8a5e\u8a9e\u5728\u8a72\u8a5e\u8a9e\u7684\u8a9e\u5883\u4e2d\u7684\u8a55\u50f9</td></tr><tr><td>\u503c\u53ca\u6700\u4f73\u7fa9\u9805\u3002\u672c\u7bc0\u4e2d\u4ecb\u7d39\u5982\u4f55\u5728\u4e2d\u9593\u8a9e\u8a00\u7d50\u69cb\u4e2d\u9032\ufa08\u8a5e\u7fa9\u7684\u9078\u64c7\uff0c\u5373\u8a5e\u7fa9\u6392\u6b67\u3002</td><td/></tr><tr><td colspan=\"2\">\u672c\u6587\u6309\u7167\u8a5e\u6027\u7684\u8a9e\u6cd5\u7279\u6027\uf92d\u4ee5\u4e00\u5b9a\u7684\u5148\u5f8c\u9806\u5e8f\u5c0d\uf967\u540c\u8a5e\u8a9e\u9032\ufa08\u7fa9\u9805\u9078\u64c7\u3002\u4e00\u822c\uf92d</td></tr><tr><td colspan=\"2\">\uf96f\uff0c\u52d5\u8a5e\u7684\u642d\u914d\u95dc\u4fc2\u5c0d\u7d50\u69cb\u7684\u8a9e\u7fa9\u8a9e\u7fa9\u5f71\u97ff\u6700\u5927\uff0c\u5176\u6b21\u662f\u5f62\u5bb9\u8a5e\u3002\u6839\u64da\u9019\u4e00\u539f\u5247\uff0c\u6211\u5011</td></tr><tr><td>\u63d0\u51fa\uf9ba\u4ee5\u4e0b\u8a5e\u7fa9\u9078\u64c7\u7684\u6b65\u9a5f\u6f14\u7b97\u6cd5\uff1a</td><td/></tr><tr><td colspan=\"2\">1\uff1a\u5275\u5efa\u4e00\u500b RESULT \u54c8\u5e0c\u8868\uff0c\u5b83\u7684\u5165\u53e3\u662f\u7d50\u69cb\u4f9d\u5b58\u6a39\u4e2d\u5404\u7d50\u9ede\u5c0d\u61c9\u7684\u8a5e\u8a9e\uff0c\u6bcf\u500b</td></tr><tr><td colspan=\"2\">\u5165\u53e3\u9805\u7684\u503c\u5c31\u662f\u5165\u53e3\u8a5e\u8a9e\u7684\u4faf\u9078\u7fa9\u9805\u96c6\u5408\u53ca\u7d50\u9ede\u7684\u8a55\u50f9\u5206\u503c\u3002\u5982\u679c\u8a5e\u8a9e\u5728\u5206\u6790\u968e\u6bb5\u5df2\u7d93</td></tr><tr><td colspan=\"2\">\u8a08\u7b97\u51fa\u7fa9\u9805\u96c6\u53ca\u8a55\u50f9\u5206\uf969\uff0c\u5c31\u5c07\u5b83\u5011\u8a2d\u7232\u76f8\u61c9\u7684\u9805\u503c\uff0c\u5426\u5247\u521d\u59cb\u7684\u5019\u9078\u7fa9\u9805\u7232\u8a5e\u8a9e\u7684\u6240</td></tr><tr><td>\u6709\u7fa9\u9805\uff0c\u800c\u8a5e\u8a9e\u8a55\u50f9\u5206\u503c\u7232 MINIUM\u3002</td><td/></tr><tr><td colspan=\"2\">2\uff1a\u6309\u81f3\u5e95\u5411\u4e0a\u7684\u9806\u5e8f\u5c0d\u7d50\u69cb\u4f9d\u5b58\u6a39\u4e2d\u7684\u52d5\u8a5e\u7d50\u9ede\u9032\ufa08\u8a5e\u7fa9\u9078\u64c7\uff0c\u5c07\u9078\u64c7\u7d50\u679c\u96c6\u5408(\u53ef</td></tr><tr><td colspan=\"2\">\u80fd\u542b\u6709\u4e00\u500b\u6216\u591a\u500b\u7684\u5143\u7d20)\u53ca\u8a55\u50f9\u5206\u503c\u586b\u5165 RESULT \u54c8\u5e0c\u8868\u7684\u76f8\u61c9\u9805\u4e2d\u3002\u5c0d\u53d7\u8a72\u52d5\u8a5e\u9650</td></tr><tr><td colspan=\"2\">\u5b9a\u7684\u6240\u6709\u8a5e\u8a9e\uff0c\u6211\u5011\u53ef\u4ee5\u5f97\u5230\u5728\u8a72\u52d5\u8a5e\u4e0b\u7684\u8a55\u50f9\u5206\u503c\u53ca\u6700\u512a\u7fa9\u9805\uff0c\u5c07\u9019\u500b\u8a55\u50f9\u5206\u503c\u8207</td></tr><tr><td colspan=\"2\">RESULT \u8868\u4e2d\u5c0d\u61c9\u9805\u7684\u8a55\u50f9\u503c\u9032\ufa08\u6bd4\u8f03\uff0c\u5982\u679c\uf901\u5927\uff0c\u5247\u5c07\u6b64\u6700\u512a\u7fa9\u9805\u96c6\u5408\u8207\u5206\u503c\u66ff\u4ee3\u539f</td></tr><tr><td>\uf92d\u7684\u5167\u5bb9\u3002</td><td/></tr><tr><td colspan=\"2\">1\u2264i\u2264z } 3\uff1a\u8207\u7b2c 2 \u6b65\uf9d0\u4f3c\uff0c\u5c0d\u6a39\u4e2d\u6240\u6709\u5f62\u5bb9\u8a5e\u53ca\u5176\u9650\u5b9a\u8a5e\u9032\ufa08\u8a5e\u7fa9\u9078\u64c7\u4e26\uf901\u6539\u76f8\u61c9\u7684 (3.5) \u5982\u679c OP \u7232*AND*\uff0c\u6839\u64da 3.3.1 \u4e2d\u7684\u5b9a\u7fa9\uff0c\u53ea\u8981 RS \u96c6\u5408\u4e2d\u6709\u4e00\u500b\u5143\u7d20\u7684\u6bd4\u8f03\u503c\u8f03\u5c0f\uff0c \u8fd4\u56de\u7684\u503c\u5c31\u61c9\u8a72\u5c0f\uff0c\u6b64\u6642\u61c9\u5f9e RS \u4e2d\u9078\u53d6\u4e00\u500b\u6700\u5c0f\u503c\uff1b\u5982\u679c OP \u7232*OR*\uff0c\u5247\u8868\u793a\u96c6\u5408\u53ef RESULT \u8868\u9805\u3002</td></tr><tr><td colspan=\"2\">\u53d6\u4efb\u4f55\u4e00\u6bd4\u8f03\u503c\uff0c\u9019\u6a23\u6b64\u6642\u61c9\u5f9e RS \u4e2d\u9078\u53d6\u4e00\u500b\u6700\u5927\u503c\uff1b\u5982\u679c OP \u7232*NOT*\uff0c\u8868\u793a\u96c6\u5408\u4e2d 4\uff1a\u8207\u7b2c 2 \u6b65\uf9d0\u4f3c\uff0c\u5c0d\u6a39\u4e2d\u6240\u6709\uf97e\u8a5e\u53ca\u5176\u9650\u5b9a\u8a5e\u9032\ufa08\u8a5e\u7fa9\u9078\u64c7\u4e26\uf901\u6539\u76f8\u61c9\u7684 RESULT</td></tr><tr><td colspan=\"2\">\u53ea\u6709\u4e00\u500b\u5143\u7d20\uff0c\u4e26\u4e14\u5982\u679c\u8a72\u5143\u7d20\u503c\u6bd4\u8f03\u5927\uff0c\u8fd4\u56de\u7684\u503c\u53cd\u800c\u61c9\u8a72\u5c0f\uff0c\u53cd\u4e4b\u5143\u7d20\u7684\u503c\u6bd4\u8f03\u5c0f \u8868\u9805\u3002</td></tr><tr><td colspan=\"2\">\u7684\u8a71\uff0c\u8fd4\u56de\u7684\u503c\u61c9\u8a72\u5927\u3002\u56e0\u6b64\uff0c\u6211\u5011\u53ef\u4ee5\u7528 1 \u8207\u5143\u7d20\u503c\u7684\u5dee\uf92d\u8a08\u7b97\u6bd4\u8f03\u76f8\u4f3c\ufa01\u3002 5\uff1a\u5c0d\u4f9d\u5b58\u6a39\u4e2d\u9084\u672a\u9032\ufa08\u904e\u8a5e\u7fa9\u6392\u6b67\u7684\u7d50\u9ede\uff0c\u5982\u679c\u5176\u6240\u5728\u8a9e\u7fa9\u683c\u6709\u901a\u7528\u7684\u9650\u5236\u898f\u5247\uff0c</td></tr><tr><td>\u7d9c\u4e0a\u6240\u8ff0\uff0c\u53ef\u4ee5\u5b9a\u7fa9 C \u8207 Entry \u7684\u76f8\u4f3c\ufa01\u7232 \u5247\u6839\u64da\u901a\u7528\u898f\u5247\u9032\ufa08\u8a5e\u7fa9\u9078\u64c7\uff0c\uf901\u6539\u76f8\u61c9\u7684 RESULT \u8868\u9805\u3002</td><td/></tr><tr><td colspan=\"2\">\u23a7 MAX 6\uff1a\u5c0d\u65bc\u4f9d\u5b58\u6a39\u7684\u6bcf\u500b\u8a5e\u8a9e\u7d50\u9ede\uff0c\u5230 RESULT \u8868\u67e5\u51fa\u5c0d\u61c9\u9805\u7684\u7fa9\u9805\u96c6\u5408\uff0c\u5c07\u9019\u4e9b\u7fa9 = \u65f6 OP \u5f53 * *OR RS ; \u23aa \u23a8 = \u2212 \u2212 = \u65f6 \u5f53OP * *AND RS ; CaseRule) SC(Entry, ENTRY SIM (3.6) \u9805\u7684\u6240\u6709\u82f1\u6587\u5c0d\u8b6f\u8a5e\u4f5c\u7232\u5206\u6790\u7d50\u69cb\u7684\u8a5e\u8a9e\u8b6f\u6587\u3002 MIN \u23aa \u23a9 = \u65f6 \u5f53OP * *NOT ; -1 RS MAX 3.4.4 \u8a5e\u7fa9\u6392\u6b67\u6f14\u7b97\u6cd5\u793a\uf9b5</td></tr><tr><td>3)\u8a5e\u8a9e\u7684\u8a55\u50f9\u6f14\u7b97\u6cd5 \u6f14\u7b97\u6cd5 3.3\uff1a\u8a5e\u8a9e\u7684\u8a55\u50f9\u6f14\u7b97\u6cd5 \u4e0b\u9762\u6211\u5011\uf92d\u770b\u770b\u5982\u4f55\u904b\u7528\u672c\u7ae0\u7d66\u51fa\u7684\u6f14\u7b97\u6cd5\u9032\ufa08\u8a5e\u7fa9\u7684\u6392\u6b67\uff0c\u8a2d\u6709\u8f38\u5165\uf906</td><td/></tr></table>",
+                "text": "\u694a\u66c9\u5cf0\u3001\uf9e1\u5802\u79cb \u5404\u898f\u5247\u4e2d\u683c\u7684\u9650\u5236\u63cf\u8ff0\u8207 word \u7684\u8a9e\u7fa9\u74b0\u5883 ENV \u4e2d\u5404\u8a5e\u8a9e\u7684\u7fa9\u9805\u9032\ufa08\u76f8\u4f3c\ufa01\u7684\u8a08\u7b97\uff0c\u9019 \u4e9b\u8a5e\u8a9e\u76f8\u5c0d\u65bc word \u7576\u524d\u7fa9\u9805\u7684\u6700\u5927\u76f8\u4f3c\ufa01\u503c\u53ca\u5c0d\u61c9\u7fa9\u9805\u8a18\uf93f\u65bc\u4e00\u500b\u4e8c\u7dad\u9663\uf99c\u4e2d\u3002\u63a5\u8457\u6f14 \u7b97\u6cd5\u6839\u64da ENV \u4e2d\u5404\u8a5e\u8a9e\u7684\u6700\u5927\u76f8\u4f3c\ufa01\u8a08\u7b97 word \u7576\u524d\u7fa9\u9805\u7684\u8a55\u50f9\u503c\u3002\u6700\u5f8c\uff0c\u6f14\u7b97\u6cd5\u628a word \u7684\u5177\u6709\u6700\u5927\u8a55\u50f9\u503c\u7684\u7fa9\u9805\u4f5c\u7232 word \u7684\u6700\u4f73\u7fa9\u9805 BestEntry\uff0c\u9019\u4e00\u6700\u5927\u8a55\u50f9\u503c\u4f5c\u7232 word \u7684 \u8a55\u50f9\u503c\u3002\u540c\u6642\uff0c\u6f14\u7b97\u6cd5\u5f9e\u4e8c\u7dad\u9663\uf99c\u4e2d\u53d6\u51fa ENV \u4e2d\u5404\u8a5e\u8a9e\u76f8\u5c0d\u65bc BestEntry \u7684\u6700\u5927\u76f8\u4f3c\ufa01 \u53ca\u7fa9\u9805\uff0c\u4e26\u5c07\u5176\u4f5c\u7232\u8a72\u8a5e\u8a9e\u5728 word \u7684\u8a9e\u5883\u4e2d\u7684\u8a55\u50f9\u503c\u53ca\u6700\u4f73\u7fa9\u9805\u3002"
+            },
+            "TABREF9": {
+                "html": null,
+                "num": null,
+                "type_str": "table",
+                "content": "<table><tr><td>\u4e00\u7a2e\u57fa\u65bc\u77e5\u7db2\u7684\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b\u7814\u7a76 \u4e00\u7a2e\u57fa\u65bc\u77e5\u7db2\u7684\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b\u7814\u7a76</td><td>75 \u694a\u66c9\u5cf0\u3001\uf9e1\u5802\u79cb 77</td></tr><tr><td colspan=\"2\">\u8868 3.6 \u5404\u5206\u6790\u7d50\u69cb\u7684\u8a55\u50f9\u5206\uf969 \u5b9a\u7fa9\u662f\"return|\u9084\uff02\uff0c\u9019\u4e00\u7fa9\u9805\u4e2d\u7684\u642d\u914d\u53d7\u4e8b\u9ad4\u4e00\u822c\uf92d\uf96f\u662f\u548c\u9322\u8ca1\u6709\u95dc\u7684\uff0c\u4f46\"return| \u8a9e\u7fa9\u6709\u95dc\u7684\u8655\uf9e4\uff0c\u5982\u8a9e\u6cd5\u683c\u7684\u8abf\u6574\u7b49\u3002</td></tr><tr><td colspan=\"2\">\u5206\u6790\u7d50\u69cb ( (\u4fee\uf9e4 \u7238\u7238 )\u7684 \u81ea\ufa08\uf902) 0 \u9084\uff02\u7684\u7fa9\u539f\u898f\u5247\u4e2d\u7684\u53d7\u4e8b\u9ad4\u7684\u8a9e\u7fa9\u7279\u5fb5\uf967\u4e00\u5b9a\u662f\u548c\u9322\u6709\u95dc\uff0c\u9019\u5c31\u6709\u53ef\u80fd\u5c0e\u81f4\u5982\"\u627e\u66f8\uff02 \u5404\u8a5e\u8a9e\u8a55\u50f9\u5206\uf969 \u5206\u6790\u7d50\u69cb\u8a55\u50f9\u5206\uf969 \u4fee\uf9e4 \u7238\u7238 \u81ea\ufa08\uf902 0 25.0 \u672c\u6587\u6240\u63d0\u51fa\u7684\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b\u5df2\u5728\u6a5f\u5668\u7ffb\u8b6f\u7cfb\u7d71\u4e2d\u5177\u9ad4\u5730\u5be6\u73fe\u3002\u5be6\u9a57\uf9b5\uf906\u7684\u6e2c\u8a66\u8868\u660e \u4e2d\u7684\"\u627e\uff02\u4e5f\u6703\u8aa4\u9078\u7232\"Return\"\u3002 \u9019\u4e00\u6392\u6b67\u5c0d\u8655\uf9e4\u8fad\u5f59\u6b67\u7fa9\u3001\u7d50\u69cb\u6b67\u7fa9\u662f\u6709\u6548\u7684\u3002 0.0 (\u4fee\uf9e4 (\u7238\u7238 \u7684 \u81ea\ufa08\uf902) ) 89.28 0 89.28 89.28 ( (\u4fee\uf9e4 \u81ea\ufa08\uf902) \u7684 \u7238\u7238) 94.64 100 89.28 94.64 (\u4fee\uf9e4 (\u81ea\ufa08\uf902 \u7684 \u7238\u7238 )) 0 0 0 (\uff12)\uff1a\u300a\u77e5\u7db2\u300b\u4e2d\u8a5e\u8a9e\u542b\u6709\u7684\u4e00\u4e9b\u6587\u8a00\u7528\u6cd5\u5c0d\u5e72\u64fe\uf9ba\u8a5e\u7fa9\u7684\u6b63\u78ba\u9078\u64c7\u3002\u5982\"\u53bb\uff02\u4e00 \u7531\u65bc\u7814\u7a76\u7684\u6642\u9593\u95dc\u4fc2\uff0c\u672c\u6587\u7684\u6392\u6b67\u6a21\u578b\u5f9e\u529f\u80fd\u4e0a\uf92d\u770b\u9084\u53ea\u662f\u4e00\u500b\u5be6\u9a57\u7cfb\u7d71\uff0c\u9084\u6709\uf967 \u8a5e\uff0c\u5728\u300a\u77e5\u7db2\u300b\u4e2d\u6709\"leave|\uf9ea\u958b\uff02\u7684\u7fa9\u9805\u89e3\u91cb\uff0c\u9019\u7a2e\u7528\u6cd5\u5728\u73fe\u4ee3\u6587\u4e2d\u5f88\u5c11\u51fa\u73fe\uff0c\u800c\u5b83\u7684 \u5c11\u53ef\u4ee5\u6539\u9032\u7684\u5730\u65b9\u3002\uf9b5\u5982\uff1a \u5b58\u5728\u5c0d\u6703\u5f71\u97ff\"\u53bb\uff02\u7684\u5e38\ufa0a\u7fa9\u9805\"go|\u53bb\uff02\u7684\u9078\u64c7\u3002 0\uff0e0 1\uff1a\u672c\u6587\u4e3b\u8981\u662f\u5728\u8a5e\u6b67\u6d88\u6b67\u7684\u57fa\u790e\u4e0a\u9032\ufa08\u7d50\u69cb\u6d88\u6b67\uff0c\u56e0\u6b64\u53ea\u8003\u616e\uf9ba\u5e38\u7528\u7684\u5e7e\u7a2e\u6b67\u7fa9</td></tr><tr><td colspan=\"2\">\u5c0f\u65bc\u95be\u503c\uff0c\u5247 \u5f9e\u8868\u4e2d\u53ef\u4ee5\u78ba\u5b9a(\u4fee\uf9e4 (\u7238\u7238 \u7684 \u81ea\ufa08\uf902))\u8207((\u4fee\uf9e4 \u81ea\ufa08\uf902)\u7684 \u7238\u7238)\u5206 \u5225\u662f\uf906 1 \u8207\uf906 2 \u7684\u6700\u4f73\u5206\u6790\u7d50\u679c\u3002 \u683c\u5f0f\uff0c\u9084\u6709\u5f88\u591a\u5176\u4ed6\u7684\u6b67\u7fa9\u7d50\u69cb\u6709\u5f85\u7e3d\u7d50\u8207\u8655\uf9e4\u3002\u540c\u6642\u4e00\u4e9b\u5982\"\u5403\u98df\u5802\uff02\uff0c\"\u5403\u5927\u9910\uff02 3.8 \u5c0f\u7d50 \u7b49\u5177\u6709\u7279\u6b8a\u7684\u8a9e\u7fa9\u683c\u8f49\u63db\u6b67\u7fa9\u7684\u73fe\u8c61\uff0c\u9700\u8981\u9032\u4e00\u6b65\u7684\u6df1\u5165\u7814\u7a76\u3002</td></tr><tr><td colspan=\"2\">\u8868\u660e\u88ab\u4fee\u98fe\u7684 NP \u5728 VP \uf967\u80fd\u5145\u7576\u5408\u9069\u7684\u8a9e\u7fa9\u6210\u5206\uff0c\u6216\u8a9e\u7fa9\u6210\u5206\uf967\u80fd\u78ba\u5b9a\u3002\u9019\u7a2e\u60c5\u6cc1\u5982 \u672c\u6f14\u7b97\u6cd5\u5177\u6709\u5982\u4e0b\u7279\u9ede\uff1a 2\uff1a\u672c\u6587\u7684\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b\u4e3b\u8981\u91dd\u5c0d\u5be6\u7fa9\u8a5e\u9032\ufa08\uff0c\u5c0d\u65bc\u865b\u8a5e\u8a5e\u8a9e\u6b67\u7fa9\u554f\u984c\u6c92\u6709\u904e\u591a\u5730</td></tr><tr><td colspan=\"2\">1\uff1a\u4ed6\u4e2d\u734e\u7684\u6d88\u606f\uf9f7\u523b\u50b3\u958b\uf9ba\u3002(\"\u6d88\u606f\uff02\u7232\"\u4ed6\u4e2d\u734e\uff02\u7684\u540c\u6307) 3.7 \u5be6\u9a57\u7d50\u679c\u53ca\u8a0e\uf941 (1) \u5728\u300a\u77e5\u7db2\u300b\u4e2d\u8a5e\u8a9e\u7684\u7fa9\u9805\u7531\u591a\u500b\u7fa9\u539f\u5b9a\u7fa9\uff0c\uf967\u8c61\u50b3\u7d71\u7684\u5206\uf9d0\u8a5e\u5178\u4e2d\u7fa9\u9805\u53ea\u662f\u4e00 \u8003\u616e\u3002</td></tr><tr><td colspan=\"2\">2\uff1a\u5275\u4f5c\u65b9\u6cd5\u5f88\u91cd\u8981\u3002 (\"\u65b9\u6cd5\uff02\u5be6\u969b\u4e0a\u662f\"\u5275\u4f5c\uff02\u7684\"\u65b9\u5f0f\uff02\u6210\u5206\uff0c\u4f46\u5275\u4f5c\u7684\u898f\u5247 \u4e2d\u6c92\u6709\u5c0d\"\u65b9\u5f0f\uff02\u7684\u9650\u5236\u63cf\u8ff0) \u500b\uf9d0\u4ee3\u78bc\uff0c\u9019\u6a23\u5c0d\u8a5e\u8a9e\u7fa9\u9805\u7684\u610f\u7fa9\u63cf\u8ff0\uf901\u52a0\u5168\u9762\uff0c\u8c50\u5bcc\u3002 3\uff1a\u672c\u6587\u8655\uf9e4\u90fd\u662f\u55ae\uf906\u6f22\u8a9e\u7684\u6b67\u7fa9\u6392\u6b67\uff0c\u672a\u6d89\u53ca\u5230\u7bc7\u7ae0\u7d1a\u7684\u4e0a\u4e0b\u6587\u8a9e\u5883\uf9e4\u89e3\u3002\u800c\u5be6 3.7.1 \u5be6\u9a57\u7d50\u679c (2) \uf9dd\u7528\u7fa9\u539f\u7684\u898f\u5247\u8207\u7576\u524d\u8a5e\u8a9e\u6240\u5728\u8a9e\u7fa9\u74b0\u5883\u9032\ufa08\u76f8\u4f3c\ufa01\u7684\u6bd4\u8f03\u9032\ufa08\u8a5e\u7fa9\u6392\u6b67\uff0c\u53ef \u969b\u4e0a\uf967\u5c11\u6b67\u7fa9\u9700\u8981\u653e\u5728\u4e0a\u4e0b\u6587\u4e2d\u624d\u80fd\u5f97\u4ee5\u6d88\u9664\u3002</td></tr><tr><td colspan=\"2\">\u5c0d\u65bc\u4e0a\u8ff0\u9019\uf978\u7a2e\u7684\u60c5\u6cc1\uff0c\u53ef\u4ee5\u5c07\u8a5e\u8a9e\u8207\u901a\u7528\u683c\u898f\u5247\u9032\ufa08\u5339\u914d\uff0c\u5177\u6709\u6700\u5927\u5339\u914d\u76f8\u4f3c\ufa01 \u5728\u7fa9\u539f\u7684\u540c\u73fe\u96c6\u5408\u7372\u53d6\u4e2d\u6211\u5011\u4f7f\u7528\u300a\uf95a\u8005 20 \uf98e\u6587\u96c6\u300b\u4f5c\u7232\u7d71\u8a08\u8a9e\uf9be\u5eab\uff0c\u6e2c\u8a66\u6642\u4e5f\u4f7f\u7528\u540c \uf9d0\u7684\u8a9e\uf9be\u3002\u6e2c\u8a66\u76ee\u7684\u662f\u6aa2\u9a57\u672c\u6587\u63d0\u51fa\u7684\u6392\u6b67\u6f14\u7b97\u6cd5\u7684\u662f\u5426\u6709\u6548\uff0c\u540c\u6642\u901a\u904e\u4e2d\u9593\u8a9e\u8a00\u7684\u7d50 \u4ee5\u63d0\u9ad8\u8907\u96dc\uf906\u5f0f\u7d50\u69cb\u7684\u8a5e\u7fa9\u6392\u6b67\u6b63\u78ba\uf961\u3002 \u4ee5\u4e0a\u9019\u4e9b\u554f\u984c\uff0c\u90fd\u6709\u5f85\u5728\u5f8c\u7e7c\u7684\u5de5\u4f5c\u4e2d\uf967\u65b7\u5730\u52a0\u4ee5\u88dc\u5145\u8207\u6539\u9032\uff0c\u4f7f\u7528\u8a9e\u7fa9\u6392\u6b67\u6a21\u578b \u6240\u5c0d\u61c9\u7684\u683c\u53ef\u4f5c\u7232\u88ab\u4fee\u98fe\u8a5e\u5728 VP \u4e2d\u6240\u5145\u7576\u7684\u8a9e\u7fa9\u6210\u5206\u3002 \u69cb\u512a\u53d6\u53ca\u8abf\u6574\u7684\u6b63\u78ba\uf961\uf92d\u8003\u67e5\u8a5e\u7fa9\u8a55\u50f9\u6a21\u578b\u662f\u5426\u5408\uf9e4\u3002\u6211\u5011\u5f9e\u8a9e\uf9be\u5eab\u4e2d\u9078\u53d6\uf9ba 2\uff0c000 (3) \u5c07\u6392\u6b67\u77e5\uf9fc\u5efa\uf9f7\u5728\u7fa9\u539f\u7684\u57fa\u790e\u4e0a\uff0c\u7fa9\u539f\u7684\uf969\u76ee\u662f\u6709\u9650\u7684\uff0c\u9019\u6a23\u907f\u514d\uf9ba\u624b\u5de5\u7de8\u5236 \uf901\u52a0\u6709\u6548\u3001\u5be6\u7528\u3002</td></tr><tr><td colspan=\"2\">\u5927\u898f\u6a21\u8a5e\u7fa9\u6392\u6b67\u77e5\uf9fc\u7684\u7e41\u91cd\uf92f\u52d5\u3002\u540c\u6642\u7fa9\u539f\u7684\u6392\u6b67\u77e5\uf9fc\u662f\uf96b\u8003\u7fa9\u539f\u7684\u540c\u73fe\u96c6 \u500b\u6e2c\u8a66\uf906\u9032\ufa08\u6392\u6b67\u5be6\u9a57\u3002\u4e0b\u8868\u662f\u6e2c\u8a66\u7684\u6307\u6a19\u53ca\u6e2c\u8a66\u7d50\u679c\uff1a 3.6 \uf9dd\u7528\u8a5e\u7fa9\u6392\u6b67\u9032\ufa08\u7d50\u69cb\u6392\u6b67 \u4e00\u500b\u6e2c\u8a66\uf906\u7d93\u904e\u8a9e\u6cd5\u5206\u6790\u5f8c\u6709\u53ef\u80fd\u7523\u751f\u591a\u500b\u4e2d\u9593\u7d50\u679c\uff0c\u9019\u5c31\u9700\u8981\u5c0d\u4e2d\u9593\u7d50\u679c\u9032\ufa08\u8a55\u50f9\uff0c \u5f9e\u4e2d\u512a\u9078\u51fa\u4e00\u500b\u6700\u512a\u7d50\u679c\u3002\u5373\u9032\ufa08\u7d50\u69cb\u6392\u6b67\u3002 \u5408\uff0c\u800c\u9019\u4e00\u96c6\u5408\u662f\u901a\u904e\u5c0d\u8a9e\uf9be\u5eab\u7121\u6307\u5c0e\u5b78\u7fd2\u7372\u53d6\u7684\u3002\u9019\u6a23\u77e5\uf9fc\u7684\u7372\u53d6\u7684\u5de5\u4f5c\uf97e \uf96b\u8003\u6587\u737b \u8868 3.7 \u6e2c\u8a66\u7d50\u679c \u6e2c\u8a66\u6307\u6a19 \u6307\u6a19\u63cf\u8ff0 \u9032\u4e00\u6b65\u7684\u6e1b\u5c11\uf9ba\u3002 \u99ae\u5fd7\u5049 &lt;\uf941\u6b67\u7fa9\u7d50\u69cb\u7684\u6f5b\u5728\u6027&gt;\u300a\u4e2d\u6587\u8cc7\u8a0a\u5b78\u5831\u300b\uff0c1995\uff0c\u7b2c 9 \u5377(4) \u6e2c\u8a66\u503c (4) \uf9dd\u7528\u8a5e\u8a9e\u7fa9\u9805\u7684\u8a55\u50f9\u6f14\u7b97\u6cd5\uff0c\u5728\u8a5e\u7fa9\u6392\u6b67\u904e\u7a0b\u4e2d\u53ef\u4ee5\u5c0d\u4e2d\u9593\u8a9e\u8a00\u7684\u8a9e\u7fa9\u6846\u67b6\u505a \uf980\u53d4\u6e58 &lt;\u6b67\u7fa9\uf9d0\u578b&gt;\u300a\u4e2d\u570b\u8a9e\u6587\u300b1984 \uf98e\u7b2c 5 \u671f</td></tr><tr><td colspan=\"2\">\u6f22\u8a9e\u7684\u7d50\u69cb\u6b67\u7fa9\u932f\u7d9c\u8907\u96dc\uff0c\u8a31\u591a\u7684\u6f22\u8a9e\u8a00\u6587\u5b78\u7814\u7a76\u5b78\u8005\u90fd\u5c0d\u5176\u9032\ufa08\u6df1\u5165\u7684\u7814\u7a76\uff0c\u4e26 \u8a5e\u7fa9\u6392\u6b67\u7684\u6b63\u78ba\uf961 \u8a5e\u7fa9\u5224\u65b7\u6b63\u78ba\u7684\u8a5e\u8a9e\uf969/\u6e2c\u8a66\u8a9e\uf9be\u4e2d\u6b67\u7fa9\u8a5e\u7684\u7e3d 0.92 \u9069\u7576\u8abf\u6574\u3002 \u99ae\u5fd7\u5049 &lt;\u6b67\u7fa9\u6d88\u89e3\u7b56\uf976\u521d\u63a2&gt;\u300a\u8a08\u7b97\u8a9e\u8a00\u5b78\u9032\u5c55\u8207\u61c9\u7528\u300b\u6e05\u83ef\u5927\u5b78\u51fa\u7248\u793e\uff0c1995</td></tr><tr><td colspan=\"2\">\u7e3d\u7d50\uf9ba\u8a31\u591a\u7684\u6b67\u7fa9\u77ed\u8a9e\u7d44\u5408\u683c\u5f0f\u3002 \uf969 (5) \u5728\u9032\ufa08\u8a5e\u7fa9\u6392\u6b67\u7684\u540c\u6642\u89e3\u6c7a\u591a\u500b\u5206\u6790\u7d50\u679c\u7684\u7d50\u69cb\u6392\u6b67\u3002 \u82d1\u6625\u6cd5\uff0c\u9ec3\u9326\u8f1d\uff0c\uf9e1\u6587\u6377 &lt;\u57fa\u65bc\u8a9e\u7fa9\u77e5\uf9fc\u7684\u6f22\u8a9e\uf906\u6cd5\u7d50\u69cb\u6392\u6b67&gt; \u300a\u4e2d\u6587\u8cc7\u8a0a\u5b78\u5831\u300b\u7b2c 13</td></tr><tr><td colspan=\"2\">\u672c\u6587\u5c0d\u7d50\u69cb\u6b67\u7fa9\u7684\u6d88\u9664\u6f14\u7b97\u6cd5\u662f\u5efa\uf9f7\u5728\u8a5e\u7fa9\u6392\u6b67\u7684\u57fa\u790e\u4e4b\u4e0a\u3002\u6700\u4f73\u7684\u4e2d\u9593\u7d50\u679c\u61c9\u662f \u6700\u7b26\u5408\u8a9e\u7fa9\u8207\u5e38\uf9fc\u7684\uff0c\u800c\u4e2d\u9593\u7d50\u679c\"\u512a\u9078\uff02\u7684\u539f\u5247\u4e5f\u61c9\u662f\u9078\u64c7\u6700\u6eff\u8db3\u8a9e\u7fa9\u7684\u7d50\u69cb\u3002\u5728\"\u7d04 \u7d50\u69cb\u6392\u6b67\u7684\u6b63\u78ba\uf961 \u7d50\u69cb\u9078\u64c7\u6b63\u78ba\u7684\uf906\uf969/\u6e2c\u8a66\u8a9e\uf9be\u4e2d\u6709\u591a\u500b\u5019\u9078\u5206 \u6790\u7d50\u69cb\u7684\uf906\uf969 4. \u7e3d\u7d50 0.82 \u5377\u7b2c 1 \u671f</td></tr><tr><td colspan=\"2\">\u675f\uff02\u6392\u6b67\u53ca\u672c\u7ae0\u524d\u9762\u4ecb\u7d39\u7684\u8a5e\u7fa9\u6392\u6b67\u4e2d\uff0c\u6211\u5011\u5c0d\u8a5e\u8a9e\u9032\ufa08\u7fa9\u9805\u9078\u64c7\u7684\u540c\u6642\u9084\u5f97\u5230\uf9ba\u8a5e\u8a9e \u7d50\u69cb\u8abf\u6574\u7684\u53ec\u56de\uf961 \u9032\ufa08\u7d50\u69cb\u8abf\u6574\u7684\u7d50\u679c\uf969/\u6e2c\u8a66\u8a9e\uf9be\u4e2d\u9700\u9032\ufa08\u8abf\u6574 0.95 \u6f22\u82f1\u6a5f\u5668\u7ffb\u8b6f\u4e2d\u8981\u89e3\u6c7a\u5206\u6790\u7523\u751f\u7684\u8fad\u5f59\u6b67\u7fa9\u3001\u8a9e\u7fa9\u6b67\u7fa9\uff0c\u5f97\u5230\u4e00\u500b\u6bd4\u8f03\u597d\u7684\uf906\u6cd5\u5206\u6790\u7d50</td></tr><tr><td colspan=\"2\">\u7684\u8a55\u50f9\u5206\u503c\uff0c\u5b83\u5011\u53cd\u6620\uf9ba\u8a5e\u8a9e\u5728\u7576\u524d\u8a9e\u5883\u4e2d\u7b26\u5408\u8a9e\u7fa9\u7684\u7a0b\ufa01\u3002\u524d\u6587\u4e2d\u5b9a\u7fa9\uf9ba\u8a5e\u8a9e\u7684\u9650\u5b9a \u95dc\u4fc2\uff0c\u90a3\u4e9b\u8d77\u9650\u5b9a\u4f5c\u7528\u7684\u8a5e\u8a9e\u5728\u7576\u524d\u7d50\u69cb\u4e2d\u8d77\u8457\u95dc\u9375\u7684\u642d\u914d\u4f5c\u7528\uff0c\u5b83\u5011\u5c0d\u7d50\u69cb\u7684\u8a9e\u7fa9\u5177 \u7684\u5206\u6790\u7d50\u679c\uf969 \u7d50\u69cb\u8abf\u6574\u7684\u6b63\u78ba\uf961 \u7d93\u81ea\u52d5\u8abf\u6574\u5f8c\u6b63\u78ba\u7684\u7d50\u679c\uf969/\u9032\ufa08\u81ea\u52d5\u7d50\u69cb\u8abf\u6574 \u679c\uff0c\u5fc5\u9808\u8981\u5f15\u9032\u8a9e\u7fa9\u77e5\uf9fc\uff0c\u9032\ufa08\u8a9e\u7fa9\u5206\u6790\u3002\u672c\u6587\u69cb\u9020\uf9ba\u4e00\u500b\u7528\u65bc\u6a5f\u5668\u7ffb\u8b6f\u6587\u672c\u5206\u6790\u7684\u8a9e 0.98 \u7fa9\u6392\u6b67\u6a21\u578b\uff0c\u5b83\u80fd\u5920\u7d50\u5408\u8a9e\u7fa9\u77e5\uf9fc\u9032\ufa08\u6709\u6548\u7684\u8a5e\u7fa9\u6d88\u6b67\uff0c\u9032\u4e00\u6b65\u9032\ufa08\u7d50\u69cb\u6d88\u6b67\u3002 \u7684\u7d50\u679c\uf969 \u6709\u6700\u76f4\u63a5\u7684\u5f71\u97ff\uff0c\u56e0\u6b64\u6211\u5011\u5728\u5c0d\u4e00\u500b\u4e2d\u9593\u7d50\u679c\u9032\ufa08\u8a5e\u7fa9\u6392\u6b67\u5f8c\uff0c\u53ef\u4ee5\u5c07\u6392\u6b67\u5f97\u5230\u7684\u5404\u9650 \u5b9a\u8a5e\u7684\u8a55\u50f9\u5206\u503c\u7684\u7e3d\u548c\u4f5c\u7232\u7576\u524d\u5206\u6790\u7d50\u679c\u7684\u8a55\u50f9\u503c\u3002\u8a55\u50f9\u503c\u7684\u5927\u5c0f\u662f\u512a\u9078\u7684\u6839\u64da\uff0c\u5177\u6709 \u6e2c\u8a66\u7d50\u679c\u8868\u660e\uff0c\uf9dd\u7528\u8a9e\uf9be\u5eab\u7684\u540c\u73fe\u7fa9\u539f\uf92d\u69cb\u9020\u7fa9\u539f\u7684\u8a9e\u7fa9\u9650\u5236\u898f\u5247\uff0c\u4e26\u4ee5\u6b64\u9032\ufa08\u8a5e \u672c\u6587\u63d0\u51fa\u7684\u8a9e\u7fa9\u6392\u6b67\u7cfb\u7d71\u6709\u4ee5\u4e0b\u7279\u9ede\uff1a</td></tr><tr><td colspan=\"2\">\u6700\u9ad8\u8a55\u50f9\u503c\u7684\u4e2d\u9593\u5206\u6790\u7d50\u69cb\u5c31\u4f5c\u7232\u6700\u7d42\u7684\u7d50\u69cb\u6392\u6b67\u7d50\u679c\u3002 \u7fa9\u6392\u6b67\u7684\u601d\u60f3\u662f\u5408\uf9e4\u7684\u3002\u4e26\u4e14\u5728\u8a5e\u7fa9\u6392\u6b67\u7684\u904e\u7a0b\u4e2d\u540c\u6642\u4e5f\u80fd\u5be6\u73fe\u9ad8\u6b63\u78ba\uf961\u7684\u7d50\u69cb\u6392\u6b67\u3001 1\uff1a\u5728\uf906\u6cd5\u5206\u6790\u904e\u7a0b\u4e2d\u9032\ufa08\u7684\u7d04\u675f\u6392\u6b67\u5927\u5927\u6e1b\u5c11\uf9ba\u4e2d\u9593\u8a9e\u8a00\u7684\u751f\u6210\uf969\u76ee\uff0c\u6e1b\u8f15\uf9ba\u91dd</td></tr><tr><td>\uf9b5\u5982\uff1a\"vp+np+\u7684+np\uff02\u7d50\u69cb\u53ef\u80fd\u5b58\u5728\uf978\u7a2e\u6b67\u7fa9\uff1a \u7d50\u69cb\u8abf\u6574\u3002 \u5c0d\u4e2d\u9593\u7d50\u679c\u9032\ufa08\u7684\"\u512a\u9078\uff02\u5de5\u4f5c\u3002</td><td/></tr><tr><td colspan=\"2\">1\uff1a 2\uff1a 2\uff1a\u5728\"\u512a\u9078\uff02\u7684\u65b9\u6cd5\u4e2d\u6211\u5011\uf9dd\u7528\uf9ba\u5f9e\u8a9e\uf9be\u5eab\u4e2d\u69cb\u9020\u7fa9\u9805\u8a9e\u7fa9\u9650\u5236\u898f\u5247\u7684\u65b9\u6cd5\uff0c\u6e1b (vp (np \u7684 np)) \u5982\"\u4fee\uf9e4\u7238\u7238\u7684\u81ea\ufa08\uf902\uff02 3.7.2 \u6f14\u7b97\u6cd5\u5b58\u5728\u7684\u554f\u984c \u8f15\uf9ba\u4eba\u5de5\u5236\u5b9a\u8a9e\u7fa9\u77e5\uf9fc\u7684\u5de5\u4f5c\uf97e\u3002\u540c\u6642\u8a9e\u7fa9\u898f\u5247\u4e5f\u907f\u514d\u51fa\u73fe\u55ae\u7d14\uf9dd\u7528\u7d71\u8a08\u9032\ufa08\u8a9e\u7fa9\u6392\u6b67 ((vp np) \u7684 np) \u5982\"\u4fee\uf9e4\u81ea\ufa08\uf902\u7684\u7238\u7238\uff02 (\uff11)\uff1a\u6f14\u7b97\u6cd5\u627e\u51fa\u7684\u7fa9\u539f\u898f\u5247\u662f\u5177\u6709\u666e\u904d\u642d\u914d\u95dc\u4fc2\u7684\u8a5e\u7fa9\u6392\u6b67\u898f\u5247\uff0c\u6211\u5011\u901a\u904e\u7fa9\u539f\u7684\u9650 \u6642\u5c0d\u8907\u96dc\uf906\u5f0f\u7684\u8655\uf9e4\u6548\u679c\uf967\uf9e4\u60f3\u7684\u73fe\u8c61\u3002</td></tr><tr><td colspan=\"2\">\u9019\uf978\uf906\u8a71\u53ef\u80fd\u5b58\u5728 4 \u7a2e\u6b67\u7fa9\u7d50\u69cb\uff0c\u5c0d\u5b83\u5011\u5206\u5225\u9032\ufa08\u8a5e\u7fa9\u6392\u6b67\uff0c\u5f97\u5230\u7684\u8a55\u50f9\u60c5\u6cc1\u5982\u4e0b\uff1a \u5236\uf92d\u69cb\u9020\u7fa9\u9805\u7684\u898f\u5247\uff0c\u6709\u53ef\u80fd\u7fa9\u539f\u898f\u5247\u9650\u5236\u7684\uf9f9\ufa01\u6bd4\u7fa9\u9805\u61c9\u6709\u8a9e\u7fa9\u9650\u5236\u5927\uff1b\u9019\u5c31\u9020\u6210\uf9ba \u7523\u751f\u7684\u7fa9\u9805\u898f\u5247\u8a9e\u7fa9\uf9f9\ufa01\u904e\u5927\u3002\uf9b5\u5982\u8c61\"\u627e\uff02\u4e00\u8a5e\uff0c\u5b83\u5728\u4f5c\"\u627e\uf9ba\uf978\u5143\u9322\uff02\u6642\u7684\u7fa9\u9805\u7684 3\uff1a\u6392\u6b67\u904e\u7a0b\u4e2d\u7d66\u51fa\u7684\u7fa9\u9805\u8a55\u50f9\u503c\u53ef\u4ee5\u4f7f\u5f97\u7cfb\u7d71\u5728\u5b8c\u6210\u6392\u6b67\u7684\u540c\u6642\u53ef\u4ee5\u9032\ufa08\u5225\u7684\u8207</td></tr></table>",
+                "text": "Dan Roth. \"Learning to Resolve Natural Language Ambiguities: A Unified Approach.\" AAAI-98, 1998 \u8d99\u9435\u8ecd\u7b49\u300a\u6a5f\u5668\u7ffb\u8b6f\u539f\uf9e4\u300b\uff0e\u54c8\u723e\u6ff1\uff1a\u54c8\u723e\u6ff1\u5de5\u696d\u5927\u5b78\u51fa\u7248\u793e\uff0c2000 Wilks, Y. Stevenson, M. \"Word Sense Disambiguation Using Optimized Combinations of Knowledge Sources\", In Proceedings of joint COLING-ACL'98. 1998. Montreal, Canada. Philip Resnik ,David Yarowsky. A Perspective on Word Sense Disambiguation Methods and Their Evaluation. Proceedings of the SIGLEX Workshop \"Tagging Text with Lexical Semantics: What, why and how?\", pp. 79-86, Washington, D.C. \u8463\u632f\u6771\uff0c\u8463\u5f37\uff0e\u300a\u77e5\u7db2\u300b\uff0ehttp://www.how-net.com \uf9e1\u6d93\u5b50\uff0c\u9ec3\u660c\uf95f \uff1c\u57fa\u65bc\u8f49\u63db\u7684\u7121\u6307\u5c0e\u8a5e\u7fa9\u6a19\u6ce8\u65b9\u6cd5\uff1e\u300a\u6e05\u83ef\u5927\u5b78\u5b78\u5831\u300b(\u81ea\u7136\u79d1\u5b78\u7248) \uff0c 1999 \uf98e\u7b2c 39 \u5377(7) \u6885\u5bb6\u99d2 \u300a\u73fe\u4ee3\u6f22\u8a9e\u642d\u914d\u8fad\u5178\u300b\u6f22\u8a9e\u5927\u8a5e\u5178\u51fa\u7248\u793e. 1999 \uf98e 12 \u6708\u7b2c 1 \u7248"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O02/O02-1004.json

@@ -0,0 +1,532 @@
+{
+    "paper_id": "O02-1004",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:05:50.391784Z"
+    },
+    "title": "Cross-Language Text Filtering Based on Text Concepts and kNN",
+    "authors": [
+        {
+            "first": "",
+            "middle": [],
+            "last": "\u8607\u5049\u5cf0",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Xiamen University",
+                "location": {
+                    "addrLine": "Xiamen\uff0c361005"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Weifeng",
+            "middle": [],
+            "last": "Su",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Xiamen University",
+                "location": {
+                    "addrLine": "Xiamen\uff0c361005"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Shaozi",
+            "middle": [],
+            "last": "Li",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Xiamen University",
+                "location": {
+                    "addrLine": "Xiamen\uff0c361005"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Tanqiu",
+            "middle": [],
+            "last": "Li",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Xiamen University",
+                "location": {
+                    "addrLine": "Xiamen\uff0c361005"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Wenjian",
+            "middle": [],
+            "last": "You",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Xiamen University",
+                "location": {
+                    "addrLine": "Xiamen\uff0c361005"
+                }
+            },
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "The WWW is increasingly being used source of information. The volume of information is accessed by users using direct manipulation tools. It is obviously that we'd like to have a tool to keep those texts we want and remove those texts we don't want from so much information flow to us. This paper describes a module that sifts through large number of texts retrieved by the user.",
+    "pdf_parse": {
+        "paper_id": "O02-1004",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "The WWW is increasingly being used source of information. The volume of information is accessed by users using direct manipulation tools. It is obviously that we'd like to have a tool to keep those texts we want and remove those texts we don't want from so much information flow to us. This paper describes a module that sifts through large number of texts retrieved by the user.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "The module is based on HowNet, a knowledge dictionary developed by Mr. Zhendong Dong. In this dictionary, the concept of a word is divided into sememes. In the philosophy of HowNet, all concepts in the world can be expressed by a combination more than 1500 sememes. Sememe is a very useful concept in settle the problem of synonym which is the most difficult problem in text filtering. We classified the set of sememes into two sets of sememes: classfiable sememes and unclassficable semems. Classfiable sememes includes those sememes that are more 80 \u8607\u5049\u5cf0 \u7b49 useful in distinguishing a document's class from other documents. Unclassfiable sememes include those sememes that have similar appearance in all documents. Classfiable includes about 800 sememes. We used these 800 classficable sememes to build Classficable Sememes Vector Space (CSVS) .",
+                "cite_spans": [
+                    {
+                        "start": 837,
+                        "end": 843,
+                        "text": "(CSVS)",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "A text is represented as a vector in the CSVS after the following step:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "1. text preprosessing: Judge the language of the text and do some process attribute to its language.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "2. Part-of-Speech tagging 3. keywords extraction 4. keyword sense disambiguation based on its environment by calculating its classifiable sememes relevance with it's environment's classifiable sememes.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "We add the weight of a semantic item if there are classifiable sememes the same as classifiable sememe in the its environment word's semantic item. This is not a strict disambiguation algorithm. We just adjust the weights of those semantic items.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "5. Those keywords are reduced to sememes and the weight of all keywords 's all semantic items 's classifiable sememes are calculated to be the weight of its vector feature.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "A user provides some texts to express the text he interested in. They are all expressed as vectors in the CSVS. Then those vectors represent the user's preference. The relevance of two texts can be measured by using the cosine angle between the two text's vectors. When a new text comes, it is expressed as a vector in CSVS too. We find its k nearest neighbours in the texts provided by the user in the CSVS . Calculating the relevance of the new text to its k nearest neighbours and if it is bigger than a certain valve, than it means it is of the user's interest if smaller, it means that it is not belong to the user's interesting. The k is determined by calculated every training vector its neighbours.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Information filtering based on classifiable sememes has several advantage:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "1. Low dimentional input space. We use 800 sememes instead of 10000 words.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "2. Few irrelevant feature after the keyword extraction and unclassifiable sememes's removal.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "3. Document vector's feature's weight are big.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "We made use of documents from eight different users in our experiments. All these users provides texts both in Chinese and English. We took into account the user's feedback and got a result of about 88 percent of recall and precision. It demonstrates that this is a success method. ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u57fa\u65bc\u6587\u672c\u6982\uf9a3\u548ckNN\u7684\u8de8\u8a9e\u7a2e\u6587\u672c\u904e\uf984 81",
+                "sec_num": null
+            },
+            {
+                "text": "D \uff0c\u800c\u6bcf\u4e00\u500b\u5206\uf97e d i \u662f\u77e5\u7db2\u4e2d\u7684\u4e00\u500b\u53ef\u5206\u7fa9\u539f\uff0c\u90a3\u6587\u672c\u5c31\u8868\u793a\u6210\u5411\uf97e \u2192 V \uff0c\u5176\u5206\uf97e v i \u7232 \u5c0d\u61c9\u65bc d i \u7684\u503c\uff0c\uf974\u6587\u672c\u4e2d\u6c92\u6709\u5305\u542b d i \uff0c\u5247 v i =0\u3002 \u7136\u800c\u4e26\u975e\u6587\u4ef6\u7576\u4e2d\u6240\u6709\u7684\u8a5e\u90fd\u7528\u65bc\u69cb\u9020\u6587\u672c\u5411\uf97e\uff0c\u53ea\u6709\u90a3\u4e9b\u6700\u80fd\u4ee3\u8868\u6587\u4ef6\u6240\u8981\u8868\u9054 \u7684\u610f\u601d\u7684\u8a5e\u4e5f\u5c31\u662f\u95dc\u9375\u5b57\u5f59\u53ef\u88ab\u7528\uf92d\u69cb\u9020\u5411\uf97e\u3002\u6211\u5011\u53ef\u4ee5\u63a1\u7528\u7d71\u8a08\u7684\u65b9\u6cd5\uf92d\u6c7a\u5b9a\u54ea\u4e9b\u8fad \u5f59\u662f\u95dc\u9375\u5b57\u5f59\uff0c\u9084\u6709\uff0c\u7531\u65bc\u8fad\u5f59\u7684\u5c90\u7fa9\uff0c\u6211\u5011\u4e5f\u8981\u4f5c\u4e00\u5b9a\u7a0b\ufa01\u4e0a\u7684\u6392\u5c90\u3002\u6587\u672c\u8868\u793a\u65b9\u6cd5",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u57fa\u65bc\u6587\u672c\u6982\uf9a3\u548ckNN\u7684\u8de8\u8a9e\u7a2e\u6587\u672c\u904e\uf984 81",
+                "sec_num": null
+            },
+            {
+                "text": "V V V V a = (2) \u5176\u4e2d ) ( 2 , 1 text text V V \u662f\u6307\u7528\u6236\u5411\uf97e\u548c\u6587\u672c\u5411\uf97e\u7684\u5167\u7a4d\uff0c | | text V \u8868\u793a\u6587\u672c\u5411\uf97e\u7684\u6a19\uf97e\u3002 \u5728\u6587\u672c\u904e\uf984\u7576\u4e2d\uff0c\u6211\u5011\u63a1\u7528\uf9ba k \u500b\u6700\u8fd1\u9130\u5c45(kNN)\u7684\u65b9\u6cd5\uff1a\u5c0d\u65bc\u67d0\u4e00\u8f38\u5165\u6587\u672c s\uff0c \u6309\u7167\u4e0a\u9762\u6240\u8ff0\u7684\u65b9\u6cd5\u5c07\u5176\u8868\u793a\u7232\u53ef\u5206\u7fa9\u539f\u7a7a\u9593\u7684\u5411\uf97e\uff0c\u5728\u7528\u6236\u793a\uf9b5\u4e2d\uff0c\uf9dd\u7528\u516c\u5f0f(2)\u6311 \u9078\u51fa k(k<<m)\u500b\u8207\u4e4b\u6700\u76f8\u8fd1\u7684\u9130\u5c45\u6587\u672c\uff0c\u6839\u64da\u516c\u5f0f(3)\u8a08\u7b97\u5b83\u8207\u9019 k \u500b\u6587\u672c\u7684\u76f8\u4f3c \u7a0b\ufa01 Si\uff0c\u5176\u503c\u8d8a\u9ad8\uff0c\u5247\u6211\u5011 r \u8a8d\u7232\u5b83\u8d8a\u662f\u7528\u6236\u6240\u611f\u8208\u8da3\u7684\u6587\u672c\u3002 \u8607\u5049\u5cf0 \u7b49 \u2211 = = k i i a Si S 1 2 )) (cos( (3) \u5176\u4e2d \u23a9 \u23a8 \u23a7 = x x S 0 ) ( \u5728\u6240\u9700\u904e\uf984\u7684\u6240\u6709\u6587\u672c\u7576\u4e2d\uff0c\u6211\u5011\u53ef\u4ee5\u6839\u64da Si",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u57fa\u65bc\u6587\u672c\u6982\uf9a3\u548ckNN\u7684\u8de8\u8a9e\u7a2e\u6587\u672c\u904e\uf984 81",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Dictionary-based methods for cross-lingual information retieval",
+                "authors": [
+                    {
+                        "first": "L",
+                        "middle": [],
+                        "last": "Ballesteros",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "B"
+                        ],
+                        "last": "Croft",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1996,
+                "venue": "Proc. Of the 7 th Int. DEXA Conference on Database and Expert Systems Applications",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "L.Ballesteros,W.B. Croft. \"Dictionary-based methods for cross-lingual information retieval.\" Proc. Of the 7 th Int. DEXA Conference on Database and Expert Systems Applications ,1996.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "\u8463\u632f\u6771\u3001\u8463\u5f37 \u300a\u77e5\u7db2\u300b",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "\u8463\u632f\u6771\u3001\u8463\u5f37 \u300a\u77e5\u7db2\u300b http://www.keenage.com/html/index.html",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "A Conceptual Framework for Text Filtering",
+                "authors": [
+                    {
+                        "first": "Douglas",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Oard",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Gary",
+                        "middle": [],
+                        "last": "Marchionini",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Douglas W.Oard, Gary Marchionini, \"A Conceptual Framework for Text Filtering.\" http://citeseer.nj.nec.com \u5f35\u6708\u5091\u3001\u59da\u5929\u9806 \uff1c\u57fa\u65bc\u7279\u5fb5\u76f8\u95dc\u6027\u7684\u6f22\u8a9e\u6587\u672c\u81ea\u52d5\u5206\uf9d0\u6a21\u578b\u7684\u7814\u7a76\uff1e\u300a\u5c0f\u578b\u5fae\u578b\u96fb\u8166\u7cfb \u7d71\u300b\uff0c1998 \uf98e\u7b2c 8 \u671f",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Texts Filtering using Linguistically-Motivated Indexing Terms",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [
+                            "T"
+                        ],
+                        "last": "Armapatzis",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Th",
+                        "middle": [
+                            "P"
+                        ],
+                        "last": "Van Der Weide",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [
+                            "H A"
+                        ],
+                        "last": "Koster",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "P",
+                        "middle": [],
+                        "last": "Van Bommel",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "A.T.Armapatzis and Th.P. van der Weide and C.H.A.Koster and P.van Bommel. \"Texts Filtering using Linguistically-Motivated Indexing Terms.\" http://citeseer.nj.nec.com",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "A Learning Personal Agent for Texts Filtering and Notification",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Anandeep",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Katia",
+                        "middle": [],
+                        "last": "Pannu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Sycara",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Anandeep S.Pannu and Katia Sycara. \"A Learning Personal Agent for Texts Filtering and Notification.\" http://citeseer.nj.nec.com",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Natural Laguage Understanding. The Benjamin",
+                "authors": [
+                    {
+                        "first": "James",
+                        "middle": [],
+                        "last": "Allen",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "James Allen, Natural Laguage Understanding. The Benjamin/Cumming Publishing Company, Inc.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Texts categorization with support vector machines: Learning with many relevant features",
+                "authors": [
+                    {
+                        "first": "Thorsten",
+                        "middle": [],
+                        "last": "Joachims",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Thorsten Joachims. \"Texts categorization with support vector machines: Learning with many relevant features.\" http://citeseer.nj.nec.com",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "On Automatic Filtering of Multilingual",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [],
+                        "last": "Douglas",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Nicholas",
+                        "middle": [],
+                        "last": "Oard",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Declaris",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Douglas W.Oard and Nicholas DeClaris. \"On Automatic Filtering of Multilingual.\" http://citeseer.nj.nec.com",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Information extraction asbasis forhigh-precision textclassification",
+                "authors": [
+                    {
+                        "first": "Ellen",
+                        "middle": [],
+                        "last": "Riloff",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Wendy",
+                        "middle": [],
+                        "last": "Lehnert",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "ACM Trams-actions on Information System",
+                "volume": "12",
+                "issue": "3",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ellen Riloff and Wendy Lehnert. \"Information extraction asbasis forhigh-precision textclassification.\" ACM Trams-actions on Information System, vol. 12, No 3, July 1994",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "Text Cateorization Using Weight Adjusted k-Nearest Neighbor Classification",
+                "authors": [
+                    {
+                        "first": "Eui-Hong ;",
+                        "middle": [],
+                        "last": "Sam",
+                        "suffix": ""
+                    },
+                    {
+                        "first": ")",
+                        "middle": [],
+                        "last": "Han",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Geoge",
+                        "middle": [],
+                        "last": "Karypis",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Vipin",
+                        "middle": [],
+                        "last": "Kumar",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Eui-Hong(Sam)Han , Geoge Karypis and Vipin Kumar. Text Cateorization Using Weight Adjusted k-Nearest Neighbor Classification. http://citeseer.nj.nec.com \u8607\u5049\u5cf0\u3001\uf9e1\u7d39\u6ecb\u3001\uf9e1\u5802\u79cb\u3001\u5c24\u6587\u5efa \uff1c\u53ef\u5206\u7fa9\u539f\u5411\uf97e\u7a7a\u9593\u4e2d\u7684\u8de8\u8a9e\u7a2e\u6587\u672c\u904e\uf984\u6a21\u578b\uff1e\u300a\u81ea \u7136\u8a9e\u8a00\uf9e4\u89e3\u8207\u6a5f\u5668\u7ffb\u8b6f\u300b2001 \uf98e",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF2": {
+                "type_str": "table",
+                "content": "<table><tr><td/><td>\u57fa\u65bc\u6587\u672c\u6982\uf9a3\u548ckNN\u7684\u8de8\u8a9e\u7a2e\u6587\u672c\u904e\uf984 \u57fa\u65bc\u6587\u672c\u6982\uf9a3\u548ckNN\u7684\u8de8\u8a9e\u7a2e\u6587\u672c\u904e\uf984</td><td>87 \u8607\u5049\u5cf0 \u7b49 89</td></tr><tr><td/><td colspan=\"2\">If km&gt;biggestequal then \u55ae\u8a5e\uff0c\u5982\u6b64\ufa09\u4f4e\u7dad\uf969\u53ef\u6975\u5927\u5730\u63d0\u9ad8\u53ec\u56de\uf961\uff0c\u9084\u6709\uff0c\u53ef\u4ee5\ufa09\u4f4e\u8a08\u7b97\u8907\u96dc\ufa01\u3002</td></tr><tr><td/><td colspan=\"2\">Begin 2. \u76f8\u95dc\u5206\uf97e\u503c\u8f03\u5927\uff1a\u6bd4\u5982\u5728\u4e00\u7bc7\u75c5\u4eba\u4e0a\u91ab\u9662\u53bb\u770b\u75c5\u7684\u6587\u672c\u88cf\uff0c\u53ef\u80fd\u6703\u6703\u51fa\u73fe\u8a31\u591a\uf9d0</td></tr><tr><td>100</td><td colspan=\"2\">biggestequal:=km; \u4f3c\"\u75c5\u4eba\uff02\u3001\"\u91ab\u751f\uff02\u3001\"\u91ab\u9662\uff02\u3001\"\u6cbb\uf9c1\uff02\u7b49\u7247\u8a9e\uff0c\u9019\u4e9b\u8a5e\u90fd\u5305\u542b\u6709\"\u91ab\u6cbb\uff02</td></tr><tr><td>80</td><td colspan=\"2\">\u5f53 x&lt;h \u65f6 \u7b49\u7fa9\u539f\uff0c\u5f9e\u800c\u4f7f\"\u91ab\u6cbb\uff02\u9019\u500b\u7fa9\u539f\u5206\uf97e\u7684\u503c\u6bd4\u8f03\u5927\uff0c\u9019\u6a23\u5c31\u80fd\u7a81\u51fa\u672c\u6587\u7684\u6240\u8981\u8b1b bigestk:=k;</td></tr><tr><td colspan=\"3\">\uf92d\u9032\ufa08\u76f8\u95dc\ufa01\u6392\u5e8f\u53cd\u994b\u7d66\u7528\u6236\uff0c\u4e5f\u53ef \u4ee5\u8a2d\u4e00\u95a5\u503c t\uff0c\u7576\u67d0\u6587\u672c\u8207\u7528\u6236\u9700\u6c42\u7684\u76f8\u95dc\ufa01\u5927\u65bc t \u6642\u5247\u8a8d\u7232\u8a72\u6587\u672c\u7b26\u5408\u7528\u6236\u9700\u6c42\uff0c\u628a\u6587 \u5f53 x&gt;=h \u65f6 end; \u8ff0\u7684\u5167\u5bb9\u4e3b\u8981\u662f\u95dc\u65bc\u91ab\uf9c1\u9019\u4e00\u65b9\u9762\u7684\uff0c\u9019\u6709\u52a9\u65bc\u63d0\u9ad8\u7cbe\u78ba\uf961\u8207\u53ec\u6709\uf961\u3002 60 ENDFOR \u8f03\u5c0f\u3002 20 4. \u904e\uf984\u6a21\u578b\u7684\u5be6\u9a57\u7d50\u679c\u53ca\u5be6\u9a57\u5206\u6790 kNN 3. \u5e72\u64fe\u9805\u8f03\u5c11\uff1a\u7d93\u904e\uf9ba\u95dc\u9375\u5b57\u63d0\u53d6\u3001\u8a5e\u8a9e\u6392\u5c90\u548c\uf967\u53ef\u5206\u7fa9\u539f\u7684\u53bb\u9664\u5f8c\uff0c\u6240\u5269\u4e0b\u7684\u7fa9 \u539f\u5927\u591a\u8207\u6587\u672c\u6709\u91cd\u8981\u7684\uf997\u7e6b\uff0c\u800c\u8207\u6587\u672c\u76f8\u95dc\ufa01\u8f03\u5c11\u7684\u5176\u4ed6\u5206\uf97e\u7684\u503c\u76f8\u6bd4\u4e4b\u4e0b\u660e\u986f 40 ? \u5411\uf97e</td></tr><tr><td colspan=\"3\">\u672c\u6309\u76f8\u95dc\ufa01\u5927\u5c0f\u7684\u9806\u5e8f\u8fd4\u56de\u7d66\u7528\u6236\uff0c\u628a\u4f4e\u65bc\u8a72\u503c\u7684\u6240\u6709\u6587\u672c\u53bb\u9664\u6216\u5b58\u5728\u67d0\u8655\u4ee5\u5099\u7528\u6236\u5728 \u6709\u7a7a\u6642\u8655\uf9e4\u3002\u6211\u5011\u53ef\u4ee5\u628a\u7528\u6236\u7684\u56de\u994b\u8003\u616e\u9032\u53bb\uff0c\uf974\u7528\u6236\u8a8d\u7232\u5e7e\u4e4e\u6240\u6709\u6211\u5011\u6240\u904e\uf984\u51fa\u7684\u6587 \u4ef6\u90fd\u662f\u4ed6\u6240\u611f\u8208\u8da3\u7684\uff0c\u5247\u6211\u5011\u53ef\u8abf\u4f4e t \u503c\uff0c\u53cd\u904e\uf92d\uff0c\uf974\u6709\u5f88\u591a\u6587\u672c\uf967\u7b26\u5408\u7528\u6236\u7684\u8208\u8da3\uff0c \u5247\u6211\u5011\u8abf\u9ad8 t \u503c\u3002 3.4 \u6587\u672c\uf9d0\u5225\u7684\u6b78\uf9d0 \u6211\u5011\u63a1\u7528 kNN \u7684\u65b9\u6cd5\u3002\u9996\u5148\u6211\u5011\u8a13\uf996\u7684\u6642\u5019\uff0c\u6211\u5011\u628a\u9019\u4e9b\u5df2\u7d93\u5206\u597d\uf9d0\u7684\u6309\u662f\u5426\u7232\u7528\u6236\u7684 \u9700\u8981\u5168\u90e8\u6309\u4e0a\u8ff0\u65b9\u6cd5\u8868\u793a\u6210\u53ef\u5206\u7fa9\u539f\u5411\uf97e\u7a7a\u9593\u7684\u5411\uf97e\uff0c\u5c0d\u4e00\u65b0\u9032\uf92d\u7684\u4e00\u500b\u65b0\u7684\u6587\u672c\uff0c\u6211 \u6211\u5011\u7372\u5f97\uf9ba\u516b\u500b\u7528\u6236\u7684\u5be6\u9a57\u8cc7\uf9be\uff0c\u9019\u516b\u500b\u7528\u6236\u90fd\u63d0\u4f9b\uf9ba\u4ed6\u6240\u611f\u8208\u8da3\u7684\u5167\u5bb9\u76f8\u8fd1\u7684\u4e2d\u82f1\u6587 \u6587\u672c\u5404 60 \u7bc7\u4f5c\u7232\u76f8\u95dc\u6587\u672c\uff0c\u53e6\u5916\u63d0\u4f9b 1000 \u7bc7\u5176\u4ed6\u5167\u5bb9\u7684\u6587\u672c\u4f5c\u7232\u5e72\u64fe\u6587\u672c\uff0c\u5176\u4e2d\u4e2d\u82f1 \u5728\u6211\u5011\u4ee5\u524d\u7684\u5de5\u4f5c\u7576\u4e2d\uff0c\u6211\u5011\u628a\u7528\u6236\u8868\u793a\u6210\u7232\u4e00\u500b\u5411\uf97e\uff0c\u4e26\u4ee5\u7528\u6236\u5411\uf97e\u8207\u6587\u672c\u5411\uf97e 0 \u7528? 1 \u7528? 2 \u7528? 3 \u7528? 4 \u7528? 5 \u7528? 6 \u7528? 7 \u7528? 8 \u7684\u593e\u89d2\uf92d\u8868\u793a\u6587\u672c\u8207\u7528\u6236\u7684\u76f8\u95dc\u6027\uff0c\u800c\u63a1\u7528\uf9ba kNN \u6280\u8853\uff0c\u53ef\u5728\u4ee5\u4e0b\u9019\u4e9b\u65b9\u9762\u9ad4\u73fe\u51fa\u5176\u512a \u6587\u5404 500 \u7bc7\uff0c\u5c0d\u65bc\u6bcf\u500b\u7528\u6236\uff0c\u6211\u5011\u4f7f\u7528\u5f9e\u5176\u6240\u63d0\u4f9b\u7684\u76f8\u95dc\u6587\u672c\u96a8\u6a5f\u62bd\u53d6\u4e2d\u82f1\u6587\u6587\u672c\u5404 30 \u52e2\uff1a \u7bc7\u69cb\u9020\u5176\u7528\u6236\u6a21\u677f\uff0c\u5176\u9918\u7684\u76f8\u95dc\u6587\u672c\u8207\u5e72\u64fe\u6587\u672c\u6df7\u96dc\u4e00\u8d77\u69cb\u6210\uf9ba\u6e2c\u8a66\u96c6\uff0c\u6211\u5011\u5c31\u60f3\u5f9e\u5176 1. \u9996\u5148\u5c0d\u65bc\u67d0\u4e00\u500b\u7528\u6236\u53ef\u80fd\u6709\u6bd4\u8f03\u5ee3\u6cdb\u7684\u8208\u8da3\uff0c\u5247\u53d6\u5176\u5e73\u5747\u5411\uf97e\u53ef\u80fd\u6703\u5c0e\u81f4\u6bd4\u8f03\u5927 \u4e2d\u904e\uf984\u51fa\u90a3\u4e9b\u76f8\u95dc\u6587\u672c\u3002 \u7684\u8aa4\u5dee\u3002 \u6211\u5011\u4f7f\u7528\uf9ba\uf978\u500b\uf96b\uf969\uf92d\u8a55\u50f9\u6211\u5011\u7684\u6a21\u578b\uff1a\u53ec\u56de\uf961\u548c\u7cbe\u78ba\uf961\u3002\u53ec\u56de\uf961\u662f\u6307\u6211\u5011\u904e\uf984\u51fa 2. \u5c0d\u65bc\u540c\u4e00\u500b\uf9b4\u57df\uff0c\uf967\u540c\u9ad4\u88c1\u7684\u6587\u7ae0\u5176\u5728\u5411\uf97e\u7a7a\u9593\u7576\u4e2d\u4e5f\u53ef\u80fd\u6709\u8f03\u5927\u7684\u5dee\u8ddd\uff0c\u53d6\u5e73 \u7684\u76f8\u95dc\u6587\u672c\u5360\u6240\u6709\u76f8\u95dc\u6587\u672c\u7684\u6bd4\uf961\uff0c\u7cbe\u78ba\uf961\u662f\u6307\u5728\u6211\u5011\u6240\u6709\u904e\uf984\u51fa\u7684\u6587\u672c\u7576\u4e2d\uff0c\u76f8\u95dc\u6587 \u5747\u5411\uf97e\u4e5f\u6703\u9020\u6210\u8f03\u5927\u7684\u8aa4\u5dee\u3002 \u672c\u6240\u5360\u7684\u6bd4\uf961\uff0c\u4e00\u822c\u800c\u8a00\uff0c\u53ec\u56de\uf961\u4e0a\u5347\uff0c\u5247\u7cbe\u78ba\uf961\u6703\u4e0b\ufa09\uff0c\u800c\u7cbe\u78ba\uf961\u4e0a\u5347\uff0c\u5247\u53ec\u56de\uf961\u6703 \u4e0b\ufa09\u3002 90 3. \u5982\u679c\u7528\u6236\u8208\u8da3\u7523\u751f\u8b8a\u5316\uff0c\u5e73\u5747\u5411\uf97e\u7684\u6539\u8b8a\u8f03\u7232\u9072\u7de9\uff0c\u4e26\u4e14\u5728\u9019\u500b\u904e\u7a0b\u7576\u4e2d\u4e5f\u6709\u8f03 \u5011\u63a1\u7528\u4e0a\u9762\u7684\u65b9\u6cd5\u8f49\u5316\u7232\u53ef\u5206\u7fa9\u539f\u5411\uf97e\u7a7a\u9593\u4e2d\u7684\u7a7a\u9593\u5411\uf97e\uff0c\u5047\u8a2d\u7232 d\uff0c\u5f9e\u4e2d\u627e\u51fa k \u500b\u8207 \u8868 1 \u5c31\u662f\u6211\u5011\u5be6\u9a57\u7684\u7d50\u679c\uff0c\u7d50\u679c\u8868\u660e\u7528\u8a72\u65b9\u6cd5\u9032\ufa08\u904e\uf984\u7684\u65b9\u6cd5\u6548\u679c\u975e\u5e38\u597d\uff0c\u7cbe\u78ba\uf961 \u5927\u7684\u8aa4\u5dee\u3002 85 \u5176\u6700\u7232\u9130\u8fd1\u7684\u5411\uf97e\uff0c\u7136\u5f8c\u6aa2\u67e5\u9019 k \u500b\u5df2\u7d93\u78ba\u5b9a\u597d\uf9d0\u5225\u7684\u5411\uf97e\u7684\uf9d0\u5225\u4f5c\u7232\u9019\u500b\u5411\uf97e\u7684\uf9d0 \u5225\u3002\u9019 k \u500b\u5411\uf97e\u7684\u6b0a\u91cd\u53ef\u4ee5\u901a\u904e\u5176\u8207 d \u7684\u76f8\u8fd1\u7a0b\ufa01\u9032\ufa08\u8ce6\u503c\u3002 \u5f88\u9ad8\uff0c\u5728\u5be6\u969b\u61c9\u7528\u7576\u4e2d\uff0c\u6211\u5011\u9084\u53ef\u4ee5\u628a\u7528\u6236\u53cd\u994b\u7684\u60c5\u6cc1\u8003\u616e\u9032\u53bb\uff0c\u5f62\u6210\u53ef\u6839\u64da\u7528\u6236\u7684\u8208 \u800c kNN \u5247\u6070\u6070\u76f8\u53cd\uff0c 80 kNN</td></tr><tr><td colspan=\"3\">\u5c45\u6642\u9593\u8907\u96dc\ufa01\u7232 O(L*N)\uff0c\u5176\u4e2d L \u662f\u53ef\u5206\u5411\uf97e\u7a7a\u9593\u7684\u53ef\u5206\u7fa9\u539f\uf969\u76ee\uff0cN \u7232\u53ef\u5206\u5411\uf97e\u7a7a\u9593\u4e2d 70 \u9130\u5c45\u3002 kNN \u662f\u4e00\u500b\u57fa\u65bc\u7bc4\uf9b5\u7684\u5b78\u7fd2\u6cd5\uff0c\u5176\u4e3b\u8981\u7684\u8a08\u7b97\uf97e\u662f\u5f9e\u5411\uf97e\u7a7a\u9593\u4e2d\u627e\u51fa k \u500b\u6700\u8fd1\u7684\u9130 \u8da3\u6539\u8b8a\u800c\u628a\u6539\u8b8a\u7528\u6236\u6a21\u677f\u5411\uf97e\u5f9e\u800c\u6539\u8b8a\u9078\u64c7\u7684\u6587\u672c\u7684\u81ea\u9069\u61c9\u7cfb\u7d71\u3002 1. \uf974\u7528\u6236\u6709\u6bd4\u8f03\u5ee3\u6cdb\u7684\u8208\u8da3\uff0c\u5247\u5728\u5411\uf97e\u7a7a\u9593\u7576\u4e2d\u5f62\u6210\uf967\u540c\u7c07\u7684\u5411\uf97e\uff0c\u5c31\u53ef\u6709\uf967\u540c\u7684 75 \u5355\u5411\uf97e</td></tr><tr><td colspan=\"3\">\u7684\u8a13\uf996\u6587\u672c\u7684\uf969\uf97e\u3002 User 1 2. \u5c0d\u65bc\u540c\u4e00\uf9b4\u57df\u800c\uf967\u540c\u9ad4\u88c1\u7684\u6587\u7ae0\uff0c\u4e5f\u53ef\u5728\u5411\uf97e\u7a7a\u9593\u4e2d\u5f62\u6210\uf967\u540c\u7c07\u7684\u5411\uf97e\uff0c\u69cb\u6210\uf967 User 2 User 3 User 4 User 5 User 6 User 7 User 8 Average \u540c\u7684\u9130\u5c45\u3002 65 \u7528\u62371 \u7528\u62372 \u7528\u62373 \u7528\u62374 \u7528\u62375 \u7528\u62376 \u7528\u62377 \u7528\u62378 k \u503c\u7684\u78ba\u5b9a\u65b9\u6cd5\uff1a \u53ec English 88.7 90 90 89 86 87 92 91 89.2 3. \uf974\u7528\u6236\u8208\u8da3\u767c\u751f\u8b8a\u5316\uff0c\u53ea\u8981\u518d\u6b21\u63d0\u4f9b\u65b0\u7684\u6240\u8208\u8da3\u7684\u6587\u672c\uff0c\u5728\u5411\uf97e\u7a7a\u9593\u7576\u4e2d\u5e7e\u4e4e\uf967 \u6211\u5011\u4e3b\u8981\u63a1\u7528\u767b\u5c71\u6cd5\uf92d\u78ba\u8a8d k \u503c\uff0c\u5728\u8a13\uf996\u6587\u672c\u5168\u90e8\u8868\u793a\u6210\u5411\uf97e\u7a7a\u9593\u7684\u5411\uf97e\u4ee5\u5f8c\uff0c\u6309 \u56de Chinese 86.6 91.5 86 85 84 87 90.6 90 87.6 \u53d7\u820a\u7684\u5411\uf97e\u7684\u5f71\u97ff\uff0c\u4e14\u53ef\u4fdd\uf9cd\u820a\u7684\u5411\uf97e\u4ee5\u5099\u53e6\u7528\u3002 \u4e0b\u9762\u6f14\u7b97\u6cd5\u9032\ufa08\u8a08\u7b97\uff1a \u6f14\u7b97\u6cd5 3 kNN \u4e2d\u7684 k \u7684\u8a08\u7b97\u6f14\u7b97\u6cd5 \uf961 \u5176\u512a\u52e2\u53ef\u5728\u5716 1 \u548c\u5716 2 \u9ad4\u73fe\u51fa\uf92d\u3002 5. \u7d50\u675f\u8a9e (%) biggestequal:=0 \u7cbe English 86 88.6 85 88.7 87.5 88.5 84.7 90 88.5 \u5f9e\u7db2\uf937\u8cc7\u8a0a\u670d\u52d9\u9700\u6c42\u51fa\u767c\uff0c\u6211\u5011\u8a8d\u7232\u6709\u5fc5\u8981\u5c0d\u8cc7\u8a0a\u6e90\u7684\u8cc7\u8a0a\u9032\ufa08\u904e\uf984\u3002\u672c\u6587\u63d0\u51fa\uf9ba\u4e00\u500b bigestk\uff1a=0\uff1b \u78ba Chinese 82 85.4 85 87.6 84.2 86.3 88.6 86.8 87.5 \u5728\u53ef\u5206\u7fa9\u539f\u7a7a\u9593\u4e2d\u63a1\u7528\u5411\uf97e\u7a7a\u9593\u6a21\u578b\u7684\u65b9\u6cd5\u9032\ufa08\u6587\u672c\u904e\uf984\u7684\u6a21\u578b\uff0c\uf9e4\uf941\u548c\u5be6\u9a57\u5747\u8868\u660e\uff0c \u7d66\u5411\uf97e\u7684\u6bcf\u500b\u5206\uf97e\u503c\u8ce6\u521d\u503c 0 \uf961 \u8a72\u6a21\u578b\u5177\u6709\u6bd4\u8f03\u597d\u7684\u904e\uf984\u6548\u679c\uff0c\u5f9e\u901f\ufa01\u548c\u670d\u52d9\u6027\u80fd\u4e0a\u9054\u5230\uf9ba\u8f03\u597d\u7a0b\ufa01\u3002 FOR k:=(\u4e00\u500b&gt;1 \u7684\u5c0f\u6574\uf969)TO (\u4e00\u500b\u5927\u6574\uf969) (%) \u5728\u6a21\u578b\u7684\u5be6\u73fe\u904e\u7a0b\u4e2d\uff0c\u6211\u5011\u767c\u73fe\u628a\u9019\u7a2e\u65b9\u6cd5\u8207\u95dc\u9375\u5b57\u7684\u65b9\u6cd5\u76f8\u7d50\u5408\u5728\u76f8\u7576\u7a0b\ufa01\u4e0a\u6703 km:=0; FOR I=1 TO (\u8a13\uf996\u6587\u672c\u7684\uf969\u76ee) \u8868 1 \u4f7f\u7528\u8a72\u65b9\u6cd5\u7684\u516b\u500b\u7528\u6236\u7684\u53ec\u56de\uf961\u548c\u7cbe\u78ba\uf961 \u63d0\u9ad8\u904e\uf984\u7684\u6027\u80fd\uff0c\u9019\u5c07\u662f\u6211\u5011\u4e0b\u4e00\u6b65\u7814\u7a76\u7684\u76ee\u6a19\u3002</td></tr><tr><td colspan=\"3\">\u5c0d\u65bc\u7b2c I \u500b\u8a13\uf996\u6587\u672c\uff0c\u8a08\u7b97 k \u500b\u6700\u8fd1\u9130\u5c45\uff0c\u4e26\uf9dd\u7528 k \u500b\u9130\u5c45\u7684\uf9d0\u5225\u5224\u5b9a \u7b2c I \u500b\u6587\u672c\u7684\uf9d0\u5225\uff0c\u5982\u679c\u76f8\u7b49\uff0c\u5247 km:=km+1; \u6211\u5011\u53ef\u4ee5\u5f9e\u4ee5\u4e0b\u5e7e\u65b9\u9762\uf92d\u5206\u6790\u9019\u500b\u904e\uf984\u6a21\u578b\u7523\u751f\u8f03\u597d\u7d50\u679c\u7684\u539f\u56e0\uff1a 1. \u4f4e\u7dad\u5206\u6790\u7a7a\u9593\uff1a\u6240\u6709\u7684\u6982\uf9a3\u90fd\u88ab\u5206\u89e3\u6210\u7fa9\u539f\uff0c\u53ea\u9808\u5728\u53ef\u5206\u7fa9\u539f\u7a7a\u9593\u4e2d\u8a08\u7b97\u76f8\u4f3c\u7a0b \uf96b\u8003\u6587\u737b</td></tr><tr><td colspan=\"3\">ENDFOR \ufa01\uff0c\u9019\u6a23\u6211\u5011\u5c31\u53ea\u8981\u8a08\u7b97 600 \u500b\u5de6\u53f3\u7684\u53ef\u5206\u7fa9\u539f\u800c\uf967\u662f 100000 \u500b\u5de6\u53f3\u7684\u4e2d\u82f1\u6587 TRANSLIB. \"</td></tr></table>",
+                "num": null,
+                "html": null,
+                "text": "Advanced Tools for Accessing Multilingual Library Catalogues.\" Technical Report, Deleveralbe D.1.4:Evaluation of Tools.Knowledge S.A., June 1995."
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O02/O02-2003.json

@@ -0,0 +1,347 @@
+{
+    "paper_id": "O02-2003",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:05:50.838878Z"
+    },
+    "title": "Word Similarity Computing Based on How-net",
+    "authors": [
+        {
+            "first": "",
+            "middle": [],
+            "last": "\uf9c7\u7fa4",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "Word similarity is broadly used in many applications, such as information retrieval, information extraction, text classification, word sense disambiguation, example-based machine translation, etc. There are two different methods used to compute similarity: one is based on ontology or a semantic taxonomy; the other is based on collocations of words in a corpus.",
+    "pdf_parse": {
+        "paper_id": "O02-2003",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "Word similarity is broadly used in many applications, such as information retrieval, information extraction, text classification, word sense disambiguation, example-based machine translation, etc. There are two different methods used to compute similarity: one is based on ontology or a semantic taxonomy; the other is based on collocations of words in a corpus.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "As a lexical knowledgebase with rich semantic information, How-net has been employed in various researches. Unlike other thesauri, such as WordNet and Tongyici Cilin, in which word similarity is defined based on the distance between words in a semantic taxonomy tree, How-net defines a word in a complicated multi-dimensional knowledge description language. As a result, a series of problems arise in the process of word similarity computation using How-net. The difficulties are outlined below:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "1. The description of each word consists of a group of sememes. For example, the Chinese word \"\u6697\u7bb1(camera obscura)\uff02 is described as: \"part|\u90e8\u4ef6, #TakePicture|\u62cd\u651d, %tool|\u7528\u5177, body|\u8eab\uff02, and the Chinese word \"\u5beb\u4fe1 (write a letter)\uff02 is described as: \"write|\u5beb, ContentProduct=letter|\u4fe1\u4ef6\uff02;",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "2. The meaning of a word is not a simple combination of these sememes. Sememes are organized using a specific knowledge description language.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "To meet these challenges, our work includes:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "1. A study on the How-net knowledge description language. We rewrite the How-net definition of a word in a more structural format, using the abstract data structure of set and feature structure.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "2. A study on the algorithm used to compute word similarity based on How-net. The similarity between sememes, that between sets, and that between feature structures are given. To compute the similarity between two sememes, we use the distance between the sememes in the semantic taxonomy, as is done in Wordnet and Tongyici Cilin. To compute the similarity between two sets or two feature structures, we first establish a one-to-one mapping between the elements of the sets or the feature structures. Then, the similarity between the sets or feature structures is defined as the weighted average of the similarity between their elements. For feature structures, a one-to-one mapping is established according to the attributes. For sets, a one-to-one mapping is established according to the similarity between their elements.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "3. Finally, we give experiment results to show the validity of the algorithm and compare them with results obtained using other algorithms. Our results for word similarity agree with people's intuition to a large extent, and they are better than the results of two comparative experiments. [Dagan et al. 1995 [Dagan et al. ,1999 \u53ef\u4ee5\u770b\u5230\uff0c\u7d55\u5927\u90e8\u5206\u7d50\u679c\u9084\u662f\u6bd4\u8f03\u5408\uf9e4\u7684\uff0c\u4f46\u4e5f\u6709\u90e8\u5206\u7d50\u679c\uf967\u5920\u5408\uf9e4\uff0c\uf9b5\u5982\"\u4e2d\u570b\uff02 \u548c\"\uf997\u5408\u570b\uff02\u3001\"\u4e2d\u570b\uff02\u548c\"\u5b89\uf9e4\u6703\uff02\u7684\u76f8\u4f3c\ufa01\u90fd\u904e\u4f4e\uff0c\u9019\u662f\u56e0\u70ba\uff0c\"\u4e2d\u570b\uff02\u3001\"\uf997\u5408 \u570b\uff02\u3001\"\u5b89\uf9e4\u6703\uff02\u5728\u300a\u77e5\u7db2\u300b\u4e2d\u7684\u7b2c\u4e00\u57fa\u672c\u7fa9\u539f\u5206\u5225\u662f\"\u5730\u65b9\uff02\u3001\"\u6a5f\u69cb\uff02\u3001\"\u90e8\u4ef6\uff02\u3002 \"\u8dd1\uff02\u548c\"\u8df3\uff02\u7684\u76f8\u4f3c\ufa01\u4e5f\u8f03\u4f4e\uff0c\u9019\u662f\u56e0\u70ba\u9019\uf978\u500b\u8a5e\u88ab\u7c21\u55ae\u5b9a\u7fa9\u70ba\uf978\u500b\u57fa\u672c\u7fa9\u539f\uff0c\u800c\u7f3a \u5c11\u5176\u4ed6\u8cc7\u8a0a\u3002\u9019\u4e5f\u5f9e\u4e00\u500b\u5074\u9762\u53cd\u6620\uf9ba\u77e5\u7db2\u7684\u67d0\u4e9b\u5b9a\u7fa9\uf967\u5408\uf9e4\u6216\uf967\u4e00\u81f4\u4e4b\u8655\u3002 ",
+                "cite_spans": [
+                    {
+                        "start": 290,
+                        "end": 308,
+                        "text": "[Dagan et al. 1995",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 309,
+                        "end": 328,
+                        "text": "[Dagan et al. ,1999",
+                        "ref_id": "BIBREF2"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u03b1 \u03b1 + = ) , ( ) , ( 2 1 2 1 W W Dis W W Sim \u2026\u2026(1) \u5716 1 \u300a\u540c\u7fa9\u8a5e\u8a5e\uf9f4\u300b\u8a9e\u7fa9\u5206\uf9d0\u6a39\uf9fa\u5716 [\u738b\u658c\uff0c1999]\u63a1\u7528\u9019\u7a2e\u65b9\u6cd5\uf9dd\u7528\u300a\u540c\u7fa9\u8a5e\u8a5e\uf9f4\u300b\uf92d\u8a08\u7b97\u6f22\u8a9e\u8a5e\u8a9e\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01(\u5982 \u5716 1 \u6240\u793a)\u3002\u6709\u4e9b\u7814\u7a76\u8005\u8003\u616e\u7684\u60c5\u6cc1\uf901\u8907\u96dc\u3002[Agirre & Rigau 1995]\u5728\uf9dd\u7528 Wordnet \u8a08\u7b97 \u8a5e\u8a9e\u7684\u8a9e\u7fa9\u76f8\u4f3c\ufa01\u6642\uff0c\u9664\uf9ba\u7d50\u9ede\u9593\u7684\uf937\u5f91\u9577\ufa01\u5916\uff0c\u9084\u8003\u616e\u5230\uf9ba\u5176\u4ed6\u4e00\u4e9b\u56e0\u7d20\u3002\uf9b5\u5982\uff1a \u6982\uf9a3\u5c64\u6b21\u6a39\u7684\u6df1\ufa01\uff1a\uf937\u5f91\u9577\ufa01\u76f8\u540c\u7684\uf978\u500b\u7d50\u9ede\uff0c\u5982\u679c\u4f4d\u65bc\u6982\uf9a3\u5c64\u6b21\u7684\u8d8a\u9ad8\u5c64\uff0c\u5176\u8a9e \u7fa9\u8ddd\uf9ea\u8f03\u5927\uff1b\u6bd4\u5982\uf96f\uff1a\"\u52d5\u7269\uff02\u548c\"\u690d\u7269\uff02\u3001\"\u54fa\u4e73\u52d5\u7269\uff02\u548c\"\u722c\ufa08\u52d5\u7269\uff02\uff0c\u9019\uf978\u5c0d\u6982 \uf9a3\u9593\u7684\uf937\u5f91\u9577\ufa01\u90fd\u662f 2\uff0c\u4f46\u524d\u4e00\u5c0d\u8a5e\u8655\u65bc\u8a9e\u7fa9\u6a39\u7684\u8f03\u9ad8\u5c64\uff0c\u56e0\u6b64\u8a8d\u70ba\u5176\u8a9e\u7fa9\u8ddd\uf9ea\u8f03\u5927\uff0c \u5f8c\u4e00\u5c0d\u8a5e\u8655\u65bc\u8a9e\u7fa9\u6a39\u7684\u8f03\u4f4e\u5c64\uff0c\u5176\u8a9e\u7fa9\u8ddd\uf9ea\u8f03\u5c0f\uff1b \u6982\uf9a3\u5c64\u6b21\u6a39\u7684\u5340\u57df\u5bc6\ufa01\uff1a\uf937\u5f91\u9577\ufa01\u76f8\u540c\u7684\uf978\u5c0d\u7d50\u9ede\uff0c\u5982\u679c\u4e00\u5c0d\u4f4d\u65bc\u6982\uf9a3\u5c64\u6b21\u6a39\u4e2d\u4f4e \u5bc6\ufa01\u5340\u57df\uff0c\u53e6\u4e00\u5c0d\u4f4d\u5143\u65bc\u9ad8\u5bc6\ufa01\u5340\u57df\uff0c\u90a3\u9ebc\u524d\u8005\u7684\u8a9e\u7fa9\u8ddd\uf9ea\u61c9\u5927\u65bc\u5f8c\u8005\u3002\u5f15\u5165\u5340\u57df\u5bc6\ufa01 \u7684\u539f\u56e0\u5728\u65bc\uff0c\u6709\u4e9b\u6982\uf9a3\u5c64\u6b21\u6a39\u4e2d\u6982\uf9a3\u63cf\u8ff0\u7684\u7c97\u7d30\u7a0b\ufa01\uf967\u5747\uff0c\uf9b5\u5982\u5728 Wordnet \u4e2d\uff0c\u52d5\u690d\u7269 \u5206\uf9d0\u7684\u63cf\u8ff0\u6975\u5176\u8a73\u76e1\uff0c\u800c\u6709\u4e9b\u5340\u57df\u7684\u6982\uf9a3\u63cf\u8ff0\u53c8\u6bd4\u8f03\u7c97\u758f\uff0c\u9019\u6703\u5c0e\u81f4\u8a9e\u7fa9\u8ddd\uf9ea\u8a08\u7b97\u7684\uf967 \u5408\uf9e4\u3002 \u53e6\u4e00\u7a2e\u8a5e\u8a9e\u76f8\u4f3c\ufa01\u7684\u8a08\u7b97\u65b9\u6cd5\u662f\u7528\u5927\u898f\u6a21\u7684\u8a9e\uf9be\uf92d\u7d71\u8a08\u3002\uf9b5\u5982\uff0c\uf9dd\u7528\u8a5e\u8a9e\u7684\u76f8\u95dc\u6027 \uf92d\u8a08\u7b97\u8a5e\u8a9e\u7684\u76f8\u4f3c\ufa01\u3002\u4e8b\u5148\u9078\u64c7\u4e00\u7d44\u7279\u5fb5\u8a5e\uff0c\u7136\u5f8c\u8a08\u7b97\u9019\u4e00\u7d44\u7279\u5fb5\u8a5e\u8207\u6bcf\u4e00\u500b\u8a5e\u7684\u76f8\u95dc \u6027(\u4e00\u822c\u7528\u9019\u7d44\u7279\u5fb5\u8a5e\u5728\u5be6\u969b\u7684\u5927\u898f\u6a21\u8a9e\uf9be\u4e2d\u5728\u8a72\u8a5e\u7684\u4e0a\u4e0b\u6587\u4e2d\u51fa\u73fe\u7684\u983b\uf961\uf92d\ufa01\uf97e)\uff0c \u65bc\u662f\uff0c\u5c0d\u65bc\u6bcf\u4e00\u500b\u8a5e\u90fd\u53ef\u4ee5\u5f97\u5230\u4e00\u500b\u76f8\u95dc\u6027\u7684\u7279\u5fb5\u8a5e\u5411\uf97e\uff0c\u7136\u5f8c\uf9dd\u7528\u9019\u4e9b\u5411\uf97e\u4e4b\u9593\u7684\u76f8 \u4f3c\ufa01(\u4e00\u822c\u7528\u5411\uf97e\u7684\u593e\u89d2\u9918\u5f26\uf92d\u8a08\u7b97)\u4f5c\u70ba\u9019\uf978\u500b\u8a5e\u7684\u76f8\u4f3c\ufa01\u3002\u9019\u7a2e\u505a\u6cd5\u7684\u5047\u8a2d\u662f\uff0c\u51e1 \u662f\u8a9e\u7fa9\u76f8\u8fd1\u7684\u8a5e\uff0c\u4ed6\u5011\u7684\u4e0a\u4e0b\u6587\u4e5f\u61c9\u8a72\u76f8\u4f3c\u3002[\uf9e1\u6d93\u5b50\uff0c1999]\uf9dd\u7528\u9019\u7a2e\u601d\u60f3\uf92d\u5be6\u73fe\u8a9e\u7fa9 \u7684\u81ea\u52d5\u6392\u6b67\uff1b[\uf939\u677e\uff0c2001]\u7814\u7a76\uf9ba\u5982\u4f55\uf9dd\u7528\u8a5e\u8a9e\u7684\u76f8\u95dc\u6027\uf92d\u8a08\u7b97\u8a5e\u8a9e\u7684\u76f8\u4f3c\ufa01\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u23a5 \u23a5 \u23a5 \u23a5 \u23a5 \u23a5 \u23a5 \u23a5 \u23a5 \u23a5 \u23a5 \u23a5 \u23a6 \u23a4 \u23a2 \u23a2 \u23a2 \u23a2 \u23a2 \u23a2 \u23a2 \u23a2 \u23a2 \u23a2 \u23a2 \u23a2 \u23a3 \u23a1 \u23a5 \u23a5 \u23a5 \u23a6 \u23a4 \u23a2 \u23a2 \u23a2 \u23a3 \u23a1 \u2026 \u2026 \u2026 \u2026 \u2026 \u2026 \u23a5 \u23a5 \u23a5 \u23a6 \u23a4 \u23a2 \u23a2 \u23a2 \u23a3 \u23a1 \u2026 \u2026 \u2026 \u2026 \uff5d \uff0c",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(S 1 ,S 2 )\uff1b \u5176\u4ed6\u57fa\u672c\u7fa9\u539f\u63cf\u8ff0\uff1a\u5c0d\u61c9\u65bc\u8a9e\u7fa9\u904b\u7b97\u5f0f\u4e2d\u9664\u7b2c\u4e00\u57fa\u672c\u7fa9\u539f\u63cf\u8ff0\u5f0f\u4ee5\u5916\u7684\u6240\u6709\u57fa\u672c\u7fa9 \u539f\u63cf\u8ff0\u5f0f\uff0c\u5176\u503c\u70ba\u4e00\u500b\u57fa\u672c\u7fa9\u539f\u7684\u96c6\u5408\uff0c\u6211\u5011\u5c07\uf978\u500b\u6982\uf9a3\u7684\u9019\u4e00\u90e8\u5206\u7684\u76f8\u4f3c\ufa01\u8a18\u70ba Sim 2 (S 1 ,S 2 )\uff1b \u95dc\u4fc2\u7fa9\u539f\u63cf\u8ff0\uff1a\u5c0d\u61c9\u65bc\u8a9e\u7fa9\u904b\u7b97\u5f0f\u4e2d\u6240\u6709\u7684\u95dc\u4fc2\u7fa9\u539f\u63cf\u8ff0\u5f0f\uff0c\u5176\u503c\u662f\u4e00\u500b\u7279\u5fb5\u7d50\u69cb\uff0c \u5c0d\u65bc\u8a72\u7279\u5fb5\u7d50\u69cb\u7684\u6bcf\u4e00\u500b\u7279\u5fb5\uff0c\u5176\u5c6c\u6027\u662f\u4e00\u500b\u95dc\u4fc2\u7fa9\u539f\uff0c\u5176\u503c\u662f\u4e00\u500b\u57fa\u672c\u7fa9\u539f\uff0c\u6216\u4e00\u500b \u5177\u9ad4\u8a5e\u3002\u6211\u5011\u5c07\uf978\u500b\u6982\uf9a3\u7684\u9019\u4e00\u90e8\u5206\u7684\u76f8\u4f3c\ufa01\u8a18\u70ba Sim 3 (S 1 ,S 2 )\uff1b \u95dc\u4fc2\u7b26\u865f\u63cf\u8ff0\uff1a\u5c0d\u61c9\u65bc\u8a9e\u7fa9\u904b\u7b97\u5f0f\u4e2d\u6240\u6709\u7684\u95dc\u4fc2\u7b26\u865f\u63cf\u8ff0\u5f0f\uff0c\u5176\u503c\u4e5f\u662f\u4e00\u500b\u7279\u5fb5\u7d50 \u69cb\uff0c\u5c0d\u65bc\u8a72\u7279\u5fb5\u7d50\u69cb\u7684\u6bcf\u4e00\u500b\u7279\u5fb5\uff0c\u5176\u5c6c\u6027\u662f\u4e00\u500b\u95dc\u4fc2\u7fa9\u539f\uff0c\u5176\u503c\u662f\u4e00\u500b\u96c6\u5408\uff0c\u8a72\u96c6\u5408 \u7684\u5143\u7d20\u662f\u4e00\u500b\u57fa\u672c\u7fa9\u539f\uff0c\u6216\u4e00\u500b\u5177\u9ad4\u8a5e\u3002\u6211\u5011\u5c07\uf978\u500b\u6982\uf9a3\u7684\u9019\u4e00\u90e8\u5206\u7684\u76f8\u4f3c\ufa01\u8a18\u70ba Sim 4 (S 1 ,S 2 )\u3002 \u65bc\u662f\uff0c\uf978\u500b\u6982\uf9a3\u8a9e\u7fa9\u904b\u7b97\u5f0f\u7684\u6574\u9ad4\u76f8\u4f3c\ufa01\u8a18\u70ba\uff1a \u5176\u4e2d\uff0c\u03b2 i (1\u2264i\u22644)\u662f\u53ef\u8abf\u7bc0\u7684\uf96b\uf969\uff0c\u4e14\u6709\uff1a \u03b2 1 +\u03b2 2 +\u03b2 3 +\u03b2 4 =1, \u03b2 1\u2265 \u03b2 2\u2265 \u03b2 3\u2265 \u03b2 4 \u5f8c\u8005\u53cd\u6620\uf9ba Sim 1 \u5230 Sim 4 \u5c0d\u65bc\u7e3d\u9ad4\u76f8\u4f3c\ufa01\u6240\u8d77\u5230\u7684\u4f5c\u7528\u4f9d\u6b21\u905e\u6e1b\u3002\u7531\u65bc\u7b2c\u4e00\u57fa\u672c\u7fa9\u539f \u63cf\u8ff0\u5f0f\u53cd\u6620\uf9ba\u4e00\u500b\u6982\uf9a3\u6700\u4e3b\u8981\u7684\u7279\u5fb5\uff0c\u6240\u4ee5\u6211\u5011\u61c9\u8a72\u5c07\u5176\u6b0a\u503c\u5b9a\u7fa9\u5f97\u6bd4\u8f03\u5927\uff0c\u4e00\u822c\u61c9\u5728 0.5",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u2211 = = 4 1 2 1 2 1 ) , ( ) , ( i i i S S Sim S S Sim \u03b2 \u2026\u2026(4) \u2211 \u220f = = = 4 1 1 2 1 2 1 ) , ( ) , ( i i j j i S S Sim S S",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "A proposal for word sense disambiguation using conceptual distance",
+                "authors": [
+                    {
+                        "first": "E",
+                        "middle": [],
+                        "last": "Agirre",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [],
+                        "last": "Rigau",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Proc. of International Conference Recent Advances in Natural Language Processing",
+                "volume": "",
+                "issue": "",
+                "pages": "258--264",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Agirre E. and Rigau G., \"A proposal for word sense disambiguation using conceptual distance\", Proc. of International Conference Recent Advances in Natural Language Processing (RANLP) , 1995, pp. 258-264, Tzigov Chark, Bulgaria.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Contextual Word Similarity and Estimation from Sparse Data",
+                "authors": [
+                    {
+                        "first": "I",
+                        "middle": [],
+                        "last": "Dagan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Marcus",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL)",
+                "volume": "",
+                "issue": "",
+                "pages": "164--171",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Dagan I., Marcus S., et al. , \"Contextual Word Similarity and Estimation from Sparse Data\", in Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL) , 1993, pp. 164-171",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Similarity-based models of word cooccurrence probabilities",
+                "authors": [
+                    {
+                        "first": "I",
+                        "middle": [],
+                        "last": "Dagan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "L",
+                        "middle": [],
+                        "last": "Lee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Pereira",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Machine Learning, Special issue on Machine Learning and Natural Language",
+                "volume": "34",
+                "issue": "",
+                "pages": "43--69",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Dagan I., Lee L. and Pereira F., \"Similarity-based models of word cooccurrence probabilities\", Machine Learning, Special issue on Machine Learning and Natural Language, 34(1-3) , 1999, pp. 43-69",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Automatic Word Similarity Detection for TREC 4 Query Expansion",
+                "authors": [
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Gauch",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Chong",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Proc. of TREC-4: The 4th Annual Text REtrieval Conf",
+                "volume": "",
+                "issue": "",
+                "pages": "527--536",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Gauch S. and Chong M. K., \"Automatic Word Similarity Detection for TREC 4 Query Expansion\", Proc. of TREC-4: The 4th Annual Text REtrieval Conf., Nov. 1995, Gaithersburg, MD, 1995, pp. 527-536",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Semantic Computation in Chinese Question-Answering System",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Li Sujian",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Huang",
+                        "middle": [],
+                        "last": "Zhang Jian",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Bai",
+                        "middle": [],
+                        "last": "Xiong",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Shuo",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2002,
+                "venue": "Journal of Computer Science and Technology",
+                "volume": "17",
+                "issue": "6",
+                "pages": "993--999",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "LI Sujian, ZHANG Jian, HUANG Xiong and BAI Shuo, \"Semantic Computation in Chinese Question-Answering System\", Journal of Computer Science and Technology 17(6) , 2002, pp. 993-999",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "A WordNet-based algorithm for word sense disambiguation",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Li Xiaobin",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S",
+                        "middle": [],
+                        "last": "Szpakowicz",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Matwin",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Proc. of the Twelth International Joint Conference on Artificial Intelligence (IJCAI)",
+                "volume": "",
+                "issue": "",
+                "pages": "1368--1374",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "LI Xiaobin, Szpakowicz S., and Matwin S., \"A WordNet-based algorithm for word sense disambiguation\", Proc. of the Twelth International Joint Conference on Artificial Intelligence (IJCAI). 1995, pp. 1368-1374",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF2": {
+                "html": null,
+                "num": null,
+                "content": "<table><tr><td>\u6211\u5011\u8a8d\u70ba\uff0c\u5728\u5be6\u969b\u7684\u6587\u672c\u4e2d\uff0c\u865b\u8a5e\u548c\u5be6\u8a5e\u7e3d\u662f\uf967\u80fd\u4e92\u76f8\u66ff\u63db\u7684\uff0c\u56e0\u6b64\uff0c\u865b\u8a5e\u6982\uf9a3\u548c\u5be6\u8a5e \u6982\uf9a3\u7684\u76f8\u4f3c\ufa01\u7e3d\u662f\u70ba\uf9b2\u3002 \u7531\u65bc\u865b\u8a5e\u6982\uf9a3\u7e3d\u662f\u7528\"{\uf906\u6cd5\u7fa9\u539f}\uff02\u6216\"{\u95dc\u4fc2\u7fa9\u539f}\uff02\u9019\uf978\u7a2e\u65b9\u5f0f\u9032\ufa08\u63cf\u8ff0\uff0c\u6240 \u9019\u6a23\uff0c\u7279\u5fb5\u7d50\u69cb\u7684\u76f8\u4f3c\ufa01\u5c31\u8f49\u5316\u70ba\u5404\u500b\u7279\u5fb5\u7684\u76f8\u4f3c\ufa01\u7684\u52a0\u6b0a\u5e73\u5747\u3002\u5176\u4e2d\u7684\u6b0a\u503c\u53cd\u6620 \u51fa\u8a72\u5c6c\u6027\u5728\u7279\u5fb5\u7d50\u69cb\u4e2d\u7684\u91cd\u8981\u7a0b\ufa01\u3002\u5728\u76ee\u524d\u6211\u5011\u8a8d\u70ba\u6240\u6709\u7279\u5fb5\u5177\u6709\u76f8\u540c\u7684\u91cd\u8981\u6027\u3002 \u5269\u4e0b\u7684\u554f\u984c\u5c31\u662f\u8a08\u7b97\uf978\u500b\u7279\u5fb5\u7684\u76f8\u4f3c\ufa01\u3002\u7279\u5fb5\u7531\"\u5c6c\u6027\uff02\u548c\"\u503c\uff02\u7d44\u6210\u3002\u7531\u65bc\"\u5c6c 4.1 \u7684\u76f8\u4f3c\ufa01\u4e4b\u6700\u5927\u503c\uff0c\u4e5f\u5c31\u662f\uf96f\uff1a \u4ee5\uff0c\u865b\u8a5e\u6982\uf9a3\u7684\u76f8\u4f3c\ufa01\u8a08\u7b97\u975e\u5e38\u7c21\u55ae\uff0c\u53ea\u9700\u8981\u8a08\u7b97\u5176\u5c0d\u61c9\u7684\uf906\u6cd5\u7fa9\u539f\u6216\u95dc\u4fc2\u7fa9\u539f\u4e4b\u9593\u7684 \u6027\uff02\u76f8\u540c\uff0c\u65bc\u662f\uff0c\uf978\u500b\u7279\u5fb5\u7684\u76f8\u4f3c\ufa01\u53ef\u4ee5\u7b49\u50f9\u65bc\u5176\"\u503c\uff02\u7684\u76f8\u4f3c\ufa01\u3002</td></tr><tr><td>\u5177\u9ad4\u8a5e \uff0c \u57fa\u672c\u7fa9\u539f \uff1d\u57fa\u672c\u7fa9\u539f \u57fa\u672c\u7fa9\u539f \u95dc\u4fc2\u7fa9\u539f \u5176\u4ed6\u57fa\u672c\u7fa9\u539f\u63cf\u8ff0\uff1d\uff5b \u672c\u7fa9\u539f \u7b2c\u4e00\u57fa\u672c\u7fa9\u539f\u63cf\u8ff0\uff1d\u57fa x 1 c b a | , \u9019\u6a23\uff0c\u6211\u5011\u5c31\u628a\uf978\u500b\u8a5e\u8a9e\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\u554f\u984c\u6b78\u7d50\u5230\uf9ba\uf978\u500b\u6982\uf9a3\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\u554f\u984c\u3002 \uff5d x ) , ( max ) , ( 2 1 ... 1 , .. 1 2 1 j i m j n i S S Sim W W Sim = = \u2026\u2026(2) \u76f8\u4f3c\ufa01\u5373\u53ef\u3002 4.4.2 \u96c6\u5408\u7684\u76f8\u4f3c\ufa01\u8a08\u7b97 = \u96c6\u5408\u7684\u76f8\u4f3c\ufa01\u8a08\u7b97\u6bd4\u7279\u5fb5\u7d50\u69cb\uf901\u70ba\u8907\u96dc\uff0c\u56e0\u70ba\u96c6\u5408\u7684\u5143\u7d20\u662f\u7121\u5e8f\u800c\u4e14\u5e73\u7b49\u7684\uff0c\u56e0\u6b64\u9996\u8981 4.4 \u5be6\u8a5e\u6982\uf9a3\u7684\u76f8\u4f3c\ufa01\u7684\u8a08\u7b97 \u4efb\u52d9\u662f\u8981\u5728\uf978\u500b\u96c6\u5408\u7684\u5143\u7d20\u4e4b\u9593\u5efa\uf9f7\u4e00\u4e00\u5c0d\u61c9\u95dc\u4fc2\u3002</td></tr><tr><td>\u5177\u9ad4\u8a5e \u7576\u7136\uff0c\u6211\u5011\u9019\u88cf\u8003\u616e\u7684\u662f\u5b64\uf9f7\u7684\uf978\u500b\u8a5e\u8a9e\u7684\u76f8\u4f3c\ufa01\u3002\u5982\u679c\u662f\u5728\u4e00\u5b9a\u4e0a\u4e0b\u6587\u4e4b\u4e2d\u7684\uf978\u500b\u8a5e \uff1d\u57fa\u672c\u7fa9\u539f \u95dc\u4fc2\u7fa9\u539f \u95dc\u4fc2\u7fa9\u539f\u63cf\u8ff0\uff1d \uff1a \u6982\uf9a3 \u5be6\u8a5e y y 2 | \u8a9e\uff0c\u6700\u597d\u662f\u5148\u9032\ufa08\u8a5e\u7fa9\u6392\u6b67\uff0c\u5c07\u8a5e\u8a9e\u6a19\u6ce8\u70ba\u6982\uf9a3\uff0c\u7136\u5f8c\u518d\u5c0d\u6982\uf9a3\u8a08\u7b97\u76f8\u4f3c\ufa01\u3002 \u5f9e\u524d\u9762\u7684\u5206\u6790\u53ef\u77e5\uff0c\u300a\u77e5\u7db2\u300b\u7684\u77e5\uf9fc\u63cf\u8ff0\u8a9e\u8a00\u53ef\u4ee5\u901a\u904e\u7fa9\u539f\u548c\u96c6\u5408\u3001\u7279\u5fb5\u7d50\u69cb\u9019\uf978\u7a2e\u62bd \uf978\u500b\u96c6\u5408\u7684\u76f8\u4f3c\ufa01\u8a08\u7b97\u6a21\u578b\uff0c\u5fc5\u9808\u6eff\u8db3\u6211\u5011\u5c0d\u65bc\u96c6\u5408\u76f8\u4f3c\ufa01\u8a08\u7b97\u7684\u4e00\u4e9b\u76f4\u89c0\u8981\u6c42\u3002 \u8c61\u8cc7\uf9be\u7d50\u69cb\uf92d\u8868\u9054\u3002\u7fa9\u539f\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\u8a08\u7b97\u554f\u984c\u5df2\u7d93\u89e3\u6c7a\uff0c\u5269\u4e0b\u7684\u554f\u984c\u5c31\u662f\u96c6\u5408\u548c\u7279\u5fb5 \u7d50\u69cb\u7684\u76f8\u4f3c\ufa01\u554f\u984c\uf9ba\u3002 \u9019\u88cf\u6211\u5011\uf99c\u51fa\u4ee5\u4e0b\uf978\u689d\uff1a</td></tr><tr><td>\u5177\u4f53\u8a5e \u5177\u4f53\u8a5e t | \u6211\u5011\u7684\u57fa\u672c\u8a2d\u60f3\u662f\uff1a\u6574\u9ad4\u76f8\u4f3c\u8981\u5efa\uf9f7\u5728\u90e8\u5206\u76f8\u4f3c\u7684\u57fa\u790e\u4e0a\u3002\u628a\u4e00\u500b\u8907\u96dc\u7684\u6574\u9ad4\u5206\u89e3 \u7fa9\u539f \u5177\u4f53\u8a5e \uff1d\uff5b\u7fa9\u539f \u95dc\u4fc2\u7b26\u865f \uff5d \uff0c \u7fa9\u539f \u5177\u4f53\u8a5e \uff1d\uff5b\u7fa9\u539f \u95dc\u4fc2\u7b26\u865f \u95dc\u4fc2\u7b26\u865f\u63cf\u8ff0\uff1d t s s 2 v v u u 1 | , | , | 1. \u4e00\u500b\u96c6\u5408\u548c\u5b83\u672c\u8eab\u7684\u76f8\u4f3c\ufa01\u70ba 1\uff1b 4.2 \u7fa9\u539f\u76f8\u4f3c\ufa01\u8a08\u7b97 \u6210\u90e8\u5206\uff0c\u901a\u904e\u8a08\u7b97\u90e8\u5206\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\u5f97\u5230\u6574\u9ad4\u7684\u76f8\u4f3c\ufa01\u3002 2. \u5047\u8a2d\uf978\u500b\u96c6\u5408\u90fd\u6709 n \u500b\u5143\u7d20\uff0c\u5176\u4e2d m(m&lt;n)\u500b\u5143\u7d20\u76f8\u540c\uff0c\u53c8\u5047\u8a2d\uf978\u500b\u5143\u7d20\u7684\u76f8 \u7531\u65bc\u6240\u6709\u7684\u6982\uf9a3\u90fd\u6700\u7d42\u6b78\u7d50\u4e8e\u7528\u7fa9\u539f(\u500b\u5225\u5730\u65b9\u7528\u5177\u9ad4\u8a5e)\uf92d\u8868\u793a\uff0c\u6240\u4ee5\u7fa9\u539f\u7684\u76f8\u4f3c\ufa01 \u8a08\u7b97\u662f\u6982\uf9a3\u76f8\u4f3c\ufa01\u8a08\u7b97\u7684\u57fa\u790e\u3002 \u7531\u65bc\u6240\u6709\u7684\u7fa9\u539f\u6839\u64da\u4e0a\u4e0b\u4f4d\u95dc\u4fc2\u69cb\u6210\uf9ba\u4e00\u500b\u6a39\uf9fa\u7684\u7fa9\u539f\u5c64\u6b21\u9ad4\u7cfb\uff0c\u6211\u5011\u9019\u88cf\u63a1\u7528\u7c21 \u4f3c\ufa01\u53ea\u80fd\u662f 0(\uf967\u540c)\u6216 1(\u76f8\u540c)\uff0c\u90a3\u9ebc\u9019\uf978\u500b\u96c6\u5408\u7684\u76f8\u4f3c\ufa01\u61c9\u8a72\u662f m/n\u3002 \u8981\u8a08\u7b97\uf978\u500b\u96c6\u5408\u7684\u76f8\u4f3c\ufa01\uff0c\u6700\u5bb9\uf9e0\u60f3\u5230\u7684\u65b9\u6cd5\u662f\u9996\u5148\u8a08\u7b97\uf978\u500b\u96c6\u5408\u7684\u6240\u6709\u5143\u7d20\uf978\uf978 \u5047\u8a2d\uf978\u500b\u6574\u9ad4 A \u548c B \u5206\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\u662f\u5426\u90fd\u5c0d\u6574\u9ad4\u7684\u76f8\u4f3c\ufa01\u767c\u751f\u5f71\u97ff\uff1f\u5982\u679c\uf967\u662f\u5168\u90e8\u90fd\u767c\u751f\u5f71\u97ff\uff0c\u90a3\u9ebc\u6211\u5011 \u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\uff0c\u7136\u5f8c\u518d\u9032\ufa08\u52a0\u6b0a\u5e73\u5747\u3002\u4f46\u662f\u9019\u6a23\u6703\u5e36\uf92d\u4e00\u500b\u554f\u984c\uff0c\u5c31\u662f\u4e00\u500b\u96c6\u5408\u548c\u5b83\u672c \u55ae\u7684\u901a\u904e\u8a9e\u7fa9\u8ddd\uf9ea\u8a08\u7b97\u76f8\u4f3c\ufa01\u7684\u8fa6\u6cd5\u3002\u5047\u8a2d\uf978\u500b\u7fa9\u539f\u5728\u9019\u500b\u5c64\u6b21\u9ad4\u7cfb\u4e2d\u7684\uf937\u5f91\u8ddd\uf9ea\u70ba \u61c9\u8a72\u5982\u4f55\u9078\u64c7\u767c\u751f\u5f71\u97ff\u7684\u90a3\u4e9b\u90e8\u5206\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\uff1f\u9078\u64c7\u51fa\uf92d\u4ee5\u5f8c\uff0c\u6211\u5011\u53c8\u5982\u4f55\u5f97\u5230\u6574\u9ad4 \u8eab\u7684\u76f8\u4f3c\ufa01\u53ef\u80fd\uf967\u70ba 1\uff0c\u9664\u975e\u5b83\u7684\u4efb\u610f\uf978\u500b\u5143\u7d20\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\u90fd\u70ba 1\u3002\u9019\u500b\u7d50\u679c\u7576\u7136\u662f\uf967 d\uff0c\u6839\u64da\u516c\u5f0f(1)\uff0c\u6211\u5011\u53ef\u4ee5\u5f97\u5230\u9019\uf978\u500b\u7fa9\u539f\u4e4b\u9593\u7684\u8a9e\u7fa9\u8ddd\uf9ea\uff1a \u03b1 \u03b1 + = p p Sim ) , ( 2 1 \u5408\uf9e4\u7684\u3002\u9019\u4e5f\u5f9e\u53e6\u4e00\u500b\u89d2\ufa01\uf96f\u660e\u6211\u5011\u5148\u524d\u5b9a\u7fa9\u7684\u539f\u5247(\u9996\u5148\u5728\uf978\u500b\u96c6\u5408\u7684\u5143\u7d20\u4e4b\u9593\u5efa\uf9f7 \u7684\u76f8\u4f3c\ufa01\uff1f \u2026\u2026(3) \u4e00\u4e00\u5c0d\u61c9\u95dc\u4fc2)\u7684\u5408\uf9e4\u6027\u3002 \u6211\u5011\u8a8d\u70ba\uff1a\u4e00\u500b\u6574\u9ad4\u7684\u5404\u500b\uf967\u540c\u90e8\u5206\u5728\u6574\u9ad4\u4e2d\u7684\u4f5c\u7528\u662f\uf967\u540c\u7684\uff0c\u53ea\u6709\u5728\u6574\u9ad4\u4e2d\u8d77\u76f8 d \u5728\u672c\u6587\u4e2d\uff0c\u6211\u5011\u63a1\u7528\u4ee5\u4e0b\u6f14\u7b97\u6cd5\uf92d\u70ba\uf978\u500b\u96c6\u5408\u7684\u5143\u7d20\u4e4b\u9593\u5efa\uf9f7\u4e00\u4e00\u5c0d\u61c9\u95dc\u4fc2\uff1a \u540c\u4f5c\u7528\u7684\u90e8\u5206\u4e92\u76f8\u6bd4\u8f03\u624d\u6709\u6548\u3002\uf9b5\u5982\u6bd4\u8f03\uf978\u500b\u4eba\u9577\u76f8\u662f\u5426\u76f8\u4f3c\uff0c\u6211\u5011\u7e3d\u662f\u6bd4\u8f03\u5b83\u5011\u7684\u81c9 \u5176\u4e2d p 1 \u548c p 2 \u8868\u793a\uf978\u500b\u7fa9\u539f (primitive) \uff0cd \u662f p 1 \u548c p 2 \u5728\u7fa9\u539f\u5c64\u6b21\u9ad4\u7cfb\u4e2d\u7684\uf937\u5f91\u9577\ufa01\uff0c \u662f\u4e00\u500b\u6b63\u6574\uf969\u3002\u03b1\u662f\u4e00\u500b\u53ef\u8abf\u7bc0\u7684\uf96b\uf969\u3002 \u578b\u3001\uf9d7\ufa0b\u3001\u773c\u775b\u3001\u9f3b\u5b50\u7b49\u76f8\u540c\u90e8\u5206\u662f\u5426\u76f8\u4f3c\uff0c\u800c\uf967\u6703\u62ff\u773c\u775b\u53bb\u548c\u9f3b\u5b50\u505a\u6bd4\u8f03\u3002 1. \u9996\u5148\u8a08\u7b97\uf978\u500b\u96c6\u5408\u7684\u6240\u6709\u5143\u7d20\uf978\uf978\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\uff1b</td></tr><tr><td>\u56e0\u6b64\uff0c\u5728\u6bd4\u8f03\uf978\u500b\u6574\u9ad4\u7684\u76f8\u4f3c\u6027\u6642\uff0c\u6211\u5011\u9996\u5148\u8981\u505a\u7684\u5de5\u4f5c\u662f\u5c0d\u9019\uf978\u500b\u6574\u9ad4\u7684\u5404\u500b\u90e8 2. \u5f9e\u6240\u6709\u7684\u76f8\u4f3c\ufa01\u503c\u4e2d\u9078\u64c7\u6700\u5927\u7684\u4e00\u500b\uff0c\u5c07\u9019\u500b\u76f8\u4f3c\ufa01\u503c\u5c0d\u61c9\u7684\uf978\u500b\u5143\u7d20\u5c0d\u61c9\u8d77 \u7528\u9019\u7a2e\u65b9\u6cd5\u8a08\u7b97\u7fa9\u539f\u76f8\u4f3c\ufa01\u7684\u6642\u5019\uff0c\u6211\u5011\u53ea\uf9dd\u7528\uf9ba\u7fa9\u539f\u7684\u4e0a\u4e0b\u4f4d\u95dc\u4fc2\u3002\u5be6\u969b\u4e0a\uff0c\u5728 \u300a\u77e5\u7db2\u300b\u4e2d\uff0c\u7fa9\u539f\u4e4b\u9593\u9664\uf9ba\u4e0a\u4e0b\u4f4d\u95dc\u4fc2\u5916\uff0c\u9084\u6709\u5f88\u591a\u7a2e\u5176\u4ed6\u7684\u95dc\u4fc2\uff0c\u5982\u679c\u5728\u8a08\u7b97\u6642\u8003\u616e \u5206\u4e4b\u9593\u5efa\uf9f7\u8d77\u4e00\u4e00\u5c0d\u61c9\u7684\u95dc\u4fc2\uff0c\u7136\u5f8c\u5728\u9019\u4e9b\u5c0d\u61c9\u7684\u90e8\u5206\u4e4b\u9593\u9032\ufa08\u6bd4\u8f03\u3002 \uf92d\uff1b</td></tr><tr><td>\u9032\uf92d\uff0c\u53ef\u80fd\u6703\u5f97\u5230\uf901\u7cbe\u7d30\u7684\u7fa9\u539f\u76f8\u4f3c\ufa01\ufa01\uf97e\uff0c\uf9b5\u5982\uff0c\u6211\u5011\u53ef\u4ee5\u8a8d\u70ba\uff0c\u5177\u6709\u53cd\u7fa9\u6216\u8005\u5c0d\u7fa9 \u9084\u6709\u4e00\u500b\u554f\u984c\uff1a\u5982\u679c\u67d0\u4e00\u90e8\u5206\u7684\u5c0d\u61c9\u7269\u70ba\u7a7a\uff0c\u5982\u4f55\u8a08\u7b97\u5176\u76f8\u4f3c\ufa01\uff1f\u6211\u5011\u9019\u88cf\u63a1\u7528\u4e00 3. \u5f9e\u6240\u6709\u7684\u76f8\u4f3c\ufa01\u503c\u4e2d\u522a\u53bb\u90a3\u4e9b\u5df2\u7d93\u5efa\uf9f7\u5c0d\u61c9\u95dc\u4fc2\u7684\u5143\u7d20\u7684\u76f8\u4f3c\ufa01\u503c\uff1b</td></tr><tr><td>\u95dc\u4fc2\u7684\uf978\u500b\u7fa9\u539f\u6bd4\u8f03\u76f8\u4f3c\uff0c\u56e0\u70ba\u5b83\u5011\u5728\u5be6\u969b\u7684\u8a9e\uf9be\u4e2d\u4e92\u76f8\u53ef\u4ee5\u66ff\u63db\u7684\u53ef\u80fd\u6027\u5f88\u5927\u3002\u5c0d\u65bc \u7a2e\u7c21\u55ae\u7684\u8655\uf9e4\u8fa6\u6cd5\uff1a 4. \u91cd\u8907\u4e0a\u8ff0\u7b2c 2 \u6b65\u548c\u7b2c 3 \u6b65\uff0c\u76f4\u5230\u6240\u6709\u7684\u76f8\u4f3c\ufa01\u503c\u90fd\u88ab\u522a\u9664\uff1b</td></tr><tr><td>\u9019\u500b\u554f\u984c\u9019\u88cf\u6211\u5011\uf967\u5c55\u958b\u8a0e\uf941\u3002 \u5c07\u4efb\u4e00\u975e\u7a7a\u503c\u8207\u7a7a\u503c\u7684\u76f8\u4f3c\ufa01\u5b9a\u7fa9\u70ba\u4e00\u500b\u6bd4\u8f03\u5c0f\u7684\u5e38\uf969(\u03b4)\uff1b 5. \u6c92\u6709\u5efa\uf9f7\u8d77\u5c0d\u61c9\u95dc\u4fc2\u7684\u5143\u7d20\u8207\u7a7a\u5143\u7d20\u5c0d\u61c9\u3002</td></tr><tr><td>\u53e6\u5916\uff0c\u5728\u77e5\u7db2\u7684\u77e5\uf9fc\u63cf\u8ff0\u8a9e\u8a00\u4e2d\uff0c\u5728\u4e00\u4e9b\u7fa9\u539f\u51fa\u73fe\u7684\u4f4d\u7f6e\u53ef\u80fd\u51fa\u73fe\u4e00\u500b\u5177\u9ad4\u8a5e(\u6982 \u4e0b\u9762\u6211\u5011\u5206\u5225\u8003\u616e\u96c6\u5408\u548c\u7279\u5fb5\u7d50\u69cb\u7684\u76f8\u4f3c\ufa01\u8a08\u7b97\u554f\u984c\u3002 \u6839\u64da\u4e0a\u8ff0\u6f14\u7b97\u6cd5\u5efa\uf9f7\u8d77\uf978\u500b\u96c6\u5408\u5143\u7d20\u7684\u4e00\u4e00\u5c0d\u61c9\u95dc\u4fc2\u5f8c\uff0c\u6211\u5011\u5c31\u5f88\u5bb9\uf9e0\u8a08\u7b97\uf978\u500b\u96c6</td></tr><tr><td>\uf9a3)\uff0c\u4e26\u7528\u5713\u62ec\u865f( )\u62ec\u8d77\uf92d\u3002\u6240\u4ee5\u6211\u5011\u5728\u8a08\u7b97\u76f8\u4f3c\ufa01\u6642\u9084\u8981\u8003\u616e\u5230\u5177\u9ad4\u8a5e\u548c\u5177\u9ad4\u8a5e\u3001\u5177 \u9ad4\u8a5e\u548c\u7fa9\u539f\u4e4b\u9593\u7684\u76f8\u4f3c\ufa01\u8a08\u7b97\u3002\uf9e4\u60f3\u7684\u505a\u6cd5\u61c9\u8a72\u662f\u5148\u628a\u5177\u9ad4\u8a5e\u9084\u539f\u6210\u300a\u77e5\u7db2\u300b\u7684\u8a9e\u7fa9\u904b \u7b97\u5f0f\uff0c\u7136\u5f8c\u518d\u8a08\u7b97\u76f8\u4f3c\ufa01\u3002\u9019\u6a23\u505a\u5c07\u5c0e\u81f4\u51fd\uf969\u7684\u905e\u8ff4\u8abf\u7528\uff0c\u9019\u6703\u4f7f\u6f14\u7b97\u6cd5\u8b8a\u5f97\u5f88\u8907\u96dc\u3002 \u5408\u7684\u76f8\u4f3c\ufa01\uf9ba\uff1a\u96c6\u5408\u7684\u76f8\u4f3c\ufa01\u7b49\u65bc\u5176\u5143\u7d20\u5c0d\u7684\u76f8\u4f3c\ufa01\u7684\u52a0\u6b0a\u5e73\u5747\u3002\u53c8\u56e0\u70ba\u96c6\u5408\u7684\u5143\u7d20\u4e4b 4.4.1 \u7279\u5fb5\u7d50\u69cb\u7684\u76f8\u4f3c\ufa01\u8a08\u7b97 \u9593\u90fd\u662f\u5e73\u7b49\u7684\uff0c\u6240\u4ee5\u6211\u5011\u53ef\u4ee5\u5c07\u6240\u6709\u7684\u6b0a\u503c\u53d6\u6210\u76f8\u540c\u7684\uff0c\u65bc\u662f\uff1a\u96c6\u5408\u7684\u76f8\u4f3c\ufa01\u7b49\u65bc\u5176\u5143 \u7279\u5fb5\u7d50\u69cb\u53ef\u4ee5\uf9e4\u89e3\u70ba\u4e00\u500b\"\u5c6c\u6027\uff1a\u503c\uff02\u5c0d(Attribute-Value Pair)\u7684\u96c6\u5408\uff0c\u6211\u5011\u5c07\u4e00\u500b\"\u5c6c \u7d20\u5c0d\u7684\u76f8\u4f3c\ufa01\u7684\u7b97\u8853\u5e73\u5747\u3002 \u6027\uff1a\u503c\uff02\u5c0d\u7a31\u70ba\u4e00\u500b\"\u7279\u5fb5\uff02(Feature)\u3002\u5728\u4e00\u500b\u7279\u5fb5\u7d50\u69cb\u4e2d\uff0c\u6bcf\u500b\"\u7279\u5fb5\uff02\u7684\"\u5c6c\u6027\uff02 \u7531\u65bc\u5177\u9ad4\u8a5e\u5728\u300a\u77e5\u7db2\u300b\u7684\u8a9e\u7fa9\u904b\u7b97\u5f0f\u4e2d\u53ea\u5360\u5f88\u5c0f\u7684\u6bd4\uf9b5\uff0c\u56e0\u6b64\uff0c\u5728\u6211\u5011\u7684\u5be6\u9a57\u4e2d\uff0c\u70ba\uf9ba \u7c21\u5316\u8d77\ufa0a\uff0c\u6211\u5011\u505a\u5982\u4e0b\u898f\u5b9a\uff1a \u662f\u552f\u4e00\u7684\u3002 4.4.3 \u5be6\u8a5e\u6982\uf9a3\u76f8\u4f3c\ufa01\u7684\u8a08\u7b97</td></tr><tr><td>\u8a08\u7b97\uf978\u500b\u7279\u5fb5\u7d50\u69cb\u7684\u76f8\u4f3c\ufa01\uff0c\u9996\u5148\u8981\u5728\uf978\u500b\u7279\u5fb5\u7d50\u69cb\u7684\u7279\u5fb5\u4e4b\u9593\u5efa\uf9f7\u8d77\u4e00\u4e00\u5c0d\u61c9\u7684 \u7531\u524d\u9762\u7684\u5206\u6790\u6211\u5011\u77e5\u9053\uff0c\u5728\u300a\u77e5\u7db2\u300b\u4e2d\u5c0d\u4e00\u500b\u5be6\u8a5e\u7684\u63cf\u8ff0\u53ef\u4ee5\u8868\u793a\u70ba\u4e00\u500b\u7279\u5fb5\u7d50\u69cb\uff0c\u8a72 \u5177\u9ad4\u8a5e\u8207\u7fa9\u539f\u7684\u76f8\u4f3c\ufa01\u4e00\uf9d8\u8655\uf9e4\u70ba\u4e00\u500b\u6bd4\u8f03\u5c0f\u7684\u5e38\uf969(\u03b3)\uff1b \u7279\u5fb5\u7d50\u69cb\u542b\u6709\u4ee5\u4e0b\u56db\u500b\u7279\u5fb5\uff1a \u95dc\u4fc2\u3002\u7531\u65bc\u6bcf\u500b\u7279\u5fb5\u7d50\u69cb\u7684\u5404\u500b\u7279\u5fb5\u90fd\u5177\u6709\uf967\u540c\u7684\u5c6c\u6027\uff0c\u56e0\u6b64\u9019\u7a2e\u4e00\u4e00\u5c0d\u61c9\u95dc\u4fc2\u901a\u904e\u7279 \u5177\u9ad4\u8a5e\u548c\u5177\u9ad4\u8a5e\u7684\u76f8\u4f3c\ufa01\uff0c\u5982\u679c\uf978\u500b\u8a5e\u76f8\u540c\uff0c\u5247\u70ba 1\uff0c\u5426\u5247\u70ba 0\u3002 \u5fb5\u7684\u5c6c\u6027\u5f88\u5bb9\uf9e0\u5efa\uf9f7\u8d77\uf92d\uff1a\u5c6c\u6027\u76f8\u540c\u7684\u7279\u5fb5\u4e4b\u9593\u4e00\u4e00\u5c0d\u61c9\uff0c\u5982\u679c\u6c92\u6709\u5c6c\u6027\u76f8\u540c\u7684\u7279\u5fb5\uff0c \u7b2c\u4e00\u57fa\u672c\u7fa9\u539f\u63cf\u8ff0\uff1a\u5176\u503c\u70ba\u4e00\u500b\u57fa\u672c\u7fa9\u539f\uff0c\u6211\u5011\u5c07\uf978\u500b\u6982\uf9a3\u7684\u9019\u4e00\u90e8\u5206\u7684\u76f8\u4f3c\ufa01\u8a18</td></tr><tr><td>4.3 \u865b\u8a5e\u6982\uf9a3\u7684\u76f8\u4f3c\ufa01\u7684\u8a08\u7b97 \u90a3\u9ebc\u8a72\u7279\u5fb5\u7684\u5c0d\u61c9\u7269\u70ba\u7a7a\u3002 \u70ba Sim 1</td></tr></table>",
+                "type_str": "table",
+                "text": "\u8a5e\u8a9e\u76f8\u4f3c\ufa01\u8a08\u7b97 \u5c0d\u65bc\uf978\u500b\u6f22\u8a9e\u8a5e\u8a9e W 1 \u548c W 2 \uff0c\u5982\u679c W 1 \u6709 n \u500b\u7fa9\u9805(\u6982\uf9a3)\uff1aS 11 \uff0cS 12 \uff0c\u2026\u2026\uff0cS 1n \uff0cW 2 \u6709 m \u500b\u7fa9\u9805(\u6982\uf9a3)\uff1aS 21 \uff0cS 22 \uff0c\u2026\u2026\uff0cS 2m \uff0c\u6211\u5011\u898f\u5b9a\uff0cW 1 \u548c W 2 \u7684\u76f8\u4f3c\ufa01\u662f\u5404\u500b\u6982\uf9a3 \u90fd\u53ef\u4ee5\u5206\u89e3\u6210\u4ee5\u4e0b\u90e8\u5206\uff1aA \u5206\u89e3\u6210 A 1 \uff0cA 2 \uff0c\u2026\u2026\uff0cA n \uff0cB \u5206 \u89e3\u6210 B 1 \uff0cB 2 \uff0c\u2026\u2026\uff0cB m \uff0c\u90a3\u9ebc\u9019\u4e9b\u90e8\u5206\u4e4b\u9593\u7684\u5c0d\u61c9\u95dc\u4fc2\u5c31\u6709 m\u00d7n \u7a2e\u3002\u554f\u984c\u662f\uff1a\u9019\u4e9b\u90e8"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O02/O02-2004.json

@@ -0,0 +1,433 @@
+{
+    "paper_id": "O02-2004",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:05:52.070530Z"
+    },
+    "title": "A Study on Noun Sense Disambiguation Based on Syntagmatic Features",
+    "authors": [
+        {
+            "first": "Wang",
+            "middle": [],
+            "last": "Hui",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "Peking University",
+                "location": {
+                    "postCode": "100871",
+                    "settlement": "Beijing",
+                    "country": "P.R.China"
+                }
+            },
+            "email": "whui@pku.edu.cn"
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "Word sense disambiguation (WSD) plays an important role in many areas of natural language processing, such as machine translation, information retrieval, sentence analysis, and speech recognition. Research on WSD has great theoretical and practical significance. The main purposes of this study were to study the kind of knowledge that is useful for WSD, and to establish a new WSD model based on syntagmatic features, which can be used to disambiguate noun sense in Mandarin Chinese effectively. Close correlation has been found between lexical meaning and its distribution. According to a study in the field of cognitive science [Choueka, 1983], people often disambiguate word sense using only a few other words in a given context (frequently only one additional word). Thus, the relationships between one word and others can be effectively used to resolve ambiguity. Based on a descriptive study of more than 4,000 Chinese noun senses, a multi-level framework of syntagmatic analysis was designed to describe the syntactic and semantic constraints of Chinese nouns. All of these polyseme nouns were surveyed, and it was found that different senses have different and complementary distributions at the syntax and/or collocation levels. This served as a foundation for establishing an WSD model by using grammatical information and a thesaurus provided by linguists.",
+    "pdf_parse": {
+        "paper_id": "O02-2004",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "Word sense disambiguation (WSD) plays an important role in many areas of natural language processing, such as machine translation, information retrieval, sentence analysis, and speech recognition. Research on WSD has great theoretical and practical significance. The main purposes of this study were to study the kind of knowledge that is useful for WSD, and to establish a new WSD model based on syntagmatic features, which can be used to disambiguate noun sense in Mandarin Chinese effectively. Close correlation has been found between lexical meaning and its distribution. According to a study in the field of cognitive science [Choueka, 1983], people often disambiguate word sense using only a few other words in a given context (frequently only one additional word). Thus, the relationships between one word and others can be effectively used to resolve ambiguity. Based on a descriptive study of more than 4,000 Chinese noun senses, a multi-level framework of syntagmatic analysis was designed to describe the syntactic and semantic constraints of Chinese nouns. All of these polyseme nouns were surveyed, and it was found that different senses have different and complementary distributions at the syntax and/or collocation levels. This served as a foundation for establishing an WSD model by using grammatical information and a thesaurus provided by linguists.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "The model uses the Grammatical Knowledge-base of Contemporary Chinese [Yu Shiwen et al. 2002] as one of its main machine-readable dictionaries (MRDs). It can provide rich grammatical information for disambiguation of Chinese lexicons, such as parts-of-speech (POS) and syntax functions.",
+                "cite_spans": [
+                    {
+                        "start": 70,
+                        "end": 93,
+                        "text": "[Yu Shiwen et al. 2002]",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Another resource of the model is the Semantic Dictionary of Contemporary Chinese [Wang Hui et al. 1998 ], which provides a thesaurus and semantic collocation information of more than 20,000 nouns. They were employed to analyze 635 Chinese polysemous nouns.",
+                "cite_spans": [
+                    {
+                        "start": 81,
+                        "end": 102,
+                        "text": "[Wang Hui et al. 1998",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "By making full use of these two MRD resources and a very large POS-tagged corpus of Mandarin Chinese, a multi-level WSD model based on syntagmatic features was developed. The experiment described at the end of the paper verifies that the approach achieves high levels of efficiency and precision. [1]\u570b\u52d9\u9662/n \u50d1\u52d9/n \u8fa6\u516c\u5ba4/n \u4e3b\u4efb/n \u90ed\u6771\u5761/nr \u5411/p \u6d77\u5916/s \u540c\u80de/n \u548c/c \u570b\u5167/s \u6b78 \u50d1/n\u3001\u50d1\u7737/n\u3001\u50d1\u52d9/n \u5de5\u4f5c\u8005/n \u767c\u8868/v \u65b0\uf98e/t \u8cc0\u8a5e/n\u3002 [1]\u4ed6/r \u6c7a\u5b9a/v \u5e36/v \u65b0/a \uf9b4\u5c0e/n \u5230/v \u9019/r \u5ea7/q \uf94c/n \u770b\u770b/v\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "[2]\u53bb\uf98e/t\uff0c\u5e02/n \u518d/d \u5c31\u696d/v \u8fa6\u516c\u5ba4/n \u63d0\u4f9b/v \uf9ba/u 3 \u842c/m \u5143/q \u8cb8\u6b3e/n\u3002 [3]\u4ed6\u5011/r \u8863\u8457/n \u9bae\uf977/a\uff0c\u4e00\u770b\uf965\u77e5/l \u662f/v \u5f9e\u4e8b/v \u8fa6\u516c\u5ba4/n \u5de5\u4f5c/n \u7684/u\u3002 [4]\u8077\u5de5\u5011/n \u8dd1\u9032/v \u5ee0\u9577/n \u8fa6\u516c\u5ba4/n, \u8208\u596e/a \u7684/u \u795e\u614b/n \u96e3\u4ee5\u8a00\u8868/l\u3002 [5]\u6bcf/r \u9593/q \u8fa6\u516c\u5ba4/n \u90fd/d \u662f/v \u73bb\u7483/n \uf925\u9580/n\u3002 [6]\u5728/p \u7c21\uf951/a \u7684/u \u8fa6\u516c\u5ba4/n \u88cf/f\uff0c\u912d\u671d\u9293/nr \u526f/b \u5ee0\u9577/n \u8868\u793a/v \uf9ba/u \u8b39\u614e/a \u7684 /u \uf914\u89c0/an\u3002 [7]\u66fe/d \u5728/p \u8fa6\u516c\u5ba4/n \u5c07/p 25 \u842c/m \u65e5\u5143/n \u7684/u \u9910\u8cbb/n \u55ae\u64da/n \u4ea4\u4e88/v",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u8a5e\u7fa9\u6d88\u6b67(WSD)\u6982\u8ff0",
+                "sec_num": "1."
+            },
+            {
+                "text": "[2]\uf94c/n \u9ad8/a \uf9ba/y \uff0c\uf934\u767e\u59d3/n \u7684/u \u751f\u6d3b/vn \u74b0\u5883/n \u6539\u5584/v \uf9ba/y\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u8a5e\u7fa9\u6d88\u6b67(WSD)\u6982\u8ff0",
+                "sec_num": "1."
+            },
+            {
+                "text": "[3]\u4ed6/r \u8d70\u9032/v \uf94c/n \u5167/f \uff0c\uf94c\u9053/n \u5341\u5206/m \u660f\u6697/a\u3002 \u53ea\u6709\u4e0a\u4e0b\u6587\u4e2d\u51fa\u73fe\uf9ba\"\uf94c\uff02 \u7fa9\u7684\u5178\u578b\u642d\u914d\u7279\u5fb5\u6642\uff0c\u5982\u4e0b\u9762\uf9b5 5 \u4e2d\"\uf94c\uff02\u524d\u9762\u6709\u52d5 \u8a5e\"\u4e0b\uff02\uff0c\uf9b5 6 \u4e2d\"\uf94c\uff02\u524d\u9762\u6709\u52d5\u8a5e\"\u4e0a\uff02\uff0c\uf9b5 7 \u4e2d\u7684\"\uf94c\uff02\u524d\u6709\uf969\u8a5e\"11\uff02\u548c\"8\uff02\uff0c \uf9b5 8 \u4e2d\u7684\"\uf94c\uff02\u524d\u6709\uf97e\u8a5e\"\u5c64\uff02\uff0c\u96fb\u8166\u501f\u52a9\u65bc\u9019\u4e9b\u7279\u5fb5\u8a5e\u624d\u5c07\u7cfb\u7d71\u7684\u9810\u8a2d\u503c\u53d6\u6d88\uff0c\u5224\u65b7 \u51fa\u9019\u5e7e\u500b\"\uf94c\uff02\u90fd\u662f \u7fa9\u3002\u5982\uff1a ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u8a5e\u7fa9\u6d88\u6b67(WSD)\u6982\u8ff0",
+                "sec_num": "1."
+            },
+            {
+                "text": "[4]\u5abd/n \uf934/a \uf9ba/y\uff0c\u817f\u8173/n \uf967/d \uf9dd\uf96a/a \uf9ba/y\uff0c\uf90d\u5f97/v \u4e0b/v \uf94c/n \u5566/y\uff01 [5]\uf90f/nr \u79d1\u9577/n \u89aa\u81ea/d \u5f9e/p 11/m \uf94c/n \u5c07/p \u5e2b\u5085/n \u6276\u5230/v 8/m \uf94c/n\u3002 [6]\u4ed6/r \u7adf\u7136/d \u6c92\u6709/d \u770b\u5230/v \u4e00/m \u68df/q \uf978/m \u5c64/q \uf94c/n \u7684/u \u623f\u5b50/n\u3002 \uf9dd\u7528\u9019\u7a2e\u8fa6\u6cd5\uff0c\u96fb\u8166\u8fc5\u901f\u6307\u51fa 67 \u500b\"\uf94c\uff02\u4e2d\u6709 23 \u500b\u8868\u793a \u7fa9\u3002\u7d93\u6aa2\u67e5\u53ea\u6709\u4e0b\u9762 1 \uf9b5\u932f\u8aa4\uff0c\u5176\u4ed6\u5168\u90e8\u6b63\u78ba\u3002 \u963f\u897f\u2022\u8cfd\u5fb7\u514b\u5df2/nr \u5192/v \u8457/u \u6f2b\u5929/z \u98db\u96ea/n \u8d95\u5f80/v \u70cf\uf939\u6728\u9f4a\u5e02/ns \u516b/m \uf94c/n \u9644\u8fd1 /f \u53bb/v \u7c3d\u8a02/v",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u8a5e\u7fa9\u6d88\u6b67(WSD)\u6982\u8ff0",
+                "sec_num": "1."
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Introduction to the Special Issue on Word Sense Disambiguation: The State of the Art",
+                "authors": [
+                    {
+                        "first": "Nancy; Jean",
+                        "middle": [],
+                        "last": "Ide",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "V\u00e9ronis",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Computational Linguistics",
+                "volume": "24",
+                "issue": "1",
+                "pages": "1--40",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ide, Nancy; Jean V\u00e9ronis. \"Introduction to the Special Issue on Word Sense Disambiguation: The State of the Art\", Computational Linguistics. Vol.24, No.1, 1998. pp1-40",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "LSD-C -A. linguistic-based word-sense disambiguation algorithm for Chinese",
+                "authors": [
+                    {
+                        "first": "Kam-Fai",
+                        "middle": [],
+                        "last": "Lam Sze-Sing",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Vincent",
+                        "middle": [],
+                        "last": "Wong",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Lum",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Computer Processing of Oriental Languages",
+                "volume": "10",
+                "issue": "4",
+                "pages": "409--422",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "LAM SZE-SING, KAM-FAI WONG, and VINCENT LUM. \"LSD-C -A. linguistic-based word-sense disambiguation algorithm for Chinese\". Computer Processing of Oriental Languages, Vol. 10, No. 4, 1997, pp 409-422",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Automatic sense disambiguation: How to tell a pine from an ice cream cone",
+                "authors": [
+                    {
+                        "first": "Michal",
+                        "middle": [],
+                        "last": "Lesk",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1986,
+                "venue": "The 1986 SIGDOC Conference. New Yark, ACM",
+                "volume": "",
+                "issue": "",
+                "pages": "24--26",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lesk, Michal. \"Automatic sense disambiguation: How to tell a pine from an ice cream cone\". In: Association for Computing Machinery, eds. The 1986 SIGDOC Conference. New Yark, ACM. 1986. pp24-26",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Statistical Sense Disambiguation with Relatively Small Corpora Using Dictionary Definitions",
+                "authors": [
+                    {
+                        "first": "Alpha",
+                        "middle": [
+                            "K"
+                        ],
+                        "last": "Luk",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "The 33 rd Annual Meeting of ACL",
+                "volume": "",
+                "issue": "",
+                "pages": "181--188",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Luk, Alpha K. \"Statistical Sense Disambiguation with Relatively Small Corpora Using Dictionary Definitions\". In: ACL eds. The 33 rd Annual Meeting of ACL, Cambridge, Massachusetts. 1995. pp181-188",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Selection and Information: A Class-Based Approach to Lexical Relation",
+                "authors": [
+                    {
+                        "first": "Philip",
+                        "middle": [],
+                        "last": "Resnik",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "23--54",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Resnik, Philip. \"Selection and Information: A Class-Based Approach to Lexical Relation\". [Ph. D. Dissertation], USA: University of Pennsylvania. 1993. pp 23-54",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Disambiguating Highly Ambiguous Words",
+                "authors": [
+                    {
+                        "first": "Ellen",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Voorhees",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Computational Linguistics",
+                "volume": "24",
+                "issue": "1",
+                "pages": "125--145",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Ellen M. Voorhees. \"Disambiguating Highly Ambiguous Words\". Computational Linguistics, Vol.24, No.1, 1998. pp125-145",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "A Study of Semantic Disambiguation Based on HowNet",
+                "authors": [
+                    {
+                        "first": "Yang",
+                        "middle": [],
+                        "last": "Xiaofeng",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Li",
+                        "middle": [],
+                        "last": "Tangqiu",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2002,
+                "venue": "Computational Linguistics and Chinese Language Processing",
+                "volume": "7",
+                "issue": "",
+                "pages": "47--78",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Yang Xiaofeng, Li Tangqiu. \"A Study of Semantic Disambiguation Based on HowNet\". Computational Linguistics and Chinese Language Processing. Vol.7, No.1, 2002, pp47-78",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Decision Lists for Lexical Ambiguity Resolution: Application to Accent Restoration in Spanish and French",
+                "authors": [
+                    {
+                        "first": "David",
+                        "middle": [],
+                        "last": "Yarowsky",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1994,
+                "venue": "The 32 nd Annual Meeting of Association for Computational Linguistics. Las Cruces, NM: ACL",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Yarowsky, David. \"Decision Lists for Lexical Ambiguity Resolution: Application to Accent Restoration in Spanish and French\". In: ACL eds. The 32 nd Annual Meeting of Association for Computational Linguistics. Las Cruces, NM: ACL, 1994. pp 88-95 \uf9e1\u6d93\u5b50. \"\u6f22\u8a9e\u8a5e\u7fa9\u6392\u6b67\u65b9\u6cd5\u7814\u7a76\" [\u535a\u58eb\u5b78\u4f4d\uf941\u6587].",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "\u6f22\u8a9e\u5168\u6587\u6aa2\uf96a\u4e2d\u7684\u7fa9\u9805\u6a19\u6ce8\u6280\u8853\u7814\u7a76\". \u300a\u8a08\u7b97\u8a9e\u8a00\u5b78\u9032\u5c55\u8207\u61c9\u7528\u300b. \uf963\u4eac\uff1a\u6e05 \u83ef\u5927\u5b78\u51fa\u7248\u793e\uff0c1995",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "\uf9c7\u958b\u745b",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "\uf9c7\u958b\u745b. \"\u6f22\u8a9e\u5168\u6587\u6aa2\uf96a\u4e2d\u7684\u7fa9\u9805\u6a19\u6ce8\u6280\u8853\u7814\u7a76\". \u300a\u8a08\u7b97\u8a9e\u8a00\u5b78\u9032\u5c55\u8207\u61c9\u7528\u300b. \uf963\u4eac\uff1a\u6e05 \u83ef\u5927\u5b78\u51fa\u7248\u793e\uff0c1995.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "\u82f1\u6f22\u6a5f\u5668\u7ffb\u8b6f\u4e2d\u8a5e\u7fa9\u6d88\u6b67\u65b9\u6cd5\u7684\u7814\u7a76",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "\uf9c7\u5c0f\u864e",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "\uf9c7\u5c0f\u864e. \"\u82f1\u6f22\u6a5f\u5668\u7ffb\u8b6f\u4e2d\u8a5e\u7fa9\u6d88\u6b67\u65b9\u6cd5\u7684\u7814\u7a76\" [\u535a\u58eb\u5b78\u4f4d\uf941\u6587].",
+                "links": null
+            },
+            "BIBREF13": {
+                "ref_id": "b13",
+                "title": "\u6f22\u8a9e\u771f\u5be6\u6587\u672c\u7684\u7fa9\u9805\u6a19\u6ce8",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "\u7ae5\u7fd4",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "\u7ae5\u7fd4. \"\u6f22\u8a9e\u771f\u5be6\u6587\u672c\u7684\u7fa9\u9805\u6a19\u6ce8\" [\u78a9\u58eb\u5b78\u4f4d\uf941\u6587].",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u7fa9\u8a5e\u5178\u7684\u8a2d\u8a08\u8207\u6982\u8981",
+                "authors": [
+                    {
+                        "first": "\u8a79\u885b\u6771",
+                        "middle": [],
+                        "last": "\u738b\u60e0",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "\uf9c7\u7fa4",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "\u738b\u60e0, \u8a79\u885b\u6771, \uf9c7\u7fa4. \"\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u7fa9\u8a5e\u5178\u7684\u8a2d\u8a08\u8207\u6982\u8981\". \u300a1998 \u4e2d\u6587\u4fe1\u606f\u8655\uf9e4\u570b\u969b\u6703\u8b70\uf941 \u6587\u96c6\u300b. \uf963\u4eac: \u6e05\u83ef\u5927\u5b78\u51fa\u7248\u793e. 1998. pp361~367",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF0": {
+                "num": null,
+                "content": "<table><tr><td>\u738b\u60e0</td></tr><tr><td>\u6d88\u6b67\u3002Voorhees [1993]\u3001Resnik [1995] \u5f9e\uf967\u540c\u89d2\ufa01\uf9dd\u7528 WordNet \u4e2d\u7684\u4e0a\u4e0b\u4f4d\u95dc\u4fc2\u3001\u540c\u7fa9 \u7684\u8a5e\u8a9e\u642d\u914d\u548c\u7fa9\uf9d0\u4fe1\u606f(\u5f8c\u8005\u4e3b\u8981\uf92d\u81ea\u65bc\u300a\u540c\u7fa9\u8a5e\u8a5e\uf9f4\u300b\u548c\"\u77e5\u7db2(Hownet)\uff02)\u3002\u7531 \u96fb\u8166\u53ef\u4ee5\u5f88\u5bb9\uf9e0\u5730\u6839\u64da\u8a5e\uf9d0\u6a19\u6ce8\u5224\u65b7\u51fa\u662f\uf9b5[1]\u4e2d\u7684\"\u88dc\u8cbc\uff02\u662f \u7fa9\uff0c\uf9b5[2]\u4e2d\u7684\"\u88dc \uff5e\uff0b\u5177\u9ad4\u7269\uff1a \uff5e\uff0b\u4eba\uff1a</td></tr><tr><td>\u95dc \u4fc2 \u9032 \ufa08 \u82f1 \u8a9e \u8a5e \u7fa9 \u6d88 \u6b67 \u63a2 \uf96a \u3002 Yarowsky[1994] \u63d0 \u51fa \u4e00 \u7a2e \u57fa \u65bc \u7fa9 \uf9d0 \u8a5e \u5178 \u300a Roget's International Thesaurus\u300b\u7684\u8a5e\u7fa9\u6d88\u6b67\u65b9\u6cd5\u3002\u4f7f\u7528\u8a5e\u5178\u4f5c\u70ba\u8a5e\u7fa9\u6d88\u6b67\u77e5\uf9fc\u6e90\u7684\u512a\u9ede\u5728\u65bc\u96fb \u8166\u53ef\u4ee5\u5f9e\u8a5e\u5178\u4e2d\u81ea\u52d5\u7372\u53d6\uf9fc\u5225\u591a\u7fa9\u8a5e\u7684\u5404\u500b\u8a5e\u7fa9\u7684\u4e00\u4e9b\u91cd\u8981\u77e5\uf9fc\u3002\u4f46\u9019\u7a2e\u65b9\u6cd5\u5c0d\u8a5e\u7684\u4e0a \u4e0b\u6587\uf967\u80fd\u9032\ufa08\u9810\u6e2c\uff0c\u800c\u4e14\uff0c\u5c0d\u8a5e\u7fa9\u6d88\u6b67\u6709\u5e6b\u52a9\u7684\u4e00\u4e9b\u7d44\u5408\u7279\u5fb5\u6c92\u6709\u5728\u8a5e\u5178\u4e2d\u5b8c\u5168\u9ad4\u73fe\u51fa \uf92d\u3002 \u65bc\u8a5e\u5178\u548c\u8a9e\uf9be\u5eab\u4e2d\uf967\u53ef\u80fd\u5305\u62ec\u6bcf\u500b\u8a5e\u7684\u6240\u6709\u642d\u914d\u5be6\uf9b5\uff1b\u800c\u6709\u4e9b\u4f4e\u983b\u8a5e\uff0c\u5728\u8a9e\uf9be\u4e2d\u51fa\u73fe\u6b21 \uf969\u4e5f\uf967\u591a\uff0c\u5f88\u96e3\u641c\u96c6\u5230\u5b83\u5011\u7684\u4e0a\u4e0b\u6587\u74b0\u5883\uff0c\u56e0\u800c\u77e5\uf9fc\u7372\u53d6\u4e2d\u666e\u904d\u5b58\u5728\u8457\u8cc7\uf9be\u7a00\u758f\u4ee5\u53ca\u81ea \u52d5\u5b78\u7fd2\u6f14\u7b97\u6cd5\u7684\uf96b\uf969\u7a7a\u9593\u592a\u5927\u7b49\u554f\u984c\u3002 2. \u57fa\u65bc\u7d44\u5408\u7279\u5fb5\u7684\u6f22\u8a9e\u8a5e\u7fa9\u6d88\u6b67 \u8cbc\uff02\u662f \u7fa9\uff0c\u5f9e\u800c\u7d66\u51fa\u6b63\u78ba\u7684\u8a9e\u7fa9\u6a19\u6ce8\u6216\u82f1\u8a9e\u8b6f\u6587\uff1a [1] This will be subsidized by the state. \uff5e\u9580/\uff5e\u7a97\u6236/\uff5e\u73bb\u7483 \uff5e\u4e3b\u4efb/\uff5e\u79d8\u66f8/\uff5e\u4eba\u54e1 \uff5e\uff0b\u540d\u8a5e \u7279\u5b9a\u642d\u914d\uff1a \u76f4\u63a5\u4f5c\u5b9a\u8a9e \uff5e\u5de5\u4f5c \uff5e\uff0b\u65b9\u4f4d\u8a5e \uff5e\u88cf/\uff5e\u524d/\uff5e\u5167/\uff5e\u5f8c\u9762 / [2] 2.2 \u8a5e\uf9d0\u76f8\u540c\uff0c\u5247\uf9dd\u7528\uf901\u7d30\u7dfb\u7684\u8a9e\u6cd5\u529f\u80fd\u8207\u8a5e\u7fa9\u642d\u914d\u5dee\uf962\u9032\ufa08\u8a5e\u7fa9\u6d88\u6b67 \uff5e\uff0b\u8655\u6240\u8a5e \uff5e\u9580\u53e3/\uff5e\u9580\u524d /</td></tr><tr><td>\u7531\u65bc\u81ea\u7136\u8a9e\u8a00\u4e2d\u4e00\u8a5e\u591a\u7fa9\u73fe\u8c61\u666e\u904d\u5b58\u5728\uff0c\u56e0\u6b64\uff0c\u8981\u8b93\u96fb\u8166\u6b63\u78ba\u5730\u5206\u6790\u548c\uf9e4\u89e3\u81ea\u7136\u8a9e\u8a00\uff0c \u4e00\u500b\u91cd\u8981\u7684\u524d\u63d0\u5c31\u662f\u80fd\u5920\u5728\u67d0\u500b\u7279\u5b9a\u4e0a\u4e0b\u6587\u4e2d\uff0c\u81ea\u52d5\u6392\u9664\u6b67\u7fa9\uff0c\u78ba\u5b9a\u591a\u7fa9\u8a5e\u7684\u610f\u7fa9\u3002\u9019 \u5c31\u662f\u901a\u5e38\u6240\uf96f\u7684\u8a5e\u7fa9\u6d88\u6b67(Word sense disambiguation\u3002 \u8a5e\u7fa9\u6d88\u6b67\u662f\u5927\u591a\uf969\u81ea\u7136\u8a9e\u8a00\u8655\uf9e4\u4efb\u52d9\u7684\u4e00\u500b\u5fc5\uf967\u53ef\u5c11\u7684\u4e2d\u9593\u5c64\u6b21\uff0c\u4f7f\u7528\u5e36\u8a5e\u7fa9\u6a19\u6ce8 \u7684\u6587\u672c\u53ef\u4ee5\u63d0\u9ad8\u8cc7\u8a0a\u6aa2\uf96a\u4e2d\u7684\u67e5\u5168\uf961\u548c\u67e5\u6e96\uf961\uff0c\u5be6\u73fe\u57fa\u65bc\u6982\uf9a3\u7684\u6aa2\uf96a\uff1b\u53ef\u4ee5\u5c0d\u6f22\u8a9e\uf906\u6cd5 \u5206\u6790\u4e2d\uf9d0\u5e8f\u540c\u5f62\u7684\u6b67\u7fa9\u554f\u984c\u7684\u89e3\u6c7a\u63d0\u4f9b\u5fc5\u8981\u7684\u8a9e\u7fa9\u4fe1\u606f\uff0c\u70ba\u81ea\u52d5\uf906\u6cd5\u6d88\u6b67\u63d0\u4f9b\u5e6b\u52a9\uff1b\u5728 \u6a5f\u5668\u7ffb\u8b6f\u4e2d\u6709\uf9dd\u65bc\u9078\u64c7\u53ef\u4ee5\u6070\u7576\u8868\u9054\u8a9e\uf906\u4e2d\u8a5e\u7684\u76ee\u6a19\u8a5e\uff0c\u4ee5\u63d0\u9ad8\u7ffb\u8b6f\u7684\u6e96\u78ba\u6027\uff1b\uf9dd\u7528\u5927 \u898f\u6a21\u5e36\u8a5e\u7fa9\u6a19\u6ce8\u7684\u8a9e\uf9be\u5eab\u9084\u53ef\u4ee5\u5efa\uf9f7\u57fa\u65bc\u8a9e\u7fa9\uf9d0\u7684\u8a9e\u8a00\u6a21\u578b\uff0c\u70ba\u8a9e\u97f3\uf9fc\u5225\u3001\u624b\u5beb\u9ad4\uf9fc\u5225 \u548c\u97f3\u5b57\u8f49\u63db\u63d0\u4f9b\u5e6b\u52a9\u3002\u56e0\u6b64\uff0c\u8a5e\u7fa9\u6d88\u6b67\u7814\u7a76\u5728\u81ea\u7136\u8a9e\u8a00\u8655\uf9e4\uf9b4\u57df\u5177\u6709\u91cd\u8981\u7684\uf9e4\uf941\u548c\u5be6\u8e10 \u610f\u7fa9\u3002\u5f9e 50 \uf98e\u4ee3\u521d\u671f\u958b\u59cb\u5c31\u4e00\u76f4\u5099\u53d7\u8a08\u7b97\u8a9e\u8a00\u5b78\u5bb6\u7684\u95dc\u6ce8[Ide, 1998]\u3002 1.1 \u8a5e\u7fa9\u6d88\u6b67\u7684\u77e5\uf9fc\u6e90 \u65e9\u671f\u4eba\u5011\u6240\u4f7f\u7528\u7684\u8a5e\u7fa9\u6d88\u6b67\u77e5\uf9fc\u4e00\u822c\u662f\u6191\u4eba\u624b\u5de5\u7de8\u5236\u7684\u898f\u5247\u3002\u4f46\u624b\u5de5\u7de8\u5beb\u898f\u5247\u8cbb\u6642\u8cbb \uf98a\uff0c\u5b58\u5728\u56b4\u91cd\u7684\u77e5\uf9fc\u7372\u53d6\u7684\"\u74f6\u9838\uff02\u554f\u984c\uff0c\u53ea\u80fd\u8655\uf9e4\u70ba\uf969\u6709\u9650\u7684\u500b\u5225\u8a5e\uff0c\u7121\u6cd5\u52dd\u4efb\u8655\uf9e4 \u5927\u898f\u6a21\u6587\u672c\u7684\u8a5e\u7fa9\u6a19\u6ce8\u5de5\u4f5c\u3002 20 \u4e16\u7d00 80 \uf98e\u4ee3\u4ee5\u5f8c\uff0c\u8a5e\u5178\u6210\u70ba\u4eba\u5011\u7372\u53d6\u8a5e\u7fa9\u6d88\u6b67\u77e5\uf9fc\u7684\u4e00\u500b\u91cd\u8981\u77e5\uf9fc\u6e90\u3002 Lesk[1986]\u3001Luk[1995]\u6839\u64da\u300aOxford Advanced Learner's Dictionary\u300b\u4e2d\u7684\u91cb\u7fa9\u6587\u672c\uf92d\u5224 \u65b7\u591a\u7fa9\u8a5e\u5728\u4e0a\u4e0b\u6587\u4e2d\u7684\u8a5e\u7fa9\u3002Dagan[1991]\u3001Gale[1993]\uf9dd\u7528\u96d9\u8a9e\u5c0d\u7167\u8a5e\u5178\uf92d\u5e6b\u52a9\u591a\u7fa9\u8a5e \u8fd1\uf98e\uf92d\uff0c\u96a8\u8457\u96fb\u8166\u5b58\u5132\u5bb9\uf97e\u548c\u904b\u7b97\u901f\ufa01\u7684\u98db\u901f\u63d0\u9ad8\uff0c\u901a\u904e\u4f7f\u7528\u5404\u7a2e\u6a5f\u7528\u8cc7\u6e90\u548c\u5927\u898f \u6a21\u8a9e\uf9be\u5eab\uff0c\u96fb\u8166\u80fd\u5920\u81ea\u52d5\u7372\u5f97\u5404\u7a2e\u52d5\u614b\u7684\u642d\u914d\u77e5\uf9fc\u53ca\u5176\u7d71\u8a08\u8cc7\uf9be\uff0c\u4ee5\u6b64\u89e3\u6c7a\u898f\u5247\u65b9\u6cd5\u4e2d \u7684\u77e5\uf9fc\u7a7a\u7f3a\u554f\u984c\u3002\u56e0\u800c\uff0c\u8a5e\u7fa9\u6d88\u6b67\u7814\u7a76\u4e2d\u6e67\u73fe\u51fa\u8a31\u591a\u57fa\u65bc\u8a9e\uf9be\u5eab\u7d71\u8a08\u7684\u65b9\u6cd5\u3002\u6bd4\u5982\uff0cGale &amp; Church[1992,1993]\u7b49\uf9dd\u7528\u96d9\u8a9e\u8a9e\uf9be\u5eab\u5c0d\u82f1\u8a9e\u591a\u7fa9\u8a5e\u9032\ufa08\u8a13\uf996\u548c\u6e2c\u8a66\u3002\u4f46\u4f7f\u7528\u96d9\u8a9e\u8a9e\uf9be \u5eab\u7684\u4e3b\u8981\u554f\u984c\u662f\uff1a\u7372\u5f97\u591a\u7fa9\u8a5e\u6d88\u6b67\u77e5\uf9fc\u7684\u524d\u63d0\u662f\u4e00\u500b\u591a\u7fa9\u8a5e\u5728\u53e6\u4e00\u7a2e\u8a9e\u8a00\u4e2d\u5177\u6709\uf967\u540c\u7684 \u7ffb\u8b6f\u8a5e\uff0c\u4e26\u4e14\u7ffb\u8b6f\u8a5e\u5728\u53e6\u4e00\u7a2e\u8a9e\u8a00\u4e2d\u5fc5\u9808\u662f\u55ae\u7fa9\u8a5e\uff0c\u9019\u6a23\u5fc5\u7136\u9650\u5b9a\uf9ba\u591a\u7fa9\u8a5e\u7684\u8655\uf9e4\u7bc4 \u570d\u3002\u5176\u6b21\uff0c\u96d9\u8a9e\u8a9e\uf9be\u5eab\u7684\u898f\u6a21\u548c\u591a\u6a23\u6027\u90fd\u5f88\u6709\u9650\uff0c\u5927\uf97e\u591a\u7fa9\u8a5e\u6216\u591a\u7fa9\u8a5e\u7684\u67d0\u500b\u8a5e\u7fa9\u5728\u8a9e \uf9be\u4e2d\u53ef\u80fd\u5f9e\u672a\u51fa\u73fe\uff1b\u800c\u4e14\u7531\u65bc\u73fe\u5728\u96d9\u8a9e\u8a9e\uf9be\u5c0d\u9f4a\u6280\u8853\u5c1a\uf967\u80fd\u9054 100%\u7684\u6b63\u78ba\uff0c\u4e5f\u4f7f\u5f97\u9019 \u7a2e\u65b9\u6cd5\u53ea\u80fd\u9650\u5b9a\u5728\u5c0f\u898f\u6a21\u7684\u5be6\u9a57\u4e2d\u3002 \u7e3d\u7684\uf92d\uf96f\uff0c\uf967\u7ba1\u662f\u57fa\u65bc\u898f\u5247\u7684\u65b9\u6cd5\uff0c\u9084\u662f\u57fa\u65bc\u8a5e\u5178\u7684\u65b9\u6cd5\uff0c\u6216\u8005\u57fa\u65bc\u5927\u898f\u6a21\u8a9e\uf9be\u5eab \u7684\u65b9\u6cd5\uff0c\u4efb\u4f55\u8a5e\u7fa9\u6d88\u6b67\u7cfb\u7d71\u90fd\uf9ea\uf967\u958b\u8a5e\u7fa9\u6d88\u6b67\u6642\u6240\u7528\u77e5\uf9fc\u7684\u8cc7\uf9be\u6e90\uff0c\u8a5e\u7fa9\u6d88\u6b67\u77e5\uf9fc\u5eab\u7684 \u8cea\uf97e\u5df2\u6210\u70ba\u8a5e\u7fa9\u6d88\u6b67\u7cfb\u7d71\u6210\u6557\u7684\u95dc\u9375\u3002\u82f1\u8a9e\u8a5e\u7fa9\u6d88\u6b67\u7814\u7a76\u5df2\u6709\u591a\uf98e\u7684\uf98c\u53f2\uff0c\u4f46\u5927\u90e8\u5206\u5de5 \u4f5c\u90fd\u7531\u65bc\u7f3a\u5c11\u8db3\u5920\u7684\u8a5e\u7fa9\u77e5\uf9fc\uff0c\u5f9e\u800c\u88ab\u9650\u5236\u5728\u4e00\u500b\u8f03\u5c0f\u7684\u898f\u6a21(\u5e7e\u500b\u6216\u5341\u5e7e\u500b\u8a5e)\uff0c\u5927 \u898f\u6a21\u82f1\u8a9e\u8a9e\uf9be\u5eab\u9032\ufa08\u8a5e\u7fa9\u6a19\u6ce8\u7684\u5de5\u4f5c\u8fc4\u4eca\u5c1a\u672a\ufa0a\u5230\u3002 1.2 \u6f22\u8a9e\u8a5e\u7fa9\u6d88\u6b67\u7814\u7a76 \u8eab\u4efd\uff0b\uff5e\uff1a \u975e\u6307\u4eba\u540d\u8a5e\uff0b\uff5e\uff1a \u6211\u5011\u77e5\u9053\uff0c\u8a5e\u7fa9\u548c\u8a5e\u7684\u5206\u4f48\u4e4b\u9593\u5177\u6709\u5bc6\ufa00\u7684\u95dc\u4fc2\u3002\u4e00\u500b\u8a5e\u7121\uf941\u5305\u542b\u591a\u5c11\u7a2e\u610f\u7fa9(sense)\uff0c \u5982\u679c\u4e00\u500b\u8a5e\u7684\u5e7e\u500b\u610f\u7fa9\u90fd\u5c6c\u65bc\u540d\u8a5e\uff0c\u8a5e\u6027\u6a19\u8a18\u5c31\u7121\u80fd\u70ba\uf98a\uf9ba\u3002\u9019\u6642\uff0c\u53ef\u4ee5\u6839\u64da\uf901\u7d30\u7dfb\u7684 \u6559 \u54e1 \uff5e / \uf934 \u5e2b \uff5e / \u6703 \u8a08 \uff5e / \u91ab \u7e23 \u8a8c \uff5e / \u570b \u52d9 \u9662 \u65b0 \u805e \uff5e / \u5916 \u5728\u4e00\u5b9a\u8a9e\uf906\u4e2d\u8d77\u4f5c\u7528\u7684\uff0c\u5f80\u5f80\u53ea\u662f\u5176\u4e2d\u67d0\u4e00\u500b\u610f\u7fa9\u3002\u8a5e\u7684\uf967\u540c\u610f\u7fa9\u5f80\u5f80\u6703\u5728\uf906\u6cd5\u6216\u8fad\u5f59 \u7d44\u5408\u7279\u5fb5\uf92d\u5340\u5206\u8a5e\u7fa9\u3002\u5c31\u73fe\u4ee3\u6f22\u8a9e\u540d\u8a5e\u800c\u8a00\uff0c\uf967\u50c5\uf969\uf97e\u5de8\u5927\uff0c\u800c\u4e14\u64da\u7b46\u8005\u7d71\u8a08\uff0c\u300a\u73fe\u4ee3 \u751f\uff5e/\u500b\u4eba\uff5e/\u4e3b\u4efb\uff5e \u4e8b\uff5e/\u6e2f\u6fb3\u4e8b\u52d9\uff5e/\u4ea4\uf9e0\u6703\uff5e \u642d\u914d\u5c64\u9762\u4e0a\u8868\u73fe\u51fa\uf967\u540c\u7684\u7d44\u5408\u7279\u5fb5\u3002\u4eba\u5011\u4e4b\u6240\u4ee5\u80fd\u5920\u5728\u4e00\u5b9a\u7684\u4e0a\u4e0b\u6587\u4e2d\uf9e4\u89e3\u591a\u7fa9\u8a5e\u7684\uf967 \u540c\u610f\u7fa9\uff0c\u6b63\u662f\u501f\u52a9\u65bc\u9019\u4e9b\u5f7c\u6b64\u7368\uf9f7\u4e26\u4e14\u5448\u4e92\u88dc\u5206\u4f48\u7684\u7279\u5fb5\u3002\u8a8d\u77e5\u8a9e\u8a00\u5b78\u5bb6 Choueka[1983] \u6f22\u8a9e\u8a9e\u6cd5\u4fe1\u606f\u8a5e\u5178\u8a73\u89e3\u300b\u6240\u5305\u542b\u7684 3491 \u500b\u540d\u8a5e\u4e2d\uff0c\u6709 23%\u662f\u591a\u7fa9\u8a5e\uff0c\u55ae\u5b57\u8a5e\u4e2d\u591a\u7fa9\u8a5e \u7684\u6bd4\uf9b5\uf901\u662f\u9ad8\u9054 47.5%\u3002\u55ae\u5b57\u8a5e\u5e73\u5747\u6709 2.8 \u500b\u610f\u7fa9\uff0c\u96d9\u97f3\u7bc0\u8a5e\u6709 2.2 \u500b\uff0c\u4e09\u97f3\u7bc0\u8a5e\u6709 2 \u8077\u4f4d\uff0b\uff5e\uff1a \u8077\u4f4d\uff0b\uff5e\uff1a \u540d\u8a5e\uff0b\uff5e \u6821\u9577\uff5e/\u6240\u9577\uff5e/\u5ee0\u9577\uff5e/ \u9662 \u6821 \u9577 \uff5e / \u5834 \u9577 \uff5e / \u6240 \u9577 \uff5e / \u5ee0 \u7684\u7814\u7a76\u8868\u660e\uff0c\u4eba\u5011\u901a\u5e38\u50c5\u50c5\uf9dd\u7528\u4e0a\u4e0b\u6587\u4e2d\u7684\u4e00\u500b\u8a5e\u6216\u5c11\uf969\u5e7e\u500b\u8a5e\u5c31\u80fd\u5920\uf9fc\u5225\u51fa\u591a\u7fa9\u8a5e\u7684 \u8a5e\u7fa9\u3002\u56e0\u6b64\uff0c\u5b8c\u5168\u53ef\u4ee5\u6839\u64da\u8a5e\u8207\u8a5e\u4e4b\u9593\u7684\u7d44\u5408\u95dc\u4fc2\uf92d\u6709\u6548\u5730\u5206\u5316\u591a\u7fa9\u8a5e\u3002 \u5c0d\u65bc\u96fb\u8166\uf92d\uf96f\uff0c\u8981\u771f\u6b63\u6709\u6548\u5730\u63d0\u9ad8\u8a5e\u7fa9\u6d88\u6b67\u7684\u6c34\u5e73\uff0c\uf967\u50c5\u9700\u8981\u7372\u53d6\u8a5e\u7684\u91cb\u7fa9\u548c\u5206\uf9d0 \u500b\u3002\u5982\uff1a \u3010\u8fa6\u516c\u5ba4\u3011 \u8fa6\u516c\u7684\u5c4b\u5b50\u3002 \u6a5f\u95dc\u3001\u5b78\u6821\u3001\u4f01\u696d\u7b49\u55ae\u4f4d\u5167\u8fa6\uf9e4\ufa08\u653f\u6027\u4e8b\u52d9\u7684\u90e8\u9580\u3002 \u591a\u7fa9\u540d\u8a5e\u5167\u90e8\u7684\u8a5e\u7fa9\u95dc\u4fc2\u4e5f\u662f\u932f\u7d9c\u8907\u96dc\u7684\uff0c\u6bd4\u5982\uff0c\u6709\u7684\u662f\"\u90e8\u5206\uff5e\u6574\u9ad4\uff02\u95dc\u4fc2\uff0c\u6709 \u9577\uff5e/\u7e3d\uf9e4\uff5e/\u7e3d\u7d71\uff5e \u9577\uff5e/ \u9662\u9577\uff5e/\u7e3d\uf9e4\uff5e/\u7e3d\u7d71\uff5e \u76f4\u63a5 \u7d44\u7e54\u6a5f\u69cb\uff0b\uff5e\uff1a \u7d44\u7e54\u6a5f\u69cb\uff0b\uff5e\uff1a \u4f5c\u4e2d\u5fc3\u8a9e \u5e02\u653f\u5e9c\uff5e/\u7d21\u7e54\u5c40\uff5e/\u5916\u6587\u7cfb\uff5e \uf98c\u53f2\u7cfb\uff5e/\u5c08\u6848\u7d44\uff5e\u5152\u7ae5\u6751\uff5e \u4fe1\u606f\uff0c\u800c\u4e14\uf901\u91cd\u8981\u7684\u662f\uff0c\u7d9c\u5408\uf9dd\u7528\u73fe\u6709\u7684\u8a9e\u8a00\u77e5\uf9fc\u8cc7\u6e90\uff0c\u5728\u8a5e\uf9d0\u5283\u5206\u57fa\u790e\u4e0a\uff0c\u589e\u52a0\u8a5e\u7fa9 \u7684\u8a9e\u6cd5\u529f\u80fd\u5206\u6790\u548c\u8a9e\u5f59\u642d\u914d\u63cf\u5beb\uff0c\u5f9e\u591a\u77e5\uf9fc\u6e90\u4e2d\u63d0\u53d6\u591a\u7fa9\u8a5e\u7684\u6bcf\u500b\u610f\u7fa9\u5728\uf967\u540c\u5c64\u7d1a\u4e0a\u76f8 \u4e92\u5340\u5225\u7684\u7d44\u5408\u7279\u5fb5\u3002 \u672c\u6587\u5728\uf963\u4eac\u5927\u5b78\u8a08\u7b97\u8a9e\u8a00\u5b78\u7814\u7a76\u6240\u958b\u767c\u7684\"\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u6cd5\u4fe1\u606f\u8a5e\u5178\uff02[\u4fde\u58eb\u6c76\u7b49, 2002]\u3001\"\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u7fa9\u8a5e\u5178\uff02[\u738b\u60e0\u7b49, 1998]\u548c\u5927\u898f\u6a21\u8a9e\uf9be\u5eab\u7684\u57fa\u790e\u4e0a\uff0c\u63d0\u51fa\uf9ba\u4e00\u7a2e\u57fa \u65bc\u591a\u7d1a\u7d44\u5408\u7279\u5fb5\u7684\u73fe\u4ee3\u6f22\u8a9e\u8a5e\u7fa9\u6d88\u6b67\u7b56\uf976\u3002 2.1 \uf9dd\u7528\u8a5e\uf9d0\u6a19\u8a18\u9032\ufa08\u8a5e\u7fa9\u6d88\u6b67 \u5f9e\u8a9e\u8a00\u8cc7\u8a0a\u8655\uf9e4\u89d2\ufa01\uf92d\u770b\uff0c\u8a5e\u7684\u7d44\u5408\u7279\u5fb5\u53ef\u4ee5\u5206\u70ba\uf978\u5927\uf9d0\uff0c\u4e00\uf9d0\u662f\u8a5e\uf9d0\u6a19\u8a18\uff0c\u4e00\uf9d0\u662f\u8a5e \u7684\u662f\u6bd4\u55bb\u95dc\u4fc2\uff0c\"\u8fa6\u516c\u5ba4\uff02\u7684 \u7fa9\u5247\u662f\u5f9e \u7fa9\u5f15\u7533\u800c\uf92d\u7684\u3002\u56e0\u6b64\uff0c\u5982\u4f55\u9078\u53d6\u6070\u7576\u7684\u8a5e\u7fa9 \uf969\uf97e\u8a5e\uff0b\uff5e \u4e00\u9593\uff5e/\u4e00\u500b\uff5e \u4e00\u500b\uff5e \u7d44\u5408\u7279\u5fb5\uf92d\u628a\u63e1\uf969\u76ee\u9f90\u96dc\u7684\u540d\u8a5e\uff0c\u6210\u70ba\u554f\u984c\u7684\u95dc\u9375\u3002 \u672c\u6587\u5728\u5c0d 4000 \u9918\u500b\u540d\u8a5e\u7fa9\u9805\u5177\u9ad4\u5206\u6790\u7684\u57fa\u790e\u4e0a\uff0c\u63d0\u51fa\uf9ba\u4e00\u500b\u591a\u7ea7\u7684\u73fe\u4ee3\u6f22\u8a9e\u540d\u8a5e \u52d5\u8a5e\uff0b\uff5e \u5c31\u696d\u5b89\u7f6e\uff5e/\u6d88\u8cbb\u6307\u5c0e\uff5e/\u6625 / \u904b\uff5e/\u4f4f\u623f\u89e3\u56f0\uff5e/\ufa03\u696d\u751f\u7522\uff5e \u8a5e\u7fa9\u7d44\u5408\u5206\u6790\u6846\u67b6\uff1a\u9996\u5148\uff0c\u8003\u5bdf\u540d\u8a5e\u5145\u7576\u4e3b\u8a9e\u3001\u8cd3\u8a9e\u3001\u5b9a\u8a9e\u3001\u4e2d\u5fc3\u8a9e\u7b49\uf906\u6cd5\u6210\u5206\u7684\u80fd\uf98a \u53ca\u5176\u6240\u7d50\u5408\u7684\u8a5e\uf9d0\uff1b\u7136\u5f8c\uff0c\u9032\u4e00\u6b65\u63ed\u793a\u5b83\u5728\u6bcf\u500b\u8a9e\u6cd5\u4f4d\u7f6e\u4e0a\u7684\u8a9e\u7fa9\u642d\u914d\u9650\u5236\u3002 \u4eba\u7a31\u4ee3\u8a5e\uff0b \u6211\u7684\uff5e/\u4f60\u7684\uff5e/\u4ed6\u7684\uff5e / \u7684\uff0b\uff5e \u9019\u500b\u5206\u6790\u6846\u67b6\u628a\u7cfb\u7d71\u7684\u8a9e\u6cd5\u5206\u6790\u8207\uf9b2\u6563\u7684\u8fad\u5f59\u8a9e\u7fa9\u642d\u914d\u6709\u6a5f\u5730\u7d50\u5408\u5728\u4e00\u8d77\u3002\uf9dd\u7528 \uf901\u91cd\u8981\u7684\u662f\uff0c\u7531\u65bc\"\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u6cd5\u4fe1\u606f\u8a5e\u5178\uff02\u4e2d\u5df2\u7d93\u5c0d 35000 \u500b\u540d\u8a5e\u5145\u7576\u4e3b\u8a9e\u3001\u8cd3 \u5b83\uff0c\u6211\u5011\u53ef\u4ee5\u5c0d\uf967\u540c\u7684\u540d\u8a5e\u90fd\u53ef\u4ee5\u63a1\u7528\u7d71\u4e00\u7684\u65b9\u6cd5\u548c\u6b65\u9a5f\u9032\ufa08\u7d44\u5408\u7279\u5fb5\u5206\u6790\u3002\u6bd4\u5982\uff0c\"\u8fa6 \u8a9e\u3001\u5b9a\u8a9e\u3001\u4e2d\u5fc3\u8a9e\u7b49\uf906\u6cd5\u6210\u5206\u7684\u80fd\uf98a\u53ca\u5176\u6240\u7d50\u5408\u7684\u8a5e\uf9d0\u505a\uf9ba\u8a73\u7d30\u7684\u63cf\u5beb\uff0c\"\u73fe\u4ee3\u6f22\u8a9e\u8a9e \u516c\u5ba4\uff02\u7684 \u7fa9\u6307\u5efa\u7bc9\u7269\uff0c \u7fa9\u662f\u4eba(\u67d0\u7a2e\u90e8\u9580)\uff0c\u628a\u5b83\u5011\u653e\u5165\u8a72\u6846\u67b6\uff0c\u53ef\u6e05\u695a\u5730\u986f\u793a\u4e8c \u7fa9\u8a5e\u5178\uff02\u5247\u9032\u4e00\u6b65\u70ba\u5b83\u5011\u4e00\u4e00\u6a19\u6ce8\uf9ba\u8a9e\u7fa9\uf9d0\uff0c\u4e26\u523b\u756b\uf9ba\u5b83\u5011\u5728\u6bcf\u500b\u8a9e\u6cd5\u4f4d\u7f6e\u4e0a\u7684\u8a9e\u7fa9\u642d \u8005\u5404\u81ea\u7684\u7d44\u5408\u7279\u5fb5\u53ca\u5176\u5728\u5206\u4f48\u7a7a\u9593\u4e0a\u7684\u5dee\uf962\uff1a \u914d\u9650\u5236\u3002\u56e0\u6b64\uff0c\u901a\u904e\u67e5\u8a5e\u5178\uff0c\u96fb\u8166\u5c31\u53ef\u7372\u5f97\u4e0a\u8ff0\u77e5\uf9fc\u3002 \u6f22\u8a9e\u8a5e\u7fa9\u6d88\u6b67\u7814\u7a76\u5f9e 20 \u4e16\u7d00 90 \uf98e\u4ee3\u4ee5\u5f8c\u624d\u958b\u59cb\uff0c\u4e3b\u8981\u662f\uf9dd\u7528\u8a5e\u5178\u63d0\u4f9b\u7684\u8a9e\u8a00\u77e5\uf9fc\u3002\u6e05 \u83ef\u5927\u5b78\u7ae5\u7fd4[1993]\uf9dd\u7528\u300a\u540c\u7fa9\u8a5e\u8a5e\uf9f4\u300b\u4e2d\u7684\u8a9e\u7fa9\u5206\uf9d0\uff0c\u5c0d\u6f22\u8a9e\u5408\u6210\u8a5e\u4e2d\u7684\u55ae\u5b57\u9032\ufa08\u7fa9\u9805 \u7136\u5f8c\u7528\u4e00\u500b\u4e8c\u5143\u6a21\u578b\u9032\ufa08\u8a13\uf996\u548c\u6e2c\u8a66\uff0c\u9032\ufa08\u6587\u672c\u6a19\u6ce8\u7814\u7a76\uff0c\u6b63\u78ba\uf961\u5728 85%\u5de6\u53f3[\u8f49\u5f15\u81ea\uf9e1 \u77e5\uf9fc\uff0c\u5c0d\u6587\u672c\u4e2d\u7684\u6bcf\u500b\u8a5e\u9032\ufa08\u8a5e\u7fa9\u6a19\u6ce8\uff0c\u5e73\u5747\u6b63\u78ba\uf961\u9054\u5230 84.77%\uff0c\u591a\u7fa9\u8a5e\u6d88\u6b67\u7684\u6b63\u78ba\uf961 [2]\u751f\u6d3b/n \u88dc\u8cbc/n \u5f88/d \u5feb/a \u767c\u5230/v \u707d\u5340/n \u4eba\u6c11/n \u624b/n \u88cf/f\u3002 \u6a23\u9762\uf9f6\u8457\u8a5e\u7fa9\u77e5\uf9fc\u7372\u53d6\u7684\"\u74f6\u9838\uff02\u554f\u984c\u3002\u73fe\u6709\u7684\u5404\u7a2e\u65b9\u6cd5\u6240\uf9dd\u7528\u7684\u77e5\uf9fc\u4e00\u822c\u50c5\u9650\u65bc\u5177\u9ad4 [1]\u9019/r \u5c07/d \u7531/p \u570b\u5bb6/n \u4e88\u4ee5/v \u88dc\u8cbc/v\u3002 \u6f22\u8a9e\u8a5e\u7fa9\u6d88\u6b67\u96d6\u7136\u5728\u8f03\u77ed\u7684\u6642\u9593\u5167\u53d6\u5f97\uf9ba\uf9a8\u4eba\u9f13\u821e\u7684\u9032\u5c55\uff0c\u4f46\u5b83\u8207\u82f1\u8a9e\u8a5e\u7fa9\u6d88\u6b67\u4e00 \u7d93\u904e\u81ea\u52d5\ufa00\u8a5e\u3001\u8a5e\uf9d0\u6a19\u6ce8[\u4ee3\u78bc\u89e3\u91cb\uf96b\ufa0a\u9644\uf93f 1]\u7684\u6587\u672c\uff1a Xiaofeng 2002]\u3002 \u5c0d\u65bc\u8a5e\uf9d0\uf967\u540c\u7684\u610f\u7fa9\uff0c\u96fb\u8166\u53ef\u76f4\u63a5\u501f\u52a9\u65bc\u8a9e\uf9be\u4e2d\u7684\u8a5e\uf9d0\u6a19\u8a18\u9032\ufa08\u5224\u65b7\u3002\u6bd4\u5982\uff0c\u9047\u5230\u4e0b\u9762 \u4ecb\u8a5e\uff0b\uff5e \u5728\uff5e/\u5f9e\uff5e(\u8d70\u904e\uf92d) \u6a5f\u5668\u7ffb\u8b6f\u7b49\u9650\u5b9a\uf9b4\u57df\u4e2d\u7684\u8a5e\u7fa9\u6d88\u6b67\u65b9\u6cd5\u5206\u5225\u9032\ufa08\uf9ba\u63a2\uf96a[\uf9c7\u958b\u745b 1995\uff1b\uf9c7\u5c0f\u864e 1998\uff1bYang \u7531\u65bc\u73fe\u6709\u7684\u6f22\u8a9e\u8a5e\uf9d0\u6a19\u6ce8\u5de5\u5177\u5df2\u7d93\u53ef\u4ee5\u9054\u5230 96%\u7684\u6b63\u78ba\uf961[\uf9e1\u6d93\u5b50 1999\uff1a30]\uff0c\u56e0\u6b64\uff0c \u8abf\u63db\uff5e/\u5750\uff5e \u70ba 52.13%\u3002\u6b64\u5916\uff0c\u5c71\u897f\u5927\u5b78\u3001\u54c8\u723e\u6ff1\u5de5\u696d\u5927\u5b78\u3001\u5ec8\u9580\u5927\u5b78\u4e5f\u5206\u5225\u5c0d\u6f22\u8a9e\u5168\u6587\u6aa2\uf96a\u3001\u82f1\u6f22 2196 \u500b\uff0c\u5176\u4e2d\u610f\u7fa9\u8a5e\uf9d0\uf967\u540c\u7684\u6709 592 \u500b\uff0c\u5360 27%\u3002\u9019\u4e5f\u5c31\u662f\uf96f\uff0c\u50c5\u50c5\uf9dd\u7528\u8a5e\uf9d0\u6a19\u8a18\u5c31\u53ef \u4ee5\u6d88\u9664\u8d85\u904e 1/5 \u7684\u6b67\u7fa9\u3002 \u7279\u5b9a\u642d\u914d\uff1a \u76f4\u63a5\u4f5c\u8cd3\u8a9e \u52d5\u8a5e\uff0b\uff5e \u5230\uff5e/\u53bb\uff5e/\u8d70\u51fa\uff5e/\uf9ea\u958b\uff5e \u8a5e\u8a5e\uf9f4\u300b\u3001\u300a\u73fe\u4ee3\u6f22\u8a9e\u8fad\u6d77\u300b\u4ee5\u53ca\u5f9e\u5927\u898f\u6a21\"\u4eba\u6c11\u65e5\u5831\uff02\u8a9e\uf9be\u5eab\u4e2d\u7372\u53d6\u7684\u8a5e\u8a9e\u52d5\u614b\u642d\u914d \u5b57\u7684\u300a\u4eba\u6c11\u65e5\u5831\u300b\u8a9e\uf9be[1998 \uf98e 1 \u6708]\u7684\u7d71\u8a08\u7d50\u679c\u8207\u6b64\u76f8\u8fd1\uff0c22744 \u500b\u540d\u8a5e\u4e2d\u5171\u6709\u591a\u7fa9\u8a5e \u95d6\u9032\uff5e/\u8d70\u9032\uff5e/\u56de\u5230\uff5e/\u9032\uff5e/ \u6210\uf9f7\uff5e/\u8a2d\uf9f7\uff5e \uf9d0\u4ee3\u78bc\uff0c\u5c0d\u5be6\u8a5e\u591a\u7fa9\u8a5e\u9032\ufa08\u8a5e\u7fa9\u6d88\u6b67\uff0c\u5e73\u5747\u6b63\u78ba\uf961\u70ba 45.5%\u3002\uf9e1\u6d93\u5b50[1999]\uf9dd\u7528\u300a\u540c\u7fa9 \u4e2d\u50cf\"\u88dc\u8cbc\uff02\u9019\u6a23\u5305\u542b\uf967\u540c\u8a5e\uf9d0\u7684\u610f\u7fa9\u7684\u540d\u8a5e\u6709 932\uff0c\u5360\u591a\u7fa9\u540d\u8a5e\u7684 23.4%\u3002\u5c0d 200 \u842c \u8da8\u5411\u52d5\u8a5e\uff0b\uff5e\uff1a \u7279\u5b9a\u642d\u914d\uff1a \u6d93\u5b50 1999\uff1a18]\u3002LAM[1997]\uf9dd\u7528\u300a\u73fe\u4ee3\u6f22\u8a9e\u8a5e\u5178\u300b\u7684\u91cb\u7fa9\u6587\u672c\u548c\u300a\u540c\u7fa9\u8a5e\u8a5e\uf9f4\u300b\u7684\u7fa9 \u3010\u88dc\u8cbc\u3011 \u8cbc\u88dc\uff1a\uff5e\u5bb6\u7528\uff5c\uff5e\uf97b\u50f9\u3002 \u8cbc\u88dc\u7684\u8cbb\u7528\uff1a\u798f\uf9dd\uff5e\uff5c\u526f\u98df\uff5e\u3002 \u64da\u7b46\u8005\u6240\u4f5c\u7684\u8abf\u67e5\uff0c\u300a\u73fe\u4ee3\u6f22\u8a9e\u8a5e\u5178\u300b\u7684 20513 \u500b\u540d\u8a5e\u4e2d\u5171\u6709\u591a\u7fa9\u8a5e 3989 \u500b\uff0c\u5176 \uff5e\uff0b\u5f62\u5bb9\u8a5e \uff5e\u5341\u5206\u5bec\u655e/\uff5e\u7a7a\uf9ba/\uff5e\u5b89\u975c / \u76f4\u63a5\u4f5c\u4e3b\u8a9e \uff5e\uff0b\u52d5\u8a5e \uff5e\u6539\u6697\u623f \uff5e\u63d0\u51fa/\uff5e\u767c\u8868\u8072\u660e/\uff5e\uf96f \u6a19\u6ce8\u3002\u6b64\u5f8c\uff0c\u4e0a\u6d77\uf966\u65e6\u5927\u5b78\u66fe\u4f7f\u7528\u300a\u540c\u7fa9\u8a5e\u8a5e\uf9f4\u300b\u7684\u4e2d\uf9d0\u8a9e\u7fa9\u7de8\u78bc\u4eba\u5de5\u6a19\u6ce8 5 \u842c\u8a9e\uf9be\uff0c \u8868 1 \"\u8fa6\u516c\u5ba4\uff02\u7684\uf978\u500b\u610f\u7fa9\u7d44\u5408\u7279\u5fb5\u5c0d\u6bd4 \uf9dd\u7528\u8868 1 \u4e2d\u7684\u7d44\u5408\u7279\u5fb5\uff0c\u6d88\u6b67\u7cfb\u7d71\u53ef\u4ee5\u5c0d\u5be6\u969b\u6587\u672c\u4e2d\u51fa\u73fe\u7684\u591a\u7fa9\u540d\u8a5e\u7684\u8a5e\u7fa9\u9032\ufa08\u5224 \u5728\u4e0a\u4e0b\u6587\u4e2d\u7684\u8a5e\u7fa9\u642d\u914d\u9650\u5236\u3002\u6f22\u8a9e\u4e2d\u6709\u4e9b\u591a\u7fa9\u8a5e\u7684\uf967\u540c\u610f\u7fa9\u5c6c\u65bc\uf967\u540c\u7684\u8a5e\uf9d0\uff0c\u5982\"\u88dc \u8cbc\uff02\u7684 \u7fa9\u662f\u52d5\u8a5e\uff0c \u7fa9\u662f\u540d\u8a5e\uff1a \u8a9e\u6cd5\u529f\u80fd \u7fa9 \u7fa9 \u65b7\u3002\u6bd4\u5982[\u4ee5\u4e0b\uf9b5\uf906\u4e2d\u7684\u8a5e\uf9d0\u4ee3\u78bc\uf96b\ufa0a\u9644\uf93f]\uff1a</td></tr></table>",
+                "html": null,
+                "text": "Living allowances were quickly handed out to the people in the stricken area.",
+                "type_str": "table"
+            },
+            "TABREF2": {
+                "num": null,
+                "content": "<table><tr><td>1998 \uf98e/t \u7684/u \u623f\u5c4b/n \u627f\u5305/vn \u5408\u540c/n\u3002</td></tr><tr><td>\u7531\u4e0a\u9762\u7684\u5206\u6790\u6211\u5011\u53ef\u4ee5\u6e05\u695a\u5730\u8a8d\uf9fc\u5230\uff0c\u8a5e\u7fa9\u7d44\u5408\u7279\u5fb5\u5206\u6790\u78ba\u5be6\u53ef\u6709\u6548\u5730\u63d0\u9ad8\u8a5e\u7fa9\u6d88</td></tr><tr><td>\u6b67\u77e5\uf9fc\u5eab\u7684\u8cea\uf97e\uff0c\u6eff\u8db3\u6f22\u8a9e\u540d\u8a5e\u8a5e\u7fa9\u81ea\u52d5\u6d88\u6b67\u7684\u9700\u8981\u3002\u4f46\u554f\u984c\u662f\u9019\u6a23\u4e00\u500b\u8a5e\u7fa9\u77e5\uf9fc\u5eab\u898f</td></tr><tr><td>\u6a21\u7a76\u7adf\u591a\u5927\u624d\u80fd\u5920\u9054\u5230\u57fa\u672c\u7684\u5be6\u7528\u6c34\u5e73\u5462\uff1f\u6839\u64da\u300a\u73fe\u4ee3\u6f22\u8a9e\u983b\uf961\u8a5e\u5178\u300b[\uf963\u4eac\u8a9e\u8a00\u5b78\u9662\u51fa</td></tr><tr><td>\u7248\u793e\uff0c1985\uff1a492-514]\u7684\u7d71\u8a08\uff0c1144 \u500b\u9ad8\u983b\u8a5e\u5c0d\u8a9e\uf9be\u7684\u8986\u84cb\u7a0b\ufa01\u7d04\u70ba 75%\uff0c\u800c\u4e14\u5176\u4e2d\uf9d1</td></tr><tr><td>\u6210\u4ee5\u4e0a\u662f\u591a\u7fa9\u8a5e\u3002\u53ef\ufa0a\uff0c\uf969\uf97e\uf967\u591a\u7684\u9ad8\u983b\u591a\u7fa9\u8a5e\u662f\u5f71\u97ff\u6f22\u8a9e\u771f\u5be6\u6587\u672c\u8a5e\u7fa9\u6d88\u6b67\u6e96\u78ba\uf961\u7684 \u672c\u6587\u5de5\u4f5c\u7684\u6700\u57fa\u672c\u601d\u60f3\u662f\u5206\u5c64\u6b21\u63cf\u5beb\u6f22\u8a9e\u8a5e\u7fa9\u7684\u7d44\u5408\u80fd\uf98a\u3002\u76ee\u524d\uff0c\u4e3b\u8981\u662f\u5c0d\u540d\u8a5e\u7684</td></tr><tr><td>\u95dc\u9375\u3002\u5982\u679c\u6211\u5011\u5728\u8a5e\u7fa9\u7d44\u5408\u5206\u6790\u57fa\u790e\u4e0a\uff0c\u5c0d\u9ad8\u983b\u591a\u7fa9\u8a5e\u7684\u5404\u500b\u610f\u7fa9\u7684\u7d44\u5408\u80fd\uf98a\u9032\ufa08\u96c6\u4e2d \u7d44\u5408\u7279\u5fb5\u5206\u6790\u53ca\u5176\u5728\u8a5e\u7fa9\u6d88\u6b67\u4e2d\u7684\u61c9\u7528\u9032\ufa08\uf9ba\u4e00\u4e9b\u8a66\u9a57\u6027\u7684\u63a2\uf96a\u3002\u521d\u6b65\u7684\u5be6\u9a57\u7d50\u679c\u662f\uf9a8</td></tr><tr><td>\u7814\u7a76\u548c\u8a73\u7d30\u63cf\u8ff0\uff0c\uf967\u50c5\u53ef\u4ee5\u6709\u6548\u5730\u63d0\u9ad8\u8a5e\u7fa9\u77e5\uf9fc\u5eab\u7684\u8cea\uf97e\uff0c\u800c\u4e14\u4e5f\u53ef\u4ee5\u6307\u5c0e\u81ea\u52d5\u5b78\u7fd2\u6f14 \u4eba\u6b23\u6170\u7684\uff0c\u6211\u5011\u5e0c\u671b\u5728\u7a4d\uf94f\uf9ba\uf901\u591a\u7684\u5be6\u8e10\u7d93\u9a57\u5f8c\uff0c\u80fd\u9032\u4e00\u6b65\u5b8c\u5584\u9019\u4e00\u8a5e\u7fa9\u7d44\u5408\u5206\u6790\u6846</td></tr><tr><td>\u7b97\u6cd5\u7684\uf96b\uf969\u8a2d\u8a08\uff0c\u5c07\u6703\u5341\u5206\u6709\u52a9\u65bc\u89e3\u6c7a\u6d88\u6b67\u8a9e\u7fa9\u77e5\uf9fc\u7372\u53d6\u7684\u74f6\u9838\u554f\u984c\u3002 \u67b6\uff0c\u4e26\u5c07\u9019\u7a2e\u601d\uf937\u61c9\u7528\u65bc\u52d5\u8a5e\u3001\u5f62\u5bb9\u8a5e\u7684\u8a5e\u7fa9\u77e5\uf9fc\u5eab\u69cb\u9020\u4e4b\u4e2d\uff0c\u540c\u6642\u52aa\uf98a\u5be6\u73fe\u7531\u96fb\u8166\u8f14</td></tr><tr><td>\u52a9\u62bd\u53d6\u8a5e\u7fa9\u7684\u7d44\u5408\u7279\u5fb5\u3002</td></tr><tr><td>4. \u7d50\u8a9e</td></tr><tr><td>\u4efb\u4f55\u8a5e\u7fa9\u6d88\u6b67\u7cfb\u7d71\u90fd\uf9ea\uf967\u958b\u8a5e\u7fa9\u6d88\u6b67\u6642\u6240\u7528\u77e5\uf9fc\u7684\u8cc7\uf9be\u6e90\u3002\u672c\u6587\u63d0\u51fa\uf9ba\u4e00\u7a2e\u5145\u5206\uf9dd\u7528 \u73fe\u6709\u8cc7\u6e90\uff0c\u628a\u8a9e\u6cd5\u529f\u80fd\u3001\u8a9e\u7fa9\u642d\u914d\u7b49\uf967\u540c\u5c64\u9762\u7684\u77e5\uf9fc\u7d71\u4e00\u8d77\uf92d\u5206\u7d1a\u63cf\u5beb\u7684\u8a5e\u7fa9\u7d44\u5408\u7279 \uf96b\u8003\u6587\u737b</td></tr><tr><td>\u5fb5\u5eab\u7684\u8a2d\u8a08\u539f\u5247\uff0c\u4e26\u7d66\u51fa\uf9ba\u4e00\u500b\u57fa\u65bc\u8a5e\u7fa9\u7d44\u5408\u7279\u5fb5\u7684\u8a5e\u7fa9\u6d88\u6b67\u6a21\u578b\uff1a</td></tr></table>",
+                "html": null,
+                "text": "Choueka, Y. and S. Lusignan, \"A Connectionist Scheme for Modeling Word Sense Disambiguation\". Cognition and Brain Theory. 6 (1) 1983, pp.89-120 Dagan, Ido, Alon Itai, and Shaul Markovitch. \"Two Languages Are More Informative Than One\". In: The 29 th Annual Meeting of Association for Computational Linguistics, Berkeley, CA: ACL, 1991. pp 130-137 Gale, William A, Kenneth W. Church, and David Yarowsky. \"Using bilingual materials to develop word sense disambiguation methods\". In: The International Conference on Theoretical and Methodological Issues in Machine Translation, 1992. pp 101-112 Gale, William A, Kenneth W. Church, and David Yarowsky. \"A Method for Disambiguation Word Senses in a Large Corpus\". Computer and the Humanities. (26) 1993. pp 415-439",
+                "type_str": "table"
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O02/O02-2005.json

@@ -0,0 +1,86 @@
+{
+    "paper_id": "O02-2005",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:06:13.781736Z"
+    },
+    "title": "",
+    "authors": [],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "We introduce the development of the Electronic Lexicon of Contemporary Newborn Chinese Words: (1) the definition of a newborn word, (2) the main principle behind constructing the lexicon, (3) the collection of newborn words and their feature descriptions of them, and (4) the classification of 40,000 newborn words. In our opinion, a new bornword is a character string that appeared after 1978 in a new form, with a new meaning and with a new usage. In addition, it must be frequently used and accepted, but the names of men and places are not newborn words according to our definition. The approach to collecting newborn words is quite unrestricted, that is, the more the better. Based on the Contemporary Chinese Grammatical Knowledge Base of the Institute of Computational Linguistics at Peking University, we have finished compiling a lexicon of almost 40,000 newborn words semi-automatically. The lexicon, we believe, is a worthy resource for research on Chinese word-building rules and Natural Language Processing. Firstly, classification is done based on the preponderant grammatical characteristics of each word, and then the detailed features are described in the database of ACCESS. The lexicon contains a total base and three grammatical bases (i.e., a noun base, verb base and adjective base); what's more, it also has an old word base, a loanword base and a acronym base. The entire base is related to the sub-bases through the fields of word, phonetic notation and semantics fields, which form a hypernymy hierarchy that is quite convenient for searching. Totally, there are more than 200 fields in the bases that give information regarding phonetic notation, semantics, sources, word building, syntax and pragmatics. Without doubt, this lexicon is one of the largest domestic lexicons available with the most detailed descriptions of newborn Chinese words.",
+    "pdf_parse": {
+        "paper_id": "O02-2005",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "We introduce the development of the Electronic Lexicon of Contemporary Newborn Chinese Words: (1) the definition of a newborn word, (2) the main principle behind constructing the lexicon, (3) the collection of newborn words and their feature descriptions of them, and (4) the classification of 40,000 newborn words. In our opinion, a new bornword is a character string that appeared after 1978 in a new form, with a new meaning and with a new usage. In addition, it must be frequently used and accepted, but the names of men and places are not newborn words according to our definition. The approach to collecting newborn words is quite unrestricted, that is, the more the better. Based on the Contemporary Chinese Grammatical Knowledge Base of the Institute of Computational Linguistics at Peking University, we have finished compiling a lexicon of almost 40,000 newborn words semi-automatically. The lexicon, we believe, is a worthy resource for research on Chinese word-building rules and Natural Language Processing. Firstly, classification is done based on the preponderant grammatical characteristics of each word, and then the detailed features are described in the database of ACCESS. The lexicon contains a total base and three grammatical bases (i.e., a noun base, verb base and adjective base); what's more, it also has an old word base, a loanword base and a acronym base. The entire base is related to the sub-bases through the fields of word, phonetic notation and semantics fields, which form a hypernymy hierarchy that is quite convenient for searching. Totally, there are more than 200 fields in the bases that give information regarding phonetic notation, semantics, sources, word building, syntax and pragmatics. Without doubt, this lexicon is one of the largest domestic lexicons available with the most detailed descriptions of newborn Chinese words.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "\u5176\u4ed6\u5404\u5eab\u5171\u6709\u7684\u5c6c\u6027\u8cc7\u8a0a\u6709\"\u8a5e\u8a9e\u3001\u62fc\u97f3\u3001\u7fa9\u9805\uff02\uff0c\u5747\u5f9e\u7e3d\u5eab\u4e2d\u7e7c\u627f\uff0c\u5176\u4ed6\u5c6c\u6027\u8cc7 \u8a0a\u5982\u4e0b\u3002 \u8a9e\u6cd5\u8cc7\u8a0a\u5eab\u4e2d\u540d\u8a5e\u5eab\u4e3b\u8981\u63cf\u8ff0\uf9ba\u8207\u540d\u8a5e\u642d\u914d\u7684\u5404\u7a2e\uf97e\u8a5e\uff0c\u540d\u8a5e\u7684\u5b50\uf9d0\uff0c\u80fd\uf967\u80fd\u76f4\u63a5 \u53d7\uf969\u8a5e\u3001\uf969\uf97e\u8a5e\u3001\u5176\u4ed6\u540d\u8a5e\u3001\u52d5\u8a5e\u7684\u4fee\u98fe\uff0c\u80fd\u53d7\u54ea\u4e9b\u4ee3\u8a5e\u76f4\u63a5\u6216\u52a0\"\u7684\uff02\u5f8c\u4fee\u98fe\uff0c\u524d\u63a5 \u6216\u5f8c\u63a5\u6210\u5206\uff0c\u80fd\uf967\u80fd\u4f5c\u5b9a\u8a9e\u3001\u4e3b\u8a9e\u3001\u8cd3\u8a9e\uff0c\u80fd\uf967\u80fd\u76f4\u63a5\u6216\u52a0\"\u5730\uff02\u4f5c\uf9fa\u8a9e\uff0c\u4ee5\u53ca\u80fd\uf967\u80fd \u91cd\u758a\u3001\uf9f6\u6642\u5145\u7576\uf97e\u8a5e\u7b49\u3002\u52d5\u8a5e\u5eab\u4e3b\u8981\u63cf\u8ff0\u7684\u8cc7\u8a0a\u6709\uff1a\u52d5\u8a5e\u7684\u5b50\uf9d0--\u4fc2\u8a5e\u3001\u52a9\u52d5\u8a5e\u3001\u8da8 \u5411\u52d5\u8a5e\u3001\u88dc\u52a9\u52d5\u8a5e\u3001\u5f62\u5f0f\u52d5\u8a5e\u3001\u81ea\u4e3b\u52d5\u8a5e\u3001\u975e\u81ea\u4e3b\u52d5\u8a5e\u3001\u5167\u5916\u52d5\u8a5e\u3001\u5b58\u73fe\u52d5\u8a5e\u3001\uf9ea\u5408\u8a5e \u7b49\uff1b\u69cb\u6210\u7684\uf906\u5f0f--\"\u628a\uff02\u5b57\uf906\u3001\"\u88ab\uff02\u5b57\uf906\u3001\u517c\u8a9e\uf906\u3001\u96d9\u8cd3\uf906\u3001\u5b58\u73fe\uf906\u7b49\uff1b\u5145\u7576\u7684\u6210 \u5206--\u5b9a\u8a9e\u3001\u540d\u8a5e\u6027\u7d50\u69cb\u7684\u4e2d\u5fc3\u8a9e\u3001\u55ae\u4f5c\u8b02\u8a9e\u3001\u8cd3\u8a9e\u3001\uf9fa\u8a9e\uff1b\u5f8c\u5e36\u7684\u6210\u5206--\u9ad4\u8b02\u51c6\u8cd3 \u8a9e\u3001\u52d5\u6642\uf97e\u88dc\u8a9e\u3001\u7d50\u679c\u88dc\u8a9e\u3001\u8da8\u5411\u88dc\u8a9e\u7b49\uff1b\u52d5\u8a5e\u81ea\u8eab\u5f62\u614b\u7684\u8b8a\u5316--\u524d\u53d7\"\uf967\u3001\u6c92\u3001\u5f88\u3001 \u6b63\uff02\u7684\u4fee\u98fe\u3001\u5f8c\u8ddf\"\u8457\u3001\uf9ba\u3001\u904e\uff02\u3001VV\u3001AABB\u3001V \u4e00 V\u3001V \uf9ba V\u3001V \uf9ba\u4e00 V\u3001VVO \u7b49\u3002 \u8cd3\u8a9e\u3001\u7d50\u679c\u88dc\u8a9e\u3001\u8da8\u5411\u88dc\u8a9e\u7684\u8a73\u7d30\u8cc7\u8a0a\u5c07\u53e6\ufa08\u63cf\u8ff0\u3002\u5f62\u5bb9\u8a5e\u5eab\u63cf\u8ff0\u7684\u4e3b\u8981\u8cc7\u8a0a\u6709\uff1a\u5b50\uf9d0\u3001 \u76f4\u63a5\u4f5c\u5b9a\u8a9e\u6216\u52a0\"\u7684\uff02\u5f8c\u4f5c\u5b9a\u8a9e\u3001\u4f5c\u8b02\u8a9e\u3001\u88dc\u8a9e\u3001\uf9fa\u8a9e\u6216\u52a0\"\u5730\uff02\u5f8c\u4f5c\uf9fa\u8a9e\u6216\u518d\u52a0\"\u5f88\uff02 \u5f8c\u4f5c\uf9fa\u8a9e\u3001\u4f5c\u6e96\u8b02\u8cd3\u3001\"\u6709\uff02\u7684\u8cd3\u8a9e\u3001\u540d\u8a5e\u6027\u7d50\u69cb\u7684\u4e2d\u5fc3\u8a9e\u3001AA \u91cd\u758a\u53ca\u91cd\u758a\u5f8c\u7684\u8a5e\u6027\u3001 ABAB\u3001A \u88cf AB\u3001\u5e36\"\u8457\uf9ba\u904e\uff02\u3001\u51c6\u8cd3\u8a9e\u3001\u8da8\u5411\u88dc\u8a9e\u7b49\u3002 \u69cb\u8a5e\u6cd5\u5eab\u63cf\u8ff0\u7684\u4e3b\u8981\u8cc7\u8a0a\u6709\uff1a1\u3001\u69cb\u8a5e\u90e8\u4ef6\uff0c\u5206\u7232\"\u6210\u5206 1\uff02\"\u6210\u5206 2\uff02\"\u6210\u5206 3\uff02\uff0c \u5206\u5225\u586b\u5165\u69cb\u6210\u8a72\u8a5e\u8a9e\u7684\u6210\u5206\u7684\uf9d0\u5225\uff0c\u5176\u4e2d\u6709\u7684\u662f\u8a9e\u7d20\u3001\u6709\u7684\u662f\u8a5e\u30022\u3001\u69cb\u8a5e\u6cd5\uff0c\u8003\u5bdf\u8a72\u8a5e \u8a9e\u7684\u69cb\u8a5e\u65b9\u5f0f\uff0c\u4e3b\u8981\u5206\u7232\uff1a\u4e3b\u8b02\u3001\u52d5\u8cd3\u3001\uf9fa\u4e2d\u3001\u5b9a\u4e2d\u3001\u88dc\u5145\u3001\uf997\u5408\u3001\u52a0\u5b57\u9996\u3001\u52a0\u5c3e\u78bc\u7b49\u3002 3\u3001\u8a5e\u6027\uff0c\u586b\u5165\u8a5e\u8a9e\u7684\u8a5e\u6027\uff0c\u5f9e\u7e3d\u5eab\u4e2d\u7e7c\u627f\uf92d\u30024\u3001\u97f3\u7bc0\uff0c\u586b\u5165\u8a72\u8a5e\u8a9e\u7684\u97f3\u7bc0\uf969\uff0c\u5f9e\u7e3d\u5eab \u4e2d\u7e7c\u627f\uf92d\u3002 \u820a\u8a5e\u5eab\u63cf\u8ff0\u7684\u4e3b\u8981\u8cc7\u8a0a\u6709\uff1a1\u3001\u820a\u7fa9\uff0c\u586b\u5165\u8a72\u8a5e\u8a9e\u539f\uf92d\u7684\u610f\u7fa9\uff1b2\u3001\u65b0\u7fa9\uff0c\u586b\u5165\u8a72\u8a5e \u8a9e\u65b0\u610f\u7fa9\u6216\u65b0\u7528\u6cd5\uff1b3\u3001\u8a5e\u6027\uff0c\u586b\u5165\u8a72\u65b0\u8a5e\u8a9e\u7684\u8a5e\u6027\uff0c\u5982\u679c\u8a5e\u6027\u8207\u539f\u8a5e\u8a9e\u8a5e\u6027\u4e00\u81f4\uff0c\u5247\u6a19 \u8a5e\u6027\u6a19\u8a18\uff1b\u5982\u679c\u6539\u8b8a\uf9ba\u8a5e\u6027\u5247\u7279\u5225\u6a19\u660e\uff0c\u5982\uf967\u53ca\u7269\u52d5\u8a5e\u8b8a\u7232\u53ca\u7269\u52d5\u8a5e\uff0c\u6a19\u7232\uff1aVt\uff1b\u53ca\u7269 \u52d5\u8a5e\u8b8a\u7232\uf967\u53ca\u7269\u52d5\u8a5e\uff0c\u6a19\u7232\uff1aVi\u30024\u3001\u8a5e\u7fa9\u6f14\u8b8a\u9014\u5f91\uff0c\u8003\u5bdf\u7531\u820a\u8a5e\u8a9e\u6f14\u8b8a\u7232\u65b0\u8a5e\u8a9e\u7684\u8a5e\u7fa9 \u7684\u6f14\u8b8a\u9014\u5f91\uff0c\u4e3b\u8981\u6709 36 \uf9d0\uff1a(1)\u540c\u7528\u76f8\u6bd4\uff0c(2)\u540c\u679c\u76f8\u55bb\uff0c(3)\u540c\u8cea\u76f8\u55bb\uff0c(4)\u540c \uf9fa\u76f8\u55bb\uff0c(5)\u7279\u5b9a\u4ee3\u666e\u901a\uff0c(6)\u5177\u9ad4\u5230\u62bd\u8c61\uff0c(7)\u540c\u4f4d\u76f8\u55bb\uff0c(8)\u8a9e\u7d20\u63db\u7fa9\uff0c(9) \u6cdb\u5316\uff0c(10)\u500b\u9ad4\u4ee3\u5168\u9ad4\uff0c(11)\u666e\u901a\u4ee3\u7279\u5b9a\uff0c(12)\u4f7f\u52d5\u5316\uff0c(13)\u540c\u611f\u5f15\u7533\uff0c 14\u7269\u4ef6\uf901\u63db\uff0c 15 ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "(1)\u97f3\u8b6f\uff0c(2)\u8ae7\u97f3\uff0c(3)\u97f3\u8b6f\u52a0\u6f22\u8a9e\u8a9e\u7d20\uff0c(4)\u97f3\u517c\u610f\u8b6f\uff0c(5)\u6309\u7167\u5916\u8a9e\u8a5e\u8a9e\u7684 \u610f\u7fa9\u5275\u9020\u4e00\u500b\u6f22\u8a9e\u8a5e\u8a9e\uff1b2\u3001\u8a9e\u97f3\u8b8a\u5316\u8cc7\u8a0a\uff0c\u4e3b\u8981\u6709\uff1a(1)\u97f3\u7d20\u7684\u66ff\u63db\uff0c(2)\u97f3\u7bc0\u7684\u589e \u6e1b\uff1b3\u3001\u610f\u7fa9\u8b8a\u5316\uff0c\u4e3b\u8981\u6709(1)\u64f4\u5927\uff0c(2)\u7e2e\u5c0f\uff0c(3)\u8f49\u79fb\uff0c(4)\u4fdd\u6301\u539f\u610f\uff1b4\u3001\u7e2e \uf976\uff0c\u8003\u5bdf\u5916\uf92d\u8a5e\u8a9e\u662f\u5426\u6709\u7e2e\uf976\uff1b5\u3001\u61c9\u7528\uf9b4\u57df\u3002",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u5916\uf92d\u8a5e\u8a5e\u5eab\u63cf\u8ff0\u7684\u4e3b\u8981\u8cc7\u8a0a\u6709\uff1a1\u3001\u9014\u5f91\uff0c\u5916\uf92d\u8a5e\u9032\u5165\u6f22\u8a9e\u7684\u4e3b\u8981\u9014\u5f91\uff0c\u4e3b\u8981\u6709\uff1a",
+                "sec_num": null
+            },
+            {
+                "text": "(1)\u7c21\u7a31\uff0c(2)\u7e2e\u8a9e\uff0c(3)\uf976\u8a9e\uff0c(4)\u51c6\u7e2e\uf976\u8a9e\u30023\u3001\u69cb\u6210\u65b9\u5f0f\uff1a\u5c07\u539f\u8a5e\u8a9e\u5283\u6bb5\uff0c\u6839 \u64da\u5be6\u969b\u60c5\u6cc1\u63cf\u8ff0\u5982\u4f55\u9032\ufa08\u7e2e\uf976\u7684\uff0c\u6bd4\u5982\"\uf963\u4eac\u5927\u5b78\uff02-\"\uf963\u5927\uff02\uff0c\u5176\u69cb\u6210\u65b9\u5f0f\u63cf\u5beb\u7232 \"a1b3\uff02\u30024\u3001\u540c\u5f62\uff1a\u5982\u6709\u540c\u5f62\u8a5e\u8a9e\uff0c\u5247\u6709\u5e7e\u500b\u586b\u76f8\u61c9\u7684\uf969\u4f4d\u30025\u3001\u7e2e\uf976\u65b9\u5f0f\uff1a(1)\u7e2e\u5408\uff0c \u5982\uff1a\u5967\uf9f4\u5339\u514b\u904b\u52d5\u6703--\"\u5967\u904b\u6703\uff02\uff1b(2)\u7bc0\u7e2e\uff0c\u5982\uff1a\u96fb\u8996\uf99a\u7e8c\u5287--\"\uf99a\u7e8c\u5287\uff02\uff1b(3) \u63d0\u53d6\uff0c\u5982\uff1a\u4e2d\u570b\u9ad8\u6280\u8853\u7814\u7a76\u767c\u5c55\u8a08\u5283\u7db1\u8981--\"863 \u8a08\u5283\uff02(\u8a72\u8a08\u5283\u7684\u63d0\u51fa\u662f 1986 \uf98e 3 \u6708)\uff1b(4)\u5176\u4ed6\uff0c\u5305\u62ec\uff1aA\u3001\u7528\u540c\u7fa9\u3001\u8fd1\u7fa9\u8a5e\u8a9e\u66ff\u63db\uff0c\u5982\uff1a\u6d6e\u5f0f\u8d77\u91cd\u6a5f--\"\u6d6e\u540a\uff02\uff1b ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "\u7c21\uf976\u8a5e\u8a5e\u5eab\u63cf\u8ff0\u7684\u4e3b\u8981\u8cc7\u8a0a\u6709\uff1a1\u3001\u539f\u8a5e\u8a9e\uff0c\u586b\u5165\u7c21\uf976\u8a5e\u7684\u539f\u578b\u30022\u3001\u7c21\uf976\u7684\uf9d0\u578b\uff1a",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "\u5f35\u5fd7\u6bc5\u3001\u5f35\u6176\u96f2\uff0e\u65b0\u6642\u671f\u65b0\u8a5e\u8a9e\u7684\u8da8\u52e2\u8207\u9078\u64c7\uff0c\u300a\u8a9e\u6587\u5efa\u8a2d\u300b\uff0c1997(3)\uff1a15-18\u3002 \uf90f\u6b63\u5805\uff0e\u300a\u6f22\u8a9e\u8a5e\u7fa9\u5f15\u7533\u5c0e\uf941\u300b\uff0c\u5357\u4eac\uff1a\u5357\u4eac\u5927\u5b78\u51fa\u7248\u793e\uff0c1996\u3002 \u4fde\u58eb\u6c76\uff0e\u300a\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u6cd5\u8cc7\u8a0a\u8a5e\u5178\u8a73\u89e3\u300b\uff0c\uf963\u4eac\uff1a\u6e05\u83ef\u5927\u5b78\u51fa\u7248\u793e\uff0c1998\u3002 \u5f90\u570b\u6176\uff0e\u300a\u73fe\u4ee3\u6f22\u8a9e\u8fad\u5f59\u7cfb\u7d71\uf941\u300b\uff0c\uf963\u4eac\uff1a\uf963\u4eac\u5927\u5b78\u51fa\u7248\u793e\uff0c1999\u3002",
+                "authors": [],
+                "year": null,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "\u5f35\u5fd7\u6bc5\u3001\u5f35\u6176\u96f2\uff0e\u65b0\u6642\u671f\u65b0\u8a5e\u8a9e\u7684\u8da8\u52e2\u8207\u9078\u64c7\uff0c\u300a\u8a9e\u6587\u5efa\u8a2d\u300b\uff0c1997(3)\uff1a15-18\u3002 \uf90f\u6b63\u5805\uff0e\u300a\u6f22\u8a9e\u8a5e\u7fa9\u5f15\u7533\u5c0e\uf941\u300b\uff0c\u5357\u4eac\uff1a\u5357\u4eac\u5927\u5b78\u51fa\u7248\u793e\uff0c1996\u3002 \u4fde\u58eb\u6c76\uff0e\u300a\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u6cd5\u8cc7\u8a0a\u8a5e\u5178\u8a73\u89e3\u300b\uff0c\uf963\u4eac\uff1a\u6e05\u83ef\u5927\u5b78\u51fa\u7248\u793e\uff0c1998\u3002 \u5f90\u570b\u6176\uff0e\u300a\u73fe\u4ee3\u6f22\u8a9e\u8fad\u5f59\u7cfb\u7d71\uf941\u300b\uff0c\uf963\u4eac\uff1a\uf963\u4eac\u5927\u5b78\u51fa\u7248\u793e\uff0c1999\u3002",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "uris": null,
+                "num": null,
+                "type_str": "figure",
+                "text": "\u5ba2\u9ad4\uf901\u63db\uff0c(16)\u6307\u7a31\u7269\u4ef6\u64f4\u5927\uff0c(17)\u5de5\u5177\u5e36\u672c\u9ad4\uff0c(18)\u7279\u5fb5\u3001 \u6a19\u8a8c\u4ee3\u672c\u9ad4\uff0c(19)\u5c08\u5316\uff0c(20)\u540c\u5f62\u76f8\u55bb\uff0c(21)\u4ee5\u679c\u4ee3\u56e0\uff0c(22)\u540c\u6240\u76f8\u55bb\uff0c(23) \u90e8\u5206\u4ee3\u5168\u9ad4\uff0c(24)\u52d5\u975c\u5f15\u7533\uff0c(25)\u6240\u5728\u4ee3(\u6bd4\u5982\uff1a\u5c71\u982d\u3001\u5927\u54e5\u5927)\uff0c(26)\u4e3b\u9ad4\u64f4 \u5927\uff0c(27)\u529f\u7528\u4ee3\u672c\u9ad4\uff0c(28)\u8a9e\u7528(\u5c0f\u59d0)\uff0c(29)\u793e\u6703\u539f\u56e0(\u8349\u696d)\uff0c(30)\u6642\u7a7a \u5f15\u7533\uff0c(31)\u6b63\u53cd\u5f15\u7533\uff0c(32)\uf906\u6cd5\u5f71\u97ff\uff0c(33)\u62bd\u8c61\u5230\u5177\u9ad4\uff0c(34)\u7279\u6307\u5316\uff0c(35) \u672c\u9ad4\u4ee3\u7279\u5fb5\uff0c(36)\u73fe\u8c61\u4ee3\u672c\u9ad4\u3002(\u6839\u64da\uf90f\u6b63\u5805\u7684\u300a\u6f22\u8a9e\u8a5e\u7fa9\u5f15\u7533\u5c0e\uf941\u300b\u548c\u5f90\u570b\u6176\u7684\u300a\u73fe \u4ee3\u6f22\u8a9e\u8fad\u5f59\u7cfb\u7d71\uf941\u300b\u6b78\u7d0d\u51fa\uf92d)5\u3001\u6f14\u8b8a\uf9d0\u578b\uff0c\u8a5e\u7fa9\u6f14\u8b8a\u7684\uf9d0\u578b\uff0c\u4e3b\u8981\u6709 9 \uf9d0\uff1a(1)\u8f49 \u79fb\uff0c(2)\u64f4\u5927\uff0c(3)\u865b\u5316\uff0c(4)\u8f49\uf9d0\uff0c(5)\u7e2e\u5c0f\uff0c(6)\u8cb6\ufa09\uff0c(7)\u63da\u5347\uff0c(8)\u5f31 \u5316\uff0c(9)\u6df1\u5316\u3002"
+            },
+            "TABREF0": {
+                "num": null,
+                "type_str": "table",
+                "html": null,
+                "content": "<table><tr><td>\u5178\uff02)\u7684\u57fa\u672c\u60c5\u6cc1\uff1a(1)\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u7684\u754c\u5b9a(2)\u65b0\u8a5e\u8a9e\u8a5e\u5178\u7684\u958b\u767c\u601d\u60f3(3)\u65b0\u8a5e \u7b49\uff1bB\u3001\u610f\u8b6f\u8a5e\uff0c\u5982\"\u71b1\u9ede(hot spot)\u3001\u97f3\uf914\u96fb\u8996(music television)\u3001\u71b1\u72d7 3.2 \u300a\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u8cc7\u8a0a\u96fb\u5b50\u8a5e\u5178\u300b\u958b\u767c\u7684\u76ee\u6a19 \u904d\u6027\u539f\u5247\u3001\u7a69\u5b9a\u6027\u539f\u5247\u3001\u97f3\u7bc0\u539f\u5247\u7b49\uff0c\u5f9e\u8a5e\u8868\u4e2d\u9074\u9078\u51fa\u65b0\u8a5e\u8a9e 3 \u842c\u591a\u500b\uff0c\u5f62\u6210\uf9ba\u65b0\u8a5e\u8a9e (6) \uf92d\u6e90\u8cc7\u8a0a\uff0c\u5373\u8a72\u8a5e\u5f9e\u90a3\u672c\u8a5e\u5178\u6216\u54ea\u4e9b\u8a9e\uf9be\u4e2d\uf92d\uff0c\u5982\u679c\u5f88\u591a\u672c\u8a5e\u5178\u90fd\u6536\uf93f\uf9ba\u8a72</td></tr><tr><td>\u6458\u8981 \u672c\u6587\u5f9e\u56db\u500b\u65b9\u9762\u4ecb\u7d39\uf9ba\u6211\u5011\u6b63\u5728\u958b\u767c\u4e2d\u7684\u300a\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u8cc7\u8a0a\u96fb\u5b50\u8a5e\u5178\u300b\uff1a \u8a9e\u7684\u63a1\u96c6\u8207\u65b0\u8a5e\u8a9e\u8a5e\u5178\u6240\u63cf\u8ff0\u7684\u5c6c\u6027\u8cc7\u8a0a(4)\u8fd1\u56db\u842c\u8a5e\u8a9e\u7684\u6b78\uf9d0\u5be6\u8e10 2. \u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u7684\u754c\u5b9a \u5c0d\u65bc\"\u65b0\u8a5e\u8a9e\uff02\u76ee\u524d\u5b78\u8853\u754c\u6709\uf967\u540c\u7684\u770b\u6cd5\uff0c\u5728\u5168\u9762\u8003\u5bdf\uf9ba\u8fd1 4 \u842c\u500b\u65b0\u8a5e\u8a9e\u4e26\u4e14\u501f\u9452\u3001\u5438 \u6536\uf9ba\u5b78\u8853\u754c\u65b0\u8a5e\u8a9e\u7814\u7a76\u6210\u679c\u7684\u57fa\u790e\u4e0a\uff0c\u6211\u5011\u8a8d\u7232\u65b0\u8a5e\u8a9e\u53ef\u4ee5\u5b9a\u7fa9\u7232\uff1a\u901a\u904e\u5404\u7a2e\u9014\u5f91\u7523\u751f \u7684\u3001\u5177\u6709\u57fa\u672c\u8a5e\u5f59\u6c92\u6709\u7684\u65b0\u5f62\u5f0f\u3001\u65b0\u610f\u7fa9\u6216\u65b0\u7528\u6cd5\u7684\u8a9e\u6587\u8a5e\u8a9e\u3002\u65b0\u8a5e\u8a9e\u7684\u7279\u9ede\u5728\u65bc \"\u65b0\uff02\uff0c\"\u65b0\uff02\u5177\u9ad4\u8868\u73fe\u5728\u8a5e\u5f62\u3001\u8a5e\u7fa9\u548c\u8a5e\u8a9e\u7684\u7528\u6cd5\u4e0a\u3002\u9452\u5b9a\u65b0\u8a5e\u8a9e\u7684\uf96b\u7167\u7cfb\u662f\u73fe\u4ee3\u6f22 \u8a9e\u57fa\u672c\u8a5e\u5f59\u7684\u8a5e\u5f62\u3001\u8a5e\u7fa9\u548c\u7528\u6cd5\u3002\u53ea\u8981\u5728\u9019\u4e09\u500b\u65b9\u9762\u7684\u4efb\u4f55\u4e00\u9ede\u4e0a\u8207\u73fe\u4ee3\u6f22\u8a9e\u57fa\u672c\u8a5e\u5f59 (hot dog)\u3001\u8d85\u7d1a\u5e02\u5834(supermarket)\uff02\uff1bC\u3001\u97f3\u8b6f\u517c\u610f\u8b6f\u8a5e\uff0c\u5982\"\u9433\u5c04\u3001\u547c \u5566\u5708\u3001\u6851\u62ff\u6d74\u3001\u8ff7\u4f60\u88d9\u3001\u5427\uf981\u3001\u9152\u5427\u3001\uff02\u7b49\uff1bD\u3001\u76f4\u63a5\u4f7f\u7528\u65e5\u8a9e\u7684\u8a5e\u8a9e\uff0c\u5982\uff1a \"\u653e\u9001\u3001\u6170\u5b89\u5a66\u3001\u7269\u8a9e\u3001\u5beb\u771f\u3001\u4eba\u6c23\uff02\u7b49\u3002 (5) \u7c21\uf976\u8a5e\uff0c\u5728\u539f\u6709\u8a5e\u8a9e\u7684\u57fa\u790e\u4e0a\u7e2e\uf976\u800c\u6210\u7684\u8a5e\u8a9e\u3002\u5206\u7232\u4e09\u7a2e\u60c5\u6cc1\uff1aA\u3001\u7c21\u7a31\u8a5e\uff0c \u5982\"\u535a\u5c0e(\u535a\u58eb\u7814\u7a76\u751f\u5c0e\u5e2b)\u3001\u6fb3\u7db2(\u6fb3\u5927\uf9dd\u4e9e\u7db2\u7403\u516c\u958b\u8cfd)\u3001\u8d85\u5e02(\u8d85\u7d1a\u5e02 \u8a5e\u5178\u7684\u57fa\u790e\u3002\u6b64\u5f8c\uff0c\u6211\u5011\uf9dd\u7528\u8a9e\u8a00\u8cc7\u8a0a\u8655\uf9e4\u6280\u8853\uf967\u65b7\u5730\u5f9e\u7db2\u4e0a\u6293\u53d6\u65b0\u8a5e\u8a9e\u53ca\u76f8\u95dc\u7684\uf9b5\uf906 \u8a5e\uff0c\u5247\uf96f\u660e\u8a72\u8a5e\u7684\u8907\u73fe\uf961\u6bd4\u8f03\u9ad8\uff0c\u65b0\u8a5e\u8a9e\u7684\u8eab\u4efd\uf901\u52a0\u78ba\u5b9a\u3002 (1) \u5e0c\u671b\u5275\u5efa\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u7814\u7a76\u7684\u57fa\u790e\u5e73\u81fa\uff0c\u5be6\u73fe\u8cc7\u6e90\u9ad8\ufa01\u5171\u7528\uff0c\u7372\u5f97\u8f03\u9ad8\u7684\u61c9 \u7528\u50f9\u503c\u3002\u672c\u9805\u7814\u7a76\uf9dd\u7528\u96fb\u8166\u8cc7\uf9be\u5eab\u6280\u8853\u548c\u76f8\u95dc\u7684\u8a9e\uf9be\u5eab\u6280\u8853\u9032\ufa08\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e \u8a9e\u7684\u8ddf\u8e64\u7814\u7a76\uff0c\u7814\u7a76\u6210\u679c\u5f62\u5f0f\u7232\u6709\u6548\u3001\u5be6\u7528\u7684\u96fb\u8166\u8cc7\uf9be\u5eab\u8edf\u9ad4\uff0c\u5176\u4e2d\u5305\u62ec\u65b0\u8a5e \u96c6\uff0c\uf967\u65b7\u5730\u64f4\u5145\u65b0\u8a5e\u8a9e\u8a5e\u5178\u3002\u78ba\u5b9a\u65b0\u8a5e\u8a9e\u8a5e\u5178\u4e2d\u7684\u8a5e\u76ee\u5f8c\uff0c\uf9dd\u7528\u65b0\u8a5e\u8a9e\u8a5e\u5178\u8cc7\u8a0a\u5eab\u548c\u5305 (7) \u6642\u9593\u8cc7\u8a0a\u3002\u8a72\u8a5e\u8a9e\u5927\u81f4\u7523\u751f\u7684\u6642\u9593\uff0c\u4ee5\u8a5e\u5178\u7684\u5f15\uf9b5\u6642\u9593\u7232\u51c6\uff1b\u4f7f\u7528\u6642\u9593\uff0c\u4ee5\u8a5e \u542b\u300a\u4eba\u6c11\u65e5\u5831\u300b1978 \uf98e\u4ee5\uf92d\u7684\u8a9e\uf9be\u3001\u300a\u5357\u65b9\u5468\u672b\u300b\u5275\u520a\u4ee5\uf92d\u7684\u8a9e\uf9be\u4ee5\u53ca\u4eba\u6c11\u65e5\u5831\u5831\u7cfb\u5176 \u5178\u51fa\u7248\u6642\u9593\u7232\u51c6\u3002 \u4ed6\u5831\u7d19\u3001\u4eba\u6c11\u7db2\u3001\u5149\u660e\u65e5\u5831\u3001\u65b0\u6c11\u665a\u5831\u7b49\u8fd1\uf98e\uf92d\u8a9e\uf9be\u7684\u8d85\u5927\u898f\u6a21\u8a9e\uf9be\u5eab\u5efa\uf9f7\u5305\u542b\u8a72\u8a5e\u8a9e \u8a9e\u96fb\u5b50\u8a5e\u5178\u548c\u5927\u898f\u6a21\u7684\u76f8\u95dc\u8a9e\uf9be\uff0c\u9019\u6a23\u53ef\u4ee5\u5be6\u73fe\u8cc7\u6e90\u7684\u9ad8\ufa01\u5171\u7528\uff0c\u4f7f\u5176\u5177\u6709\u8f03 \u7684\uf9b5\uf906\u96c6\uff0c\u8003\u5bdf\u9019\u4e9b\u8a5e\u8a9e\u7684\u610f\u7fa9\u548c\u7528\u6cd5\uff0c\u63cf\u8ff0\u5176\u7fa9\u9805\u3001\u8a9e\u6cd5\u5c6c\u6027\u3001\u8a9e\u7fa9\u5c6c\u6027\u4ee5\u53ca\u5176\u4ed6\u8cc7 \u9ad8\u7684\u61c9\u7528\u50f9\u503c\u3002 \u8a0a\u7b49\uff0c\u5f9e\u800c\u958b\u767c\u51fa\u300a\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u8cc7\u8a0a\u96fb\u5b50\u8a5e\u5178\u300b\u3002\u9019\u4e9b\u5de5\u4f5c\u5f88\u5927\u7a0b\ufa01\u4e0a\uf9dd\u7528\u96fb\u8166\u8a9e 4.3 \u65b0\u8a5e\u8a9e\u8a5e\u5178\u7684\u7d50\u69cb\u8207\u5404\u500b\u5eab\u7684\u4e3b\u8981\u5c6c\u6027\u8cc7\u8a0a \u5834)\"\uff1bB\u3001\uf976\u8a9e\u8a5e\uff0c\u5982\"\u56b4\u6253(\u56b4\u53b2\u6253\u64ca\u72af\u7f6a\u6d3b\u52d5)\u3001\u6253\u5047(\u6253\u64ca\u5047\u5192\u50de\uf99d \u5546\u54c1)\u3001\u9632\u50de(\u9632\u6b62\u5047\u5192\u50de\uf99d\u7523\u54c1)\u3001\u53f0\u8cc7(\u81fa\u7063\u4eba\u6295\u5165\u7684\u8cc7\u672c)\uff02\uff1bC\u3001\u7e2e (2) \u5e0c\u671b\u5728\u6f22\u8a9e\u7814\u7a76\u548c\u4e2d\u6587\u8cc7\u8a0a\u8655\uf9e4\u7814\u7a76\u65b9\u9762\u505a\u51fa\u7a4d\u6975\u7684\u8ca2\u737b\u3002\u4ee5\u5f80\u6f22\u8a9e\u7684\u7814\u7a76\u7684 \u8cc7\uf9be\u548c\u624b\u6bb5\u9650\u5236\uf9ba\u6f22\u8a9e\u5927\u898f\u6a21\u7684\u5be6\u7528\u5316\u7684\u7814\u7a76\uff0c\u7531\u6b64\u9020\u6210\u7684\u76f4\u63a5\u5f8c\u679c\u662f\u56b4\u91cd\u5236 \uf9be\u5eab\u7ba1\uf9e4\u6280\u8853\uff0c\u5728\u5927\u898f\u6a21\u6a5f\uf95a\u8a9e\uf9be\u5eab\u7684\u652f\u63f4\u4e0b\u9032\ufa08\uff0c\u80fd\u5920\u6bd4\u8f03\u5168\u9762\u5730\u8003\u5bdf\u6bcf\u500b\u65b0\u8a5e\u8a9e\u7684 4.3.1 \u65b0\u8a5e\u8a9e\u8a5e\u5178\u7684\u7e3d\u9ad4\u7d50\u69cb \u5206\u4f48\u74b0\u5883\uff0c\u63d0\u9ad8\u65b0\u8a5e\u8a9e\u63a1\u96c6\u3001\u6536\uf93f\u7684\u5408\uf9e4\u6027\u548c\u8cc7\u8a0a\u63cf\u8ff0\u7684\u6e96\u78ba\ufa01\u548c\u8986\u84cb\u7bc4\u570d\uff0c\u5f9e\u800c\u63d0\u5347 \u8a9e\u8a5e\uff0c\u5982\"\u4e09\u8b1b\u3001\u4e09\u500b\u4ee3\u8868\u3001\u4e09\u5047\u3001\u4e09\u966a\u3001\u4e09\uf90a\uff02\u7b49\u3002 \u7d04\uf9ba\u4e2d\u6587\u8cc7\u8a0a\u8655\uf9e4\u7684\u767c\u5c55\u3002\u672c\u9805\u7814\u7a76\uf9dd\u7528\u96fb\u8166\u6280\u8853\u9032\ufa08\uff0c\u7a4d\uf94f\uf9ba\u5927\uf97e\u7684\u6a5f\u5668\u53ef \u8a5e\u5178\u7684\u8cea\uf97e\u3002 \u65b0\u8a5e\u8a9e\u8a5e\u5178\u63a1\u7528\u6210\u719f\u7684\u95dc\uf997\u8cc7\uf9be\u5eab\u6280\u8853(\u5728 access \u8edf\u9ad4\u4e0b\u5be6\u73fe)\u3002\u586b\u5165\u7684\u8cc7\u8a0a\u5118\uf97e\u4ee5\u76f4 \uf967\u540c\uff0c\u6211\u5011\u5c31\u8a8d\u7232\u5b83\u662f\u65b0\u8a5e\u8a9e\u3002\u57fa\u672c\u8a5e\u5f59\u7684\u4ee3\u8868\u662f\uf963\u4eac\u5546\u52d9\u5370\u66f8\u9928\u51fa\u7248\u7684 96 \u7248\u7684\u300a\u73fe\u4ee3 (6) \u4fee\u8fad\u7528\u6cd5\u7a69\u5b9a\u4e0b\uf92d\u69cb\u6210\u7684\u65b0\u8a5e\u8a9e\u3002\u4e3b\u8981\u6709\uff1aA\u3001\u6bd4\u55bb\u5f15\u7533\uff0c\u5982\"\u8c46\u8150\u6e23\u5de5\u7a0b\u3001 \uf95a\u6587\u4ef6\uff0c\u7232\u5927\u898f\u6a21\u7684\u5be6\u7528\u7684\u6f22\u8a9e\u7814\u7a76\u5960\u5b9a\uf9ba\u57fa\u790e\uff0c\u5176\u7814\u7a76\u6210\u679c--\u65b0\u8a5e\u8a9e\u5c6c\u6027 \u89c0\u660e\u77ad\u7684\u6f22\u5b57\u3001\u5b57\u6bcd\u3001\uf969\u4f4d\u8868\u793a\u3002\u6839\u64da\u65b0\u8a5e\u8a9e\u5c6c\u6027\u7684\u78ba\uf9f7\uff0c\u8cc7\u8a0a\u5eab\u7e3d\u9ad4\u4e0a\u5305\u62ec\u4e09\u500b\u65b9\u9762 \u6f22\u8a9e\u8a5e\u5178\u300b\u548c\u300a\u6f22\u8a9e\u5927\u8a5e\u5178\u300b\u3002\"\u65b0\uff02\u9084\u6709\u6642\u9593\u7684\u9650\u5b9a\uff0c\u5373 1978 \uf98e\u4ee5\uf92d\u51fa\u73fe\u7684\u65b0\u8a5e\u8a9e\u3002 \u6211\u5011\u8a8d\u5b9a\u7684\u65b0\u8a5e\u8a9e\u65e2\u6709\"\u65b0\uff02\u7684\u7279\u9ede\uff0c\u540c\u6642\u5f37\u8abf\uf9ba\u65b0\u8a5e\u8a9e\u7684\u4f7f\u7528\u7bc4\u570d\uff0c\u5373\u5fc5\u9808\u662f\u5728\u793e\u6703 \u6795\u982d\u98a8\u3001\u6492\u80e1\u6912\u9eb5\u3001\u4e0b\u6bdb\u6bdb\u96e8\u3001\u6ce1\u6cab\u7d93\u6fdf\u3001\u671d\u967d\u7523\u696d\u3001\u767d\u8272\u6d88\u8cbb\u3001\u4e0b\u6d77\u3001\u6488\u4eba\uff02 \u7b49\uff1bB\u3001\u501f\u4ee3\uff0c\u5982\uff1a\"\u83dc\u7c43\u5b50\u5de5\u7a0b\u3001\u767d\u9aee\u4e16\u754c\u3001\u767d\u689d\u6848\u3001\uf934\u4eba\u982d\uff02\u7b49\u3002C\u3001\u4eff \u8cc7\u8a0a\u96fb\u5b50\u8a5e\u5178\u4ee5\u53ca\u65b0\u8a5e\u8a9e\u7684\u69cb\u8a5e\u898f\uf9d8\u53ef\u4ee5\u76f4\u63a5\u61c9\u7528\u65bc\u4e2d\u6587\u8cc7\u8a0a\u8655\uf9e4\u7684\u672a\u767b\uf93f \u8a5e\u8a9e\uf9fc\u5225\uff0c\u6709\uf9dd\u65bc\u63d0\u9ad8\u4e2d\u6587\u8cc7\u8a0a\u8655\uf9e4\u6280\u8853\u7684\u6c34\u5e73\u3002 \u4e94\u500b\u5eab\u3002\u7e3d\u5eab\u4e00\u500b\uff0c\u8a9e\u6cd5\u8cc7\u8a0a\u5eab\u4e09\u500b(\u540d\u8a5e\u5eab\u3001\u52d5\u8a5e\u5eab\u3001\u5f62\u5bb9\u8a5e\u5eab)\uff0c\u69cb\u8a5e\u6cd5\u5eab\u4e00\u500b\u3002 4.2 \u65b0\u8a5e\u8a9e\u8a5e\u5178\u5c6c\u6027\u8cc7\u8a0a\u7684\u78ba\uf9f7 \u53e6\u5916\u9084\u8a2d\uf9f7\uf9ba\u820a\u8a5e\u5eab\u3001\u5916\uf92d\u8a5e\u5eab\u3001\u7c21\uf976\u8a5e\u5eab\u548c\u65b9\u8a00\u8a5e\u5eab\uff0c\u5c0d\u65b0\u8a5e\u8a9e\u7576\u4e2d\u7684\u820a\u8a5e\u65b0\u7528\u3001\u5916 \u751f\u6d3b\u4e2d\u5ee3\u6cdb\u4f7f\u7528\u7684\u8a9e\u6587\u6027\u8cea\u7684\u65b0\u8a5e\u8a9e\uff0c\u53ef\u4ee5\u9032\u5165\u666e\u901a\u8fad\u5f59\u7684\u65b0\u8a5e\u8a9e\uff0c\u90a3\u4e9b\u65b0\u51fa\u73fe\u7684\u5c08\u696d \u64ec\uff0c\u6bd4\u5982\uff1a\"\u7159\u6c11\u3001\u80a1\u6c11\u3001\u5f69\u6c11\u3001\u7db2\u6c11\uff02\uff0c\"\u7a7a\u59d0\u3001\u6d77\u59d0\u3001\u5427\u59d0\u3001\u547c\u59d0\u3001\u7db2\u59d0\u3001 \u65b0\u8a5e\u8a9e\u8a5e\u5178\u958b\u767c\u4e3b\u8981\u662f\u7232\uf9ba\u5b78\u7fd2\u3001\u7814\u7a76\u65b0\u8a5e\u8a9e\uff0c\u7279\u5225\u662f\u7232\u4e2d\u6587\u8cc7\u8a0a\u8655\uf9e4\u63d0\u4f9b\u4e00\u500b\u57fa\u672c\u8cc7 \uf92d\u8a5e\u3001\u7c21\uf976\u8a5e\u3001\u65b9\u8a00\u8a5e\u7b49\u7684\u6709\u95dc\u8cc7\u8a0a\u9032\ufa08\uf9ba\u63cf\u8ff0\u3002\u9019\u5e7e\u500b\u5eab\u901a\u904e\"\u8a5e\u8a9e\u3001\u62fc\u97f3\u3001\u7fa9\u9805\uff02 \u8853\u8a9e\u6c92\u6709\u589e\u52a0\u65b0\u7684\u666e\u901a\u8fad\u5f59\u610f\u7fa9\u7684\uff0c\uf967\u5728\u6211\u5011\u8a8d\u5b9a\u7684\u65b0\u8a5e\u8a9e\u7bc4\u570d\u5167\u3002\u6211\u5011\u8a8d\u5b9a\u7684\u65b0\u8a5e\u8a9e \u5177\u9ad4\u5982\u4e0b\uff1a \u7a7a\u5ac2\u3001\u6d77\u5ac2\u3001\u5427\u5a18\u3001\u547c\u5ac2\uff02\uff0c\"\u6587\u76f2\u3001\u79d1\u76f2\u3001\u80a1\u76f2\u3001\u821e\u76f2\u3001\u7db2\u76f2\uff02\u3001\"\u7db2\u6c11\u3001 \u7db2\u53cb\u3001\u7db2\u54e5\u3001\u7db2\u59d0\u3001\u7db2\u8ff7\u3001\u7db2\u87f2\u3001\u7db2\u8805\uff02\u7b49\u7b49\u3002 \u6e90\u3002\u7232\uf9ba\u9054\u5230\u9019\u4e00\u76ee\u7684\uff0c\u65b0\u8a5e\u8a9e\u8a5e\u5178\u5c6c\u6027\u8cc7\u8a0a\u5305\u62ec\uf9ba\u8a9e\u97f3\u8cc7\u8a0a\u3001\uf92d\u6e90\u8cc7\u8a0a\u3001\u8a9e\u6cd5\u8cc7\u8a0a\u548c \u4e09\u500b\uf91d\u4f4d\uf99a\u63a5\u3002\u69cb\u6210\u4e00\u500b\u4e0a\u4e0b\uf99a\u63a5\u7684\u6709\u6a5f\u7cfb\u7d71\uff0c\uf965\u65bc\u8cc7\u8a0a\u7684\u63d0\u53d6\u3002\u65b0\u8a5e\u8a9e\u8a5e\u5178\u7684\u7e3d\u9ad4\u7d50 3.3 \u300a\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u96fb\u5b50\u8a5e\u5178\u300b\u7684\u958b\u767c\u5177\u9ad4\u601d\uf937 \u90e8\u5206\u8a9e\u7fa9\u3001\u8a9e\u7528\u8cc7\u8a0a\uff0c\u6d89\u53ca\uf9ba\u65b0\u8a5e\u8a9e\u5f62\u3001\u97f3\u3001\u7fa9\u4ee5\u53ca\u7528\u6cd5\u7684\u4e3b\u8981\u65b9\u9762\u3002 \u69cb\u5982\u4e0b\uff1a</td></tr><tr><td>(1)\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u7684\u754c\u5b9a\uff0c(2)\u65b0\u8a5e\u8a9e\u8a5e\u5178\u7684\u958b\u767c\u601d\u60f3\uff0c(3)\u65b0\u8a5e\u8a9e\u7684 \u63a1\u96c6\u8207\u65b0\u8a5e\u8a9e\u5c6c\u6027\u8cc7\u8a0a\u7684\u63cf\u8ff0\uff0c(4)\u8fd1\u56db\u842c\u65b0\u8a5e\u8a9e\u7684\u6b78\uf9d0\u5be6\u8e10\u3002\u6211\u5011\u8a8d\u5b9a\u7684 \u65b0\u8a5e\u8a9e\u662f\u6307 1978 \uf98e\u4ee5\uf92d\u901a\u904e\u5404\u7a2e\u9014\u5f91\u7523\u751f\u7684\u3001\u5177\u6709\u57fa\u672c\u8a5e\u5f59\u6c92\u6709\u7684\u65b0\u5f62\u5f0f\u3001 \u65b0\u610f\u7fa9\u6216\u65b0\u7528\u6cd5\u7684\u8a9e\u6587\u8a5e\u8a9e\u3002\u9664\uf9ba\u8a5e\u5f62\u3001\u8a5e\u7fa9\u6216\u7528\u6cd5\u4efb\u4f55\u4e00\u500b\u65b9\u9762\"\u65b0\uff02\u5916\uff0c \u9084\u8981\u6c42\u5fc5\u9808\u662f\u4eba\u5011\u65e5\u5e38\u751f\u6d3b\u4e2d\u666e\u904d\u3001\u5ee3\u6cdb\u4f7f\u7528\u7684\u8a9e\u6587\u8a5e\u8a9e\uff0c\u4eba\u540d\u3001\u5730\u540d\u4ee5\u53ca\u5c08 \u79d1\u8853\u8a9e\u90fd\uf967\u5c6c\u65bc\u6211\u5011\u6240\uf96f\u7684\"\u65b0\u8a5e\u8a9e\uff02\u3002\u6211\u5011\u5805\u6301\u958b\u653e\u7684\u539f\u5247\uff0c\u5118\uf97e\u5168\u9762\u7684\u63a1 \u96c6\u6536\uf93f\u65b0\u8a5e\u8a9e\uff0c\u7528\u4eba\u6a5f\uf978\u7528\u7684\u7814\u7a76\uf9e4\uf9a3\uff0c\u4ee5\uf963\u4eac\u5927\u5b78\u8a08\u7b97\u8a9e\u8a00\u5b78\u7814\u7a76\u6240\u7684\u300a\u73fe \u4ee3\u6f22\u8a9e\u8a9e\u6cd5\u8cc7\u8a0a\u8a5e\u5178\u300b\u7232\u6a21\u578b\u6253\u9020\u4e00\u90e8\u6536\u8a5e\u5168\u9762\u3001\u8cc7\u8a0a\u8c50\u5bcc\u3001\u8cc7\u6e90\u9ad8\ufa01\u5171\u7528\u7684 \u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u96fb\u5b50\u8a5e\u5178\uff0c\u7232\u65b0\u8a5e\u8a9e\u7684\u7814\u7a76\u3001\u4e2d\u6587\u8cc7\u8a0a\u8655\uf9e4\u7684\u7814\u7a76\u63d0\u4f9b\u4e00\u500b\u5bf6 \u8cb4\u7684\u8cc7\u6e90\u3002\u76ee\u524d\u5df2\u6536\uf93f\u65b0\u8a5e\u8a9e\u8fd1 4 \u842c\uff0c\u9996\u5148\u6211\u5011\u6309\u7167\u73fe\u4ee3\u6f22\u8a9e\u8a5e\uf9d0\u7684\"\u512a\u52e2\u8a9e \u6cd5\uff02\u529f\u80fd\uff0c\u7d66\u9019\u56db\u842c\u65b0\u8a5e\u8a9e\u5206\uf9d0\u4e26\u6b78\uf9d0\uff0c\u7136\u5f8c\uff0c\uf9dd\u7528\u6210\u719f\u7684\u95dc\uf997\u8cc7\uf9be\u5eab(\u5728 ACCESS \u74b0\u5883\u4e0b\u5be6\u73fe)\u8a73\u7d30\u5730\u63cf\u8ff0\uf9ba\u6bcf\u500b\u8a5e\u8a9e\u7684\u5c6c\u6027\u8cc7\u8a0a\u3002\u8a2d\uf9f7\u7e3d\u5eab\u4e00\u500b\uff0c\u8a9e \u4e9b\u5eab\u7e3d\u5171\u8a2d\uf9f7\u5c6c\u6027\uf91d\u4f4d 200 \u591a\u500b\uff0c\u5305\u62ec\u6bcf\u500b\u8a5e\u8a9e\u7684\u8a9e\u97f3\u8cc7\u8a0a\u3001\u8a9e\u7fa9\u8cc7\u8a0a\u3001\uf92d\u6e90 1 \u672c\u9805\u7814\u7a76\u5f97\u5230\u4e2d\u570b\u570b\u5bb6\u54f2\u5b78\u793e\u6703\u79d1\u5b78\u898f\u5283\u5c08\u6848(01CYY002)\u652f\u63f4\uff1b \u672c\u6587\u65bc 2002 \uf98e 4 \u6708\u5728\u81fa\uf963\u8209\ufa08\u7684\"\u7b2c\u4e09\u5c46\u4e2d\u6587\u8fad\u5f59\u8a9e\u7fa9\u5b78\u6703\u8b70\uff02\u4e0a\u5ba3\uf95a\uff0c\u6703\u5f8c\u6839\u64da\u5c08\u5bb6\u7684\u610f\ufa0a \u4f5c\uf9ba\u4fee\u6539\uff0c\u8b39\u81f4\u8b1d\u5ff1\u3002 1. \u5f15\u8a00 2001 \uf98e\u6211\u5011\u7372\u5f97\uf9ba\u4e2d\u570b\u570b\u5bb6\u793e\u79d1\u898f\u5283\u5c08\u6848\"\u300a\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u8cc7\u8a0a\u96fb\u5b50\u8a5e\u5178\u300b\u7684\u958b\u767c\u8207 \u61c9\u7528\uff02(\u5c08\u6848\u7de8\u865f\uff1a01CYY002)\u3002\u4e00\uf98e\uf92d\uff0c\u6211\u5011\u5df2\u6309\u7167\u898f\u5283\u505a\uf9ba\u5927\uf97e\u7684\u5de5\u4f5c\uff0c\u5c08\u6848\u9032 \u57ce\u3001\u8df3\uf94c\u50f9\u3001\u5927\u51fa\u8840\u3001\u5a1b\uf914\u5708\u3001\u62cd\u62d6\u3001\u4e09\u7d1a\u7247\u3001\u4e3b\u6253\u3001\uf90a\u66f2\u3001\u52c1\u6b4c\u3001\u52c1\u821e\u3001\u641e \u7b11\u3001\u723d\u3001\u975a\u3001\u99ac\u5b50\u3001\u4e8c\u5976\u3001\u5957\u78c1\u3001\u78c1\u5be6\u3001\u8c93\u81a9\u3001\u8155\u5152\u3001\u6413\u3001\u508d\u5927\u6b3e\u3001\u4f83\u5927\u5c71\u3001 \u8180\u723a\uff02\u7b49\u3002 (4) \u5916\uf92d\u8a5e\uff0c\u5f9e\u5916\u65cf\u8a9e\u501f\uf92d\u7684\u8a5e\uff0c\u53c8\u6709\uff1aA\u3001\u97f3\u8b6f\u8a5e\u5982\"\u7684\u58eb\u3001\u5df4\u58eb\u3001\u6b50\u4f69\u514b\u3001\u53ef \u53e3\u53ef\uf914\u3001\u4e01\u514b\u3001\u514b\uf9dc\u3001\u57fa\u56e0\u3001\u8a17\u798f\u3001\u5361\uf925 OK\u3001\u62dc\u62dc\u3001\u9177(cool)\u3001\u853b(cute)\u3001\u79c0 \u6fdf\u3001\u6cd5\uf9d8\u3001\u8ecd\u4e8b\u3001\u6587\u5316\u3001\u79d1\u6280\u3001\u6559\u80b2\u3001\u885b\u751f\u3001\u9ad4\u80b2\u3001\u5546\u696d\u3001\u5de5\u696d\u3001\u8fb2\u696d\u3001\u751f\u6d3b\u3001 \u9996\u5148\uf9dd\u7528\u6211\u5011\u81ea\u5df1\u958b\u767c\u597d\u7684\u300a\u65b0\u8a5e\u8a9e\u8a5e\u5178\u8cc7\u8a0a\u5eab\u300b\u548c\u8a9e\uf9be\u5eab\u6574\uf9e4\u51fa\u4e00\u500b\u65b0\u8a5e\u8a9e\u8a5e\u8868\uff0c\u7136 \u63d0\u4f9b\u7684\u4fe1\u606f\uf97e\u6975\u5176\u6709\u9650\u3002\u7531\u65bc\u4ee5\u4e0a\u7684\uf967\u8db3\uff0c\u9020\u6210\u73fe\u6709\u7684\u5404\u7a2e\u65b0\u8a5e\u8a5e\u5178\u61c9\u7528\u50f9\u503c\uf967\u9ad8\u3002 (5) \u61c9\u7528\uf9b4\u57df\u3002\u61c9\u7528\uf9b4\u57df\u7684\u5283\u5206\u662f\u4e00\u500b\u6bd4\u8f03\u68d8\u624b\u7684\u554f\u984c\uff0c\u6211\u5011\u5927\u9ad4\u4e0a\u5206\u7232\u653f\u6cbb\u3001\u7d93 4.1 \u65b0\u8a5e\u8a9e\u7684\u63a1\u96c6 \u689d\u4ef6\u7684\u9650\u5236\uff0c\u5404\u7a2e\u8a5e\u5178\u6536\u8a5e\uf97e\u6709\u9650\uff0c\u8a5e\u8a9e\u7684\u89e3\u91cb\u53ca\u5f15\uf9b5\u90fd\u6709\u6b20\u59a5\u4e4b\u8655\uff0c\uf901\u91cd\u8981\u7684\u662f\u8a5e\u5178 \u751f\u65b0\u7fa9\u3002 \u4eba\u7528\u7684\uff0c\u800c\u6c92\u6709\u8003\u616e\u5230\u6a5f\u5668\u4f7f\u7528\uff0c\u61c9\u7528\u7bc4\u570d\u53d7\u5230\uf9ba\u9650\u5236\u3002(3)\u7531\u65bc\u53d7\u5230\u7814\u7a76\u6280\u8853\u548c\u7814\u7a76 \u7684\u8fad\u5f59\uff0c\u5916\uf92d\u8a5e\uff0c\u7c21\uf976\u8a5e\uff0c\u4fee\u8fad\u7528\u6cd5\u7a69\u5b9a\u4e0b\uf92d\u69cb\u6210\u65b0\u8a5e\uff0c\u8853\u8a9e\u64f4\u5927\u4f7f\u7528\u7bc4\u570d\u7523 \u679c\u90fd\u662f\u5370\u5237\u54c1\uff0c\u6c92\u6709\u6709\u6548\u7684\u96fb\u5b50\u7248\u6210\u679c\uff0c\uf967\u80fd\u5be6\u73fe\u8cc7\u6e90\u9ad8\ufa01\u5171\u7528\u3002(2)\u9019\u4e9b\u6210\u679c\u90fd\u662f\u7232 4. \u300a\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u8cc7\u8a0a\u96fb\u5b50\u8a5e\u5178\u300b\u8a5e\u8a9e\u7684\u63a1\u96c6\u8207\u6240\u63cf\u8ff0\u7684\u5c6c\u6027\u8cc7\u8a0a (4) \u7523\u751f\u9014\u5f91\u3002\u6839\u64da\u6211\u5011\u7684\u8003\u5bdf\u4e3b\u8981\u5305\u62ec\uff1a\u65b0\u9020\u8a5e\uff0c\u820a\u8a5e\u65b0\u7528\uff0c\u65b9\u8a00\u8a5e\u9032\u5165\u666e\u901a\u8a71 \u4f46\u662f\u9019\u4e9b\u8457\u4f5c\u5c0d\u65b0\u8a5e\u8a9e\u7684\u7814\u7a76\u90fd\u6709\u4e00\u5b9a\u7684\u5c40\u9650\u3002\u4e3b\u8981\u8868\u73fe\u5728\u4ee5\u4e0b\u65b9\u9762\uff1a(1)\u9019\u4e9b\u7814\u7a76\u6210 \uf91d\u4f4d\uf997\u7e6b\u8d77\uf92d\uff0c\u69cb\u6210\uf9ba\u4e00\u500b\u5177\u6709\u4e0a\u4e0b\u4f4d\u95dc\u4fc2\u7684\u6709\u6a5f\u7cfb\u7d71\uff0c\uf965\u65bc\u8cc7\u8a0a\u7684\u63d0\u53d6\u3002\u9019 (3) \u65b9\u8a00\u8fad\u5f59\u9032\u5165\u666e\u901a\u8a71\u8fad\u5f59\u3002\u5982\"\u7092\u9b77\u9b5a\u3001\u767c\u71d2\u53cb\u3001\u57cb\u55ae\u3001\u7684\u58eb\u3001\uf99a\u9396\u5e97\u3001\u670d\u88dd \u4e9b\u5f15\u4eba\u6ce8\u76ee\u7684\u7814\u7a76\u6210\u679c\u3002\u51fa\u7248\uf9ba\u65b0\u8a5e\u8a9e\u8a5e\u5178\u53ca\u8a5e\u8a9e\u96c6\u4e09\u5341\u591a\u7a2e\u3001\u65b0\u8a5e\u8a9e\u7814\u7a76\u5c08\u8457\uf978\u672c\uff0c \u8a5e\u8a9e\uff0c\u589e\u52a0\u65b0\u8a5e\u8a9e\u5c6c\u6027\u8cc7\u8a0a\u7684\u63cf\u8ff0\uff0c\u4ee5\u6eff\u8db3\u5be6\u969b\u9700\u8981\u3002 \u5b9a\u7684\u69cb\u8a5e\u65b9\u6cd5\u69cb\u6210\u7684\u65b0\u8a5e\u8a9e\u7684\u8a5e\u6027\u7684\u898f\uf9d8\u3002 \u8a5e\u5eab\u3001\u5916\uf92d\u8a5e\u5eab\u3001\u7c21\uf976\u8a5e\u5eab\u3002\u7e3d\u5eab\u548c\u5176\u4ed6\u5404\u5eab\u901a\u904e\"\u8a5e\u8a9e\u3001\u62fc\u97f3\u3001\u7fa9\u9805\uff02\u4e09\u500b \u9ad8\u4f4e\u90fd\u53ef\u4ee5\u7528\uff0c\u8b8a\u6210\uf9ba\u4e00\u7a2e\u666e\u901a\u7684\uf96f\u6cd5\u3002 \u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u7684\u7814\u7a76\u53d7\u5230\uf9ba\u570b\u5167\u5916\u7684\u5ee3\u6cdb\u95dc\u6ce8\uff0c\u5b78\u8005\u5011\u4e5f\u505a\uf9ba\u5927\uf97e\u7684\u7814\u7a76\uff0c\u7523\u751f\uf9ba\u4e00 \u5247\uff0c\u5c07\u8ddf\u8e64\u6f22\u8a9e\u8fad\u5f59\u7684\u767c\u5c55\u8b8a\u5316\u548c\u6f22\u8a9e\u8cc7\u8a0a\u8655\uf9e4\u7684\u767c\u5c55\uff0c\uf967\u65b7\u5730\u6536\u96c6\u3001\u589e\u52a0\u65b0 \u5206\u5206\u89e3\u958b\uf92d\uff0c\u5206\u5225\u6a19\u4e0a\u8a72\u8a9e\u7d20\u6240\u5c6c\u7684\"\u8a5e\u6027\uff02\uff0c\u4ee5\uf965\u9032\u4e00\u6b65\u8003\u5bdf\u7531\u8a9e\u7d20\u6309\u7167\u4e00 \u6cd5\u8cc7\u8a0a\u5eab\u4e09\u500b\uff0c\u5305\u62ec\u540d\u8a5e\u5eab\u3001\u52d5\u8a5e\u5eab\u3001\u5f62\u5bb9\u8a5e\u5eab\uff0c\u53e6\u5916\u9084\u8a2d\uf9f7\uf9ba\u69cb\u8a5e\u6cd5\u5eab\uff0c\u820a (1) \u65b0\u9020\u8a5e\u8a9e\u3002\u6bd4\u5982\"\u6253\u5047\u3001\u6276\u8ca7\u3001\u80a1\u76f2\u3001\u5c55\u92b7\u3001\u80a1\u5e02\u3001\u9ad8\u958b\u3001\u4f4e\u8d70\u3001\u54c7\u567b\u3001\u5f69\u7968\u3001 \u8db3\u5f69\u3001\u8fa3\u59b9\u3001\u9177\u88dd\u3001\u65b0\u65b0\u4eba\uf9d0\u3001\u54c8\u97d3\u65cf\u3001\u54c8\u65e5\u65cf\u3001\u77e5\u672c\u5bb6\u3001\u9ed1\u54e8\uff02\u7b49\u7b49\u3002 (2) \u820a\u8a5e\u65b0\u7528\u3002\u9019\uf9d0\u8a5e\u8a9e\u8a5e\u5f62\u662f\u539f\u6709\u7684\uff0c\"\u65b0\uff02\u4e3b\u8981\u8868\u73fe\u5728\u7523\u751f\uf9ba\u65b0\u610f\u7fa9\u6216\u6709\uf9ba\u65b0 \u7684\u904b\u7528\u3002\u5177\u9ad4\u5206\u7232\u4e09\u7a2e\u60c5\u6cc1\uff1aA\u3001\u539f\u6709\u7684\u8a5e\u8a9e\u589e\u52a0\uf9ba\u65b0\u7684\u610f\u7fa9\uff0c\u5982\"\u4e0b\u8ab2\u3001\u4e0a \u8ab2\u3001\u6c23\u5019\u3001\u8df3\u69fd\u3001\u8d77\u98db\u3001\u7d05\u5a18\u3001\u7a97\u53e3\u3001\u4e0b\u5d17\u3001\uf977\u76f8\u3001\u65b0\u767b\u5834\u3001\u8ddf\u9032\u3001\u5145\u96fb\u3001\u8f38 \u8840\u3001\u9020\u8840\uff02\u7b49\uff1bB\u3001\u539f\u6709\u7684\u8a5e\u8a9e\u6709\uf9ba\u65b0\u7684\u7528\u6cd5\u3002\u6bd4\u5982\"\u7d50\u69cb\uff02\u672c\uf92d\u662f\u540d\u8a5e\uff0c\u4f46 \u7528\u7232\u52d5\u8a5e\uff0c\u5982\uff1a\u4f60\u7232\u6211\u7d50\u69cb\u4eba\u751f\uff1b\"\u904b\u6c23\uff02\u539f\u7232\u540d\u8a5e\uff0c\u7528\u7232\u5f62\u5bb9\u8a5e\uff0c\u5982\uff1a\u4f60\u9019 \u4eba\u5f88\u904b\u6c23\u3002\"\u706b\uff02\u539f\u7232\u540d\u8a5e\uff0c\u7528\u7232\u5f62\u5bb9\u8a5e\uff0c\u5f62\u5bb9\u4e8b\u7269\u6216\u4eba\u6709\u8072\u52e2\uff0c\u53d7\u6b61\u8fce\u3002\u5982\uff1a \u7d44\u7e54\u8005\u5011\u771f\u6c92\u60f3\u5230\u665a\u6703\u7adf\u7136\u9019\u9ebd\"\u706b\uff02\u3002C\u3001\u539f\u6709\u7684\u8a5e\u8a9e\u5f88\u9577\u4e00\u6bb5\u6642\u9593\uf967\u7528\uff0c \u53c8\u91cd\u65b0\u5553\u7528\uff0c\u6bd4\u5982\uff1a\"\u9ad8\u5c31\u3001\u8cde\u5149\u3001\u9ed1\u9053\u3001\u7d81\u7968\u3001\u6495\u7968\u3001\u591c\u7e3d\u6703\u3001\u5c0f\u59d0\u3001\u592a\u592a\u3001 \uf90a\u5a5a\u3001\u9280\u5a5a\uff02\u7b49\u3002\u5176\u4e2d\u6709\u4e9b\u610f\u7fa9\u4e5f\u767c\u751f\uf9ba\u4e00\u4e9b\u8b8a\u5316\uff0c\u6bd4\u5982\"\u9ad8\u5c31\u3001\u8cde\u5149\u3001\u592a\u592a\u3001 \u5c0f\u59d0\uff02\u7b49\u539f\uf92d\u4e3b\u8981\u7528\u65bc\u5730\u4f4d\u6bd4\u8f03\u9ad8\u7684\u4eba\uff0c\u6709\u7279\u6307\u6027\uff0c\u73fe\u5728\u5df2\u7d93\u6cdb\u5316\uff0c\uf967\uf941\u5730\u4f4d (7) \u5c08\u7528\u8853\u8a9e\u610f\u7fa9\u6cdb\u5316\u3001\u8f49\u79fb\uff0c\u64f4\u5927\u4f7f\u7528\u7bc4\u570d\uff0c\u8f49\u7232\u666e\u901a\u8fad\u5f59\u3002\u5982\"\u8edf\u9ad4\u3001\u786c\u9ad4\u3001 \u4ecb\u65bc\u76ee\u524d\u6709\u95dc\u65b0\u8a5e\u8a9e\u7684\u7814\u7a76\u6bd4\u8f03\uf9b2\u6563\uff0c\u800c\u4e14\u65b0\u8a5e\u8a9e\u7684\u7814\u7a76\u53c8\u6709\u5341\u5206\u91cd\u8981\u7684\u4f5c\u7528\uff0c\u6211\u5011\u64ec \u65b0\u8a5e\u8a9e\u8a5e\u5178\u63cf\u8ff0\u7684\u4e3b\u8981\u5c6c\u6027\u8cc7\u8a0a\u5305\u62ec\u4ee5\u4e0b\u65b9\u9762\uff1a \u7e3d\u5eab \u5c0d\u65b0\u8a5e\u8a9e\u9032\ufa08\u5927\u898f\u6a21\u7684\u6bd4\u8f03\u5b8c\u5099\u7684\u7814\u7a76\u3002\u5177\u9ad4\u601d\uf937\u7232\uff1a \u5553\u52d5\u3001\u71b1\u8655\uf9e4\u3001\uf92e\u8655\uf9e4\u3001\u9ec3\u724c\u3001\u4e3b\u65cb\uf9d8\u3001\u5957\uf946\u3001\u89f8\u96fb\u3001\u653e\u96fb\uff02\u7b49\u3002 (1) \u8a5e\u7684\u5e38\u898f\u8cc7\u8a0a\u3002\u5305\u62ec\u8a5e\u7684\uf95a\u97f3\u3001\u7fa9\u9805\u3001\u97f3\u7bc0\u3001\uf9b5\uf906\u7b49\u3002 (1) \u5118\uf97e\u7aae\u76e1\u5730\u6536\u96c6\u73fe\u6709\u7684\u65b0\u8a5e\u8a9e\uff0c\u505a\u5230\u5168\u9762\u3001\u6e96\u78ba\u3002\u76ee\u524d\u5df2\u6536\uf93f\u65b0\u8a5e\u8a9e\u8fd1 4 \u842c\uff0c (8) \u5b57\u6bcd\u8a5e\u3002\u4e3b\u8981\u6709\u4e09\uf9d0\uff1aA\u3001\u7d14\u7cb9\u7684\u5b57\u6bcd\u8a5e\uff0c\u6574\u500b\u8a5e\u7531\u82f1\u6587\u5b57\u6bcd\u69cb\u6210\uff0c\u5982\"CT\u3001 \u6536\uf93f\uf9ba\u6211\u5011\u6240\u80fd\ufa0a\u5230\u7684\u6240\u6709\u65b0\u8a5e\u8a9e\u3002 (2) \u8a9e\u6cd5\u8cc7\u8a0a\u3002\u6309\u7167\uf963\u4eac\u5927\u5b78\u8a08\u7b97\u8a9e\u8a00\u5b78\u7814\u7a76\u6240\u7684\u300a\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u6cd5\u8cc7\u8a0a\u8a5e\u5178\u300b\u7684\u898f \u540d\u8a5e\u5eab \u52d5\u8a5e\u5eab \u5f62\u5bb9\u8a5e\u5eab \u69cb\u8a5e\u6cd5\u5eab \u820a\u8a5e\u5eab \u5916\uf92d\u8a5e\u5eab \u7c21\uf976\u8a5e\u5eab \u65b9\u8a00\u8a5e\u5eab IBM\u3001CIA\u3001TOFEL\u3001GRE\u3001CEO\u3001ATM\u3001CFO\u3001BBS\u3001CVD\u3001DVD\u3001VS\u3001IT\u3001 IN\u3001Q\u3001VIP\uff02\u7b49\u7b49\uff1bB\u3001\u5b57\u6bcd\u548c\u6f22\u5b57\u7684\u7d44\u5408\uff0c\u5982\"BP \u6a5f\u3001BP \u65cf\u3001CALL \u6a5f\u3001 E \u6642\u4ee3\u3001E \u4eba\uf9d0\u3001IT \u754c\u3001IT \u696d\u3001\u5920 IN\u3001VIP \u5361\u3001\u5f88 Q\uff02\u7b49\u7b49\uff1bC\u3001\uf969\u4f4d\u548c\u5b57\u6bcd (2) \u6309\u7167\u4eba\u6a5f\uf978\u7528\u7684\u7814\u7a76\uf9e4\uf9a3\uff0c\u6253\u9020\u4e00\u90e8\u9069\u5408\u65bc\"\u4eba\uf95a\uff02\u548c\"\u6a5f\uf95a\uff02\u7684\u96fb\u5b50\u8a5e\u5178\u3002 \u683c\u63cf\u5beb\u65b0\u8a5e\u8a9e\u7684\u8a9e\u6cd5\u8cc7\u8a0a\u3002\u8a5e\uf9d0\u9ad4\u7cfb\u6cbf\u7528\u300a\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u6cd5\u8cc7\u8a0a\u8a5e\u5178\u300b\u7684 18 \u500b \u57fa\u672c\uf9d0\uff0c\u518d\u52a0\u4e0a\u6210\u8a9e\u3001\u6163\u7528\u8a9e\u3002\u8a5e\uf9d0\u6a19\u8a18\u8207\u5176\u76f8\u540c\u3002\u5404\uf9d0\u8a5e\u8a9e\u6cd5\u5c6c\u6027\u7684\u8a2d\uf9f7\u5728 4.3.2 \u5404\u500b\u5eab\u6240\u63cf\u8ff0\u7684\u4e3b\u8981\u5c6c\u6027\u8cc7\u8a0a \u589e\u52a0\u8a5e\u5178\u7684\u4fe1\u606f\uf97e\uff0c\u64f4\u5927\u8a5e\u5178\u7684\u4f7f\u7528\u7bc4\u570d\uff0c\u63d0\u9ad8\u5176\u61c9\u7528\u50f9\u503c\u3002 \u300a\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u6cd5\u8cc7\u8a0a\u8a5e\u5178\u300b\u57fa\u790e\u4e0a\u6709\u6240\u6539\u52d5\uff0c\u4f7f\u5176\uf901\u52a0\u512a\u5316\u3002 \u7e3d\u5eab\u4e3b\u8981\u63cf\u8ff0\u7684\u8cc7\u8a0a\u6709\uff1a\u8a5e\u8a9e\u3001\u62fc\u97f3\u3001\u7fa9\u9805\u3001\u8a5e\u6027\u3001\u97f3\u7bc0\u3001\u7523\u751f\u9014\u5f91\u3001\uf9b4\u57df\u3001\u6642\u9593\u3001\uf92d \u7684\u7d44\u5408\uff0c\u5982\"3D\u30013C\u30013S\uff02\u7b49\u7b49\u3002 3. \u300a\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u8cc7\u8a0a\u96fb\u5b50\u8a5e\u5178\u300b\u7684\u958b\u767c\u601d\u60f3 (4) \u4e00\u90e8\u958b\u653e\u7684\u8a5e\u5178\u3002\u672c\u8a5e\u5178\u5728\u65b0\u8a5e\u8a9e\u7684\u6536\u96c6\u53ca\u5c6c\u6027\u7684\u63cf\u8ff0\u65b9\u9762\u5747\u5805\u6301\u958b\u653e\u7684\u539f \u7a2e\uf9d0\u578b\uff1a\"\u5b57\u9996+\u8a5e\u6839\uff02\u3001\"\u8a5e\u6839+\u5c3e\u78bc\uff02\u7b49\u3002\u5c0d\u65bc\u8907\u5408\u8a5e\u5c07\u69cb\u6210\u8907\u5408\u8a5e\u7684\u5e7e\u90e8 3.1 \u65b0\u8a5e\u8a9e\u7814\u7a76\u7684\u5c40\u9650 (3) \u4ee5\uf963\u4eac\u5927\u5b78\u8a08\u7b97\u8a9e\u8a00\u5b78\u7814\u7a76\u6240\u7684\u300a\u73fe\u4ee3\u6f22\u8a9e\u8a9e\u6cd5\u8cc7\u8a0a\u8a5e\u5178\u300b\u7232\u6a21\u578b\uff0c\u63a1\u7528\u5206\uf9d0 \u8907\u5408\u5f0f\u53c8\u5206\u7232\uf997\u5408\u5f0f\u3001\u504f\u6b63\u5f0f\u3001\u88dc\u5145\u5f0f\u3001\u52d5\u8cd3\u5f0f\u548c\u4e3b\u8b02\u5f0f\u7b49\u3002\u9644\u52a0\u5f0f\u53c8\u5206\u7232\uf978 \u4e8c\u7dad\u95dc\u4fc2\uff0c\u6210\u679c\u7232\u8cc7\uf9be\u5eab\u6587\u4ef6\u683c\u5f0f\u7684\u96fb\u5b50\u8a5e\u5178\u3002 \uf997\u7dbf\u8a5e\u53c8\u5206\u7232\u96d9\u8072\u3001\u758a\u97fb\u5176\u4ed6\u7b49\u3002\u5408\u6210\u8a5e\u53c8\u5206\u7232\u8907\u5408\u5f0f\u3001\u91cd\u758a\u5f0f\u3001\u9644\u52a0\u5f0f\u4e09\uf9d0\u3002 \u9032\ufa08\u8a73\u7d30\u63cf\u8ff0\u3002\u5177\u9ad4\u63a1\u7528\u6210\u719f\u7684\u95dc\uf997\u8cc7\uf9be\u5eab\u5f62\u5f0f\u63cf\u8ff0\u8a5e\u8a9e\u548c\u8a9e\u6cd5\u3001\u8a9e\u7fa9\u5c6c\u6027\u7684 \u7232\u55ae\u97f3\u55ae\u7d14\u8a5e\u3001\u591a\u97f3\u55ae\u7d14\u8a5e\u3002\u591a\u97f3\u55ae\u7d14\u8a5e\u53c8\u5206\u7232\uf997\u7dbf\u8a5e\u3001\u97f3\u8b6f\u8a5e\u548c\u758a\u97f3\u8a5e\u7b49\u3002 \u8207\u5c6c\u6027\u63cf\u8ff0\u76f8\u7d50\u5408\u7684\u65b9\u6cd5\uff0c\u5728\u7c97\u5206\u8a5e\uf9d0\u7684\u57fa\u790e\u4e0a\u5c0d\u6bcf\u500b\u8a5e\u8a9e\u8a9e\u6cd5\u8a9e\u7fa9\u5c6c\u6027\u8cc7\u8a0a (3) \u69cb\u8a5e\u6cd5\u8cc7\u8a0a\u3002\u69cb\u8a5e\u6cd5\u4e3b\u8981\u5206\u7232\u55ae\u7d14\u69cb\u8a5e\u6cd5\u548c\u5408\u6210\u69cb\u8a5e\u6cd5\uf978\uf9d0\u3002\u55ae\u7d14\u69cb\u8a5e\u6cd5\u53c8\u5206 \u6e90\u7b49\u3002</td></tr><tr><td>* \u5c71\u6771\u7159\u81fa\u5e2b\u7bc4\u5b78\u9662\u4e2d\u6587\u7cfb(264025) \u5c55\u9806\uf9dd\u3002\u672c\u6587\u5f9e\u56db\u500b\u65b9\u9762\u4ecb\u7d39\u300a\u73fe\u4ee3\u6f22\u8a9e\u65b0\u8a5e\u8a9e\u8cc7\u8a0a\u96fb\u5b50\u8a5e\u5178\u300b(\u4ee5\u4e0b\u7c21\u7a31\"\u65b0\u8a5e\u8a9e\u8a5e E-mail:kangsy46@sohu.com Tel:0535-6672439 (show)\u3001\u812b\u53e3\u79c0(talk show)\u3001\u8840\u62fc(shopping) \u3001\u6d3e\u5c0d(party) \u3001\u4f0a\u59b9\u5152 (E-mail) \uff02 \u5f8c\u6309\u7167\u6211\u5011\u7684\u6536\u8a5e\u539f\u5247--\u5168\u9762\u6027\u539f\u5247\u3001\u898f\u7bc4\u6027\u8207\u63cf\u5beb\u6027\u76f8\u7d50\u5408\u539f\u5247\u3001\u5fc5\u8981\u6027\u539f\u5247\u3001\u666e \u901a\u7528\u7b49\uff0c\u66ab\u6642\u4f5c\u7232\u5de5\u4f5c\u898f\u7bc4\uff0c\u4ee5\u5f8c\u518d\u9010\u6f38\u8abf\u6574\u3002</td></tr><tr><td>Shandong of China: Yantai Normal College -Chinese Language Dept. (264025)</td></tr></table>",
+                "text": ""
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O02/O02-2006.json

@@ -0,0 +1,281 @@
+{
+    "paper_id": "O02-2006",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:06:14.141245Z"
+    },
+    "title": "An Experiment on Knowledge Extraction from an Encyclopedia Based on Lexicon Semantics \u5b8b\u67d4 * \uff64\u8a31\u52c7 +",
+    "authors": [
+        {
+            "first": "Song",
+            "middle": [],
+            "last": "Rou",
+            "suffix": "",
+            "affiliation": {},
+            "email": "songrou@blcu.edu.cn"
+        },
+        {
+            "first": "Xu",
+            "middle": [],
+            "last": "Yong",
+            "suffix": "",
+            "affiliation": {},
+            "email": ""
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "The typical approaches to extracting text knowledge are sentential parsing and pattern matching. Theoretically, text knowledge extraction should be based on complete understanding, so the technology of sentential parsing is used in the field. However, the fragility of systems and highly ambiguous parse results are serious problems. On the other hand, by avoiding thorough parsing, pattern matching becomes highly efficient. However, different expressions of the same information will dramatically increase the number of patterns and nullify the simplicity of the approach.",
+    "pdf_parse": {
+        "paper_id": "O02-2006",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "The typical approaches to extracting text knowledge are sentential parsing and pattern matching. Theoretically, text knowledge extraction should be based on complete understanding, so the technology of sentential parsing is used in the field. However, the fragility of systems and highly ambiguous parse results are serious problems. On the other hand, by avoiding thorough parsing, pattern matching becomes highly efficient. However, different expressions of the same information will dramatically increase the number of patterns and nullify the simplicity of the approach.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "Parsing in Chinese encounters greater barriers than that in English does. Firstly, Chinese lacks morphology. For example, recognition of base-NP in Chinese is more difficult than that in English because its left boundary is hard to discern.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Secondly, there are many stream sentences in Chinese which lack subjects and cause parsing to fail. Finally, in Chinese, the absence of verbs is also pervasive. Sentential parsing centering on verbs, which is used with English, is not always successful with Chinese.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "We are engaged in research on knowledge extraction from the Electronic Chinese Great Encyclopedia. Our goal is to extract unstructured knowledge from it and to generate a well-structured database so as to provide information services to users. The pattern-matching approach is adopted.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "The experiment was divided into two steps: (1) classifying entries based on lexicon semantics; (2) establishing a formal system based on lexicon semantics and extracting knowledge by means of pattern matching.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Classification of entries is important because in the text of the entries of different categories there are different kinds of patterns expressing knowledge. Our experiment demonstrated that an entry of the encyclopedia can be classified precisely merely according to the characters in the entry and the words in the first sentence of the entry's text. Some specific categories, e.g., organization names and Chinese place names, can be classified satisfactorily merely according to the suffix of the entry, for suffixes are closely related with semantic categories in Chinese.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "The formal system designed for knowledge extraction consists of 4 kinds of meta knowledge: concepts, mapping, relations and rules, which reflect lexicon semantic attributes. The present experiment focused on the extraction of knowledge about various areas from the texts regarding administrative places of China (how large is a place or its subdivisions). The results of the experiment show that the design of the formal system is practical. It can accurately and completely denote various expressions of simple knowledge in a Chinese encyclopedia. However, when the focus of knowledge changes, e.g., from administrative areas to habits of animals, it is a labor-intensive task to renew the formal system. Therefore the study of auto or semi-auto generation of this kind of formal system is required. [Tsujii, J. 2000\uff0cHull, R. et al. 1999 and Soderland, S. G. 1996 ]\u8655\uf9e4\u7684\u662f\u82f1\u8a9e\u6216\u65e5\u8a9e \u6587\u672c\uff0c\u4e3b\u8981\u4f7f\u7528\u57fa\u65bc\u8a9e\uf906\u5206\u6790\u7684\u65b9\u6cd5\u3002\u5176\u4e2d [Hull, R. et al. 1999] \u7684\u5de5\u4f5c\u662f\u57fa\u65bc\u5927\u5bb9\uf97e\u7684\u77e5 \uf9fc\u5eab\uff0c\u63a1\u7528\u90e8\u5206\u5206\u6790\u3001\u8a9e\u7fa9\u89e3\u91cb\u548c\u63a8\uf9e4\u7684\u6b65\u9a5f\uff1b [Soderland, S. G. 1996 \u53ec\u56de\uf961% 100 100 100 100 100 100 100 100 \u6ce8\uff1a\"\u5340\"\uf9d0\u8981\u5f9e\u5f8c\u7db4\"\u5340\"\u4e2d\u53bb\u6389\u5f8c\u7db4\"\u81ea\u6cbb\u5340\"\u3001\"\u5730\u5340\"\u3001\"\u98a8\u666f\u5340\"\u3001\"\u98a8\u666f\u540d\u52dd\u5340\"\u3001 \"\u81ea\u7136\u4fdd\u8b77\u5340\"\u3001\"\u704c\u5340\"\u3002\u8a72\uf9d0\u7684 3 \u500b\u8aa4\uf9fc\u932f\u8aa4\u662f\"\u7696\u897f\u5c71\u5340\"\u3001\"\u7696\u5357\u5c71\u5340\"\u3001\"\u795e\u8fb2\u67b6\uf9f4\u5340\"\u3002 \u7c21\u55ae\u5730\u628a\"\u5c71\u5340\"\u7576\u4f5c\u5f8c\u7db4\u5f9e\"\u5340\"\uf9d0\u4e2d\u53bb\u6389\u662f\uf967\ufa08\u7684\uff0c\u56e0\u70ba\u4e0a\u6d77\u6709\"\u5bf6\u5c71\u5340\"\uff0c\uf963\u4eac\u66fe\u6709\"\u71d5 \u5c71\u5340\"\uff0c\u7b49\u7b49\u3002 \u6ce8\uff1a\u4ee5\"\u6cb3\"\u70ba\u6700\u5f8c\u4e00\u500b\u5b57\u4f46\uf967\u662f\u6cb3\uf9ca\u7684\u8a5e\u76ee\u662f\"\u4e09\u6cb3\"\uff1b\u4ee5\"\u6c34\"\u70ba\u6700\u5f8c\u4e00\u500b\u5b57\u4f46\uf967\u662f \u6cb3\uf9ca\u7684\u8a5e\u76ee\u662f\"\u4e2d\u570b\u7684\u5730\u8868\u6c34\"\u548c\"\u4e2d\u570b\u7684\u5730\u4e0b\u6c34\"\uff1b\uf94e\uf9fc\u7684\u6cb3\uf9ca\u662f\u4ee5\"\u5e03\"\u548c\"\u66f2\"\u70ba\u5f8c\u7db4\u7684\u897f \u85cf\u5730\u5340\u6cb3\uf9ca\uff1b\u4ee5\"\u6d77\"\u70ba\u6e56\u6cca\u8a5e\u76ee\u7684\u5f8c\u7db4\uff0c\u9700\u8981\u4eba\u70ba\u5730\u53bb\u6389\u6e24\u6d77\u3001\u9ec3\u6d77\u3001\u6771\u6d77\u548c\u5357\u6d77\uff0c\u4f46 \u4ecd\u7136\u6709\u4e00\u500b\u8aa4\uf9fc\uff1a\"\u4e2d\u570b\u7684\u8fd1\u6d77\"\uff1b\"\u5c71\uf9ab\"\uf9d0\u7684\u8aa4\uf9fc\u662f\"\u4e2d\u570b\u7684\u706b\u5c71\"\uff0c\uf94e\uf9fc\u662f\"\u795e\u8fb2\u9802\"\uff1b\"\u5cf6 \u5dbc\"\uf9d0\u8981\u5f9e\u5f8c\u7db4\"\u5cf6\"\u4e2d\u53bb\u6389\u5f8c\u7db4\"\u534a\u5cf6\"\uff1b\uf94e\uf9fc\u7684\u6e56\u6cca\u662f\"\u6708\uf977\u6ce1\"\u3001\"\u5927\u5e03\u8607\u6ce1\"\u548c\u4ee5\"\uf9fe\u5361\" \u70ba\u5f8c\u7db4\u7684\u897f\u85cf\u5730\u5340\u9e79\u6c34\u6e56\uff1b\u8aa4\uf9fc\u7684\u6c99\u6f20\u662f\"\u4e2d\u570b\u7684\u6c99\u6f20\"\uff0c\uf94e\uf9fc\u7684\u6c99\u6f20\u662f\"\u6bdb\u70cf\u7d20\u6c99\u5730\"\u3002 2.2 \u95dc\u65bc\u8a5e\u76ee\u5206\uf9d0\u65b9\u6cd5\u7684\u7d50\uf941 \u6211\u5011\u7684\u8a66\u9a57\uf96f\u660e\uff0c\u50c5\u6839\u64da\u8a5e\u76ee\u7684\u7528\u5b57\u69cb\u6210\u548c\u8a5e\u76ee\u91cb\u6587\u7684\u9996\uf906\u7528\u8a5e\uff0c\u5c31\u53ef\u4ee5\u5c0d\u65bc\u767e\u79d1\u8fad\u5178 \u8a5e\u76ee\u7684\u4e3b\u8981\u984c\u6750\uf9d0\u5225\u9032\ufa08\u5206\uf9d0\uff0c\u6e96\u78ba\uf961\u548c\u53ec\u56de\uf961\u53ef\u9054\u5230\u5be6\u7528\u8981\u6c42\u3002\u5c0d\u65bc\u67d0\u4e9b\uf9d0\u5225\uff0c\u6bd4\u5982 \u6a5f\u69cb\u540d\u548c\u4e2d\u570b\u5730\u540d\uff0c\u5247\u50c5\u4f7f\u7528\u8a5e\u76ee\u5f8c\u7db4\u5c31\u80fd\u9054\u5230\u76f8\u7576\u597d\u7684\uf9fc\u5225\u6548\u679c\uff0c\u5176\u539f\u56e0\u662f\u6f22\u8a9e\u5f8c\u7db4 \u6210\u5206\u8207\u8a9e\u7fa9\uf9d0\u5225\u7dca\u5bc6\u76f8\u95dc\u3002 3 \u767e\u79d1\u8fad\u5178\u91cb\u6587\u77e5\uf9fc\u63d0\u53d6\u5be6\uf9b5 3.1 \u4e00\u500b\u57fa\u65bc\u8a5e\u5f59\u8a9e\u7fa9\u5c6c\u6027\u7684\u5f62\u5f0f\u7cfb\u7d71 \u6211\u5011\u628a\u8655\uf9e4\u5c0d\u8c61\u9650\u5b9a\u70ba\ufa08\u6587\u898f\u7bc4\u7684\u767e\u79d1\u8fad\u5178\uff0c\u76ee\u524d\u53ea\u63d0\u53d6\u6bd4\u8f03\uf9e0\u65bc\u5f62\u5f0f\u5316\u7684\u4fe1\u606f\u3002\u6211\u5011 \u7684\u57fa\u672c\u601d\u60f3\u662f\uff1a\u5efa\uf9f7\u8d77\u4e00\u500b\u57fa\u65bc\u8a5e\u5f59\u8a9e\u7fa9\u7684\u5c6c\u6027\u548c\u95dc\u4fc2\u7684\u5f62\u5f0f\u7cfb\u7d71\uff0c\u5176\u4e2d\u7684\u5c6c\u6027\u548c\u95dc\u4fc2 \u540c\u6b32\u63d0\u53d6\u7684\u4fe1\u606f\u7dca\u5bc6\u76f8\u95dc\uff1b\u4f7f\u7528\u5c6c\u6027\u6a21\u5f0f\u5339\u914d\u7684\u65b9\u6cd5\u5728\u7dda\u6027\u8a5e\uf905\u4e2d\u63d0\u53d6\u4fe1\u606f\u3002 \u6211\u5011\u9996\u5148\u505a\u7684\u662f\u4e2d\u570b\ufa08\u653f\u5730\u540d\u8a5e\u76ee\u91cb\u6587\u4e2d\u9762\u7a4d\u4fe1\u606f\u7684\u63d0\u53d6\u3002 \u5927\u90e8\u5206\u9762\u7a4d\u4fe1\u606f\u7684\u8868\u8ff0\u4e2d\u6709\"\u9762\u7a4d\"\u4e8c\u5b57\uff0c\u4f46\u662f\u5728\u6210\uf905\u7684\uf96f\u660e\u4e2d\uff0c\u6709\uf96d\uf976\u7684\u60c5\u6cc1\uff1b\"\u586b \u6d77\"\u3001\"\u7a2e\u690d\"\u7b49\u52d5\u8a5e\u5e36\uf969\u8a5e\u548c\u9762\u7a4d\uf97e\u8a5e\u8868\u793a\u9762\u7a4d\u7684\u60c5\u6cc1\u4e0b\uff0c\u6709\u6642\u4e5f\uf967\u4f7f\u7528\"\u9762\u7a4d\"\u3002\u5982\u95dc\u65bc \u9999\u6e2f\u7684\u91cb\u6587\u4e2d\u6709\uff1a \uf9d3\u5730\u9762\u7a4d 1071.8 \u5e73\u65b9\u516c\uf9e9\u3002\u5176\u4e2d\u9999\u6e2f\u5cf6 75.6 \u5e73\u65b9\u516c\uf9e9\uff0c\u4e5d\uf9c4 11.1 \u5e73\u65b9\u516c\uf9e9,\"\u65b0\u754c\uff02 (\u5305\u62ec\u5927\u5dbc\u5c71\u5cf6\u7b49\u5468\u570d 230 \u591a\u5ea7\u5cf6\u5dbc)975.1 \u5e73\u65b9\u516c\uf9e9\uff0c\u53e6\u65b0\u586b\u571f\u5730 9.2 \u5e73\u65b9\u516c\uf9e9\u3002 \u6b64\u5916\uff0c\u4e2d\u570b\ufa08\u653f\u5730\u540d\u8a5e\u76ee\u7684\u91cb\u6587\u4e2d\uff0c\u6709 4 \u8655\"\u9762\u7a4d\"\u7684\u932f\u5225\u5b57\uff1a2 \u8655\u932f\u6210\"\u5357\u7a4d\"\uff0c1 \u8655 \u932f\u6210\"\u9762\u548c\"\uff0c1 \u8655\u932f\u6210\"\u9762\u53ea\"\u3002 \u4f5c\u70ba\u4fe1\u606f\u63d0\u53d6\u7684\u521d\u6b65\u7814\u7a76\uff0c\u6211\u5011\u53ea\u8003\u616e\u51fa\u73fe\"\u9762\u7a4d\"\u4e8c\u5b57\u6642\u7684\u60c5\u6cc1\u3002 \u5728 755 \u500b\u4e2d\u570b\ufa08\u653f\u5730\u540d\u8a5e\u76ee\u7684\u91cb\u6587\u4e2d\uff0c\"\u9762\u7a4d\"\u51fa\u73fe\uf9ba 1668 \u6b21\uff0c\u5176\u4e2d 38 \u500b\"\u5927\u9762\u7a4d\" \u548c 2 \u500b\"\u55ae\u4f4d\u9762\u7a4d\"\u7528\u4f5c\u4fee\u98fe\u6210\u5206\uff0c\u5982\"\u5f62\u6210\u773e\u591a\u7684\u9e7d\u6e56\u548c\u5927\u9762\u7a4d\u6cbc\u6fa4\"\u548c\"\u6a39\u6728\u7a2e\uf9d0\u591a,\u55ae\u4f4d \u9762\u7a4d\u84c4\u7a4d\uf97e\u9ad8\"\uff0c\u5176\u9918 1628 \u8655\"\u9762\u7a4d\"\u78ba\u5be6\u8868\u9054\u9762\u7a4d\u4fe1\u606f\u3002 \uf9dd\u7528\u9762\u5411\u8a9e\u8a00\u6559\u5b78\u7814\u7a76\u7684\u6587\u672c\u6aa2\uf96a\u5de5\u5177 CCRL \u4f5c\u70ba\u8f14\u52a9\u5de5\u5177,\u6211\u5011\u7528\u4eba\u5de5\u5206\u6790\u7814\u7a76 \uf9ba\u9019\u4e9b\"\u9762\u7a4d\"\u7684\u4e0a\u4e0b\u6587\u3002 \u8207\"\u9762\u7a4d\"\u76f8\u95dc\u4e14\u5e36\u6709\uf969\u503c\u7684\u4fe1\u606f\u53ef\u4ee5\u770b\u6210\u662f\u67d0\u4e9b\u95dc\u4fc2\uff1a \uf969\uf97e\u95dc\u4fc2\u3002\uf941\u5143\u70ba\u4e3b\u9ad4\u3001\uf969\u503c\u3001\ufa01\uf97e\u55ae\u4f4d\u3002\u5982\"\u6d77\u58c7\u5cf6\u9762\u7a4d 323 \u5e73\u65b9\u516c\uf9e9\"\uff0c\"\u6d77\u58c7\u5cf6\" \u70ba\u4e3b\u9ad4\uff0c\"323\"\u70ba\uf969\u503c\uff0c\"\u5e73\u65b9\u516c\uf9e9\"\u70ba\ufa01\uf97e\u55ae\u4f4d\u3002 \u6bd4\uf9b5\u95dc\u4fc2\u3002\uf941\u5143\u70ba\u5206\u5b50\u4e3b\u9ad4\u3001\u5206\u6bcd\u4e3b\u9ad4\u3001\u6bd4\uf9b5\uf969\u3002\u591a\u6bd4\uf9b5\u95dc\u4fc2\u5247\u6d89\u53ca\u591a\u500b\u6bd4\uf9b5\u4e3b\u9ad4 \u548c\u591a\u500b\u6bd4\uf9b5\uf969\u3002\u5982\"\u9752\u6d77\u2026\u2026\u5929\u7136\u8349\u5834\u9762\u7a4d\u7d04\u5360\u5168\uf96d\u571f\u5730\u7e3d\u9762\u7a4d\u7684 46.39\uff05\"\uff0c\"\u5929\u7136\u8349\u5834\" \u70ba\u5206\u5b50\u4e3b\u9ad4\uff0c\"\u5168\uf96d\u571f\u5730\"\u70ba\u5206\u6bcd\u4e3b\u9ad4\uff0c\"46.39\uff05\"\u70ba\u6bd4\uf9b5\uf969\u3002 \u8b8a\u5316\uf969\uf97e\u95dc\u4fc2\u3002\uf941\u5143\u70ba\u4e3b\u9ad4\u3001\u64f4\u7e2e\u6a19\u8a18\u3001\uf969\u503c\u3001\ufa01\uf97e\u55ae\u4f4d\u3002\u5982\"\u2026\u2026\u57ce\u5340\u9762\u7a4d\u64f4\u5927\uf9ba 15 \u5e73\u65b9\u516c\uf9e9\"\uff0c\"\u57ce\u5340\"\u70ba\u4e3b\u9ad4\uff0c\"\u64f4\"\u70ba\u64f4\u7e2e\u6a19\u8a18\uff0c\"15\"\u70ba\uf969\u503c\uff0c\"\u5e73\u65b9\u516c\uf9e9\"\u70ba\ufa01\uf97e\u55ae\u4f4d\u3002 \u8b8a\u5316\u6bd4\uf9b5\u95dc\u4fc2\u3002\uf941\u5143\u70ba\u4e3b\u9ad4\u3001\u64f4\u7e2e\u6a19\u8a18\u3001\u500d\uf969\u6216\u6bd4\u503c\u3002\u5982\"\u8cb4\u5dde\u2026\u2026\uf9fe\u5712\u9762\u7a4d\u8f03 50 \uf98e\u4ee3\u521d\u64f4\u5927 20 \u591a\u500d\"\uff0c\"\uf9fe\u5712\u9762\u7a4d\"\u70ba\u4e3b\u9ad4\uff0c\"\u64f4\"\u70ba\u64f4\u7e2e\u6a19\u8a18\uff0c\"20 \u591a\u500d\"\u70ba\u500d\uf969\u3002 \u8b8a\u5316\uf969\uf97e\u95dc\u4fc2\u548c\u8b8a\u5316\u6bd4\uf9b5\u95dc\u4fc2\u9084\u61c9\u7576\u6d89\u53ca\u8b8a\u5316\u524d\u6642\u9593\u548c\u8b8a\u5316\u5f8c\u6642\u9593\u3002\u8b8a\u5316\u524d\u6642\u9593\u5f80 \u5f80\u986f\u5f0f\u5730\u7d66\u51fa\uff0c\u8b8a\u5316\u5f8c\u6642\u9593\u6709\u6642\uf96d\uf976\uff0c\u5176\u5be6\u5c31\u662f\u767e\u79d1\u5168\u66f8\u8cc7\uf9be\u6536\u96c6\u7684\u6642\u9593\u3002\u5982\u4e0a\u9762\u6700\u5f8c \u4e00\uf9b5\uff0c\u8b8a\u5316\u524d\u6642\u9593\u662f\"50 \uf98e\u4ee3\u521d\"\uff0c\u8b8a\u5316\u5f8c\u6642\u9593\u70ba\u767e\u79d1\u5168\u66f8\u8cc7\uf9be\u6536\u96c6\u7684\u6642\u9593\uff0c\u6587\u4e2d\uf96d\uf976\u3002 \uf969\uf97e\u95dc\u4fc2\u548c\u6bd4\uf9b5\u95dc\u4fc2\u4e5f\u61c9\u7576\u6d89\u53ca\u6642\u9593\uff0c\u88ab\uf96d\uf976\u7684\u6642\u9593\u4e5f\u662f\u767e\u79d1\u5168\u66f8\u8cc7\uf9be\u6536\u96c6\u7684\u6642\u9593\u3002\u9019 \u4e9b\u95dc\u4fc2\u5f80\u5f80\u5e36\u6709\u4fee\u98fe\u6210\u5206\uff0c\u5982\"\u7d04 10 \u516c\u9803\"\uff0c\"\uf967\u5230 30%\"\uff0c\"5 \u500d\u4ee5\u4e0a\"\uff0c\"\u64f4\u5927\u81f3 23 \u5e73\u65b9 \u516c\uf9e9\"\u7b49\u3002\u9019\u4e9b\u4e5f\u61c9\u7576\u4f5c\u70ba\uf941\u5143\u52a0\u5165\u5230\u5404\u95dc\u4fc2\u4e2d\u3002 \u4fe1\u606f\u63d0\u53d6\u7684\u4efb\u52d9\u5c31\u662f\u78ba\u5b9a\u9019\u4e9b\u95dc\u4fc2\u4e2d\u7684\uf941\u5143\u5728\u6587\u672c\u4e2d\u6240\u6307\u7684\u5167\u5bb9\u3002\u5176\u4e2d\uff0c\uf969\u503c\u3001\u6bd4 \uf9b5\uf969\u3001\u500d\uf969\u6216\u6bd4\u503c\u3001\ufa01\uf97e\u55ae\u4f4d\u3001\u64f4\u7e2e\u6a19\u8a18\u3001\u4fee\u98fe\u6bd4\u8f03\u5bb9\uf9e0\u78ba\u5b9a\uff0c\u56e0\u70ba\u5b83\u5011\u5f62\u5f0f\u898f\u7bc4\uff0c\u4f4d \u7f6e\u6bd4\u8f03\u56fa\u5b9a\uff0c\u800c\u4e14\u5f8c\u4e09\u8005\u7684\u96c6\u5408\u57fa\u672c\u4e0a\u662f\u5c01\u9589\u7684\u3002\u6642\u9593\uf941\u5143\u4e5f\u6709\u5f62\u5f0f\u6a19\u8a18\uff0c\u5305\u62ec\"\u4e16\u7d00\"\u3001 \"\uf98e\u4ee3\"\u3001\"\uf98e\"\u3001\"\u6708\"\u7b49\uff0c\u8868\u793a\u671d\u4ee3\u6216\u4e8b\u4ef6\u7684\u8a5e\u8a9e\u5f8c\u9762\u52a0\u4e0a\"\u521d\"\u3001\"\u672b\"\u3001\"\u524d\"\u3001\"\u5f8c\"\u3001\"\u671f \u9593\"\u7b49\u6642\u9593\u65b9\u4f4d\u8a5e\u3002\u78ba\u5b9a\u6642\u9593\uf941\u5143\u7684\u4e3b\u8981\u56f0\u96e3\u5728\u65bc\u51fa\u73fe\u4f4d\u7f6e\uf967\u56fa\u5b9a\u3002\u6211\u5011\u7684\u7b56\uf976\u662f\u5f9e\u5176\u5b83 \uf941\u5143\u51fa\u73fe\u7684\u4f4d\u7f6e\u5f80\u524d\u770b 6 \u500b\u9017\u865f\u6216\uf906\u865f\uff0c\u627e\u5230\uf9ba\u6642\u9593\uf941\u5143\u7279\u5fb5\u5c31\u53ef\u4ee5\u63d0\u53d6\u51fa\uf92d\uff0c\u627e\uf967\u5230 \u5c31\u6b78\u7d50\u70ba\uf96d\uf976\uff0c\u5373\u6642\u9593\uf941\u5143\u662f\u767e\u79d1\u5168\u66f8\u8cc7\uf9be\u6536\u96c6\u7684\u6642\u9593\u3002 \u6700\u5927\u7684\u56f0\u96e3\u5728\u65bc\u5404\u7a2e\u9762\u7a4d\u4e3b\u9ad4\u7684\u78ba\u5b9a\u3002\u70ba\u6b64\uff0c\u6211\u5011\u5f9e\u5be6\uf9b5\u4e2d\u63d0\u53d6\uf9ba\u4e00\u500b\u57fa\u65bc\u8a5e\u5f59\u8a9e \u5b8b\u67d4\uff64 \u8a31\u52c7 \u7fa9\u5c6c\u6027\u7684\u5f62\u5f0f\u7cfb\u7d71\uff0c\u5b83\u7684\u5167\u5bb9\u5305\u62ec 4 \uf9d0\u5143\u77e5\uf9fc\uff1a \u6982\uf9a3\uff1a \ufa08\u653f\u5340\u5283",
+                "cite_spans": [
+                    {
+                        "start": 801,
+                        "end": 838,
+                        "text": "[Tsujii, J. 2000\uff0cHull, R. et al. 1999",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 839,
+                        "end": 864,
+                        "text": "and Soderland, S. G. 1996",
+                        "ref_id": null
+                    },
+                    {
+                        "start": 896,
+                        "end": 918,
+                        "text": "[Hull, R. et al. 1999]",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 953,
+                        "end": 975,
+                        "text": "[Soderland, S. G. 1996",
+                        "ref_id": null
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "\u81ea\u7136\u5730\uf9e4\u8a5e\u76ee\u6240\u7528\u5f8c\u7db4\u548c\uf9fc\u5225\u7d50\u679c\u5982\u4e0b\uff1a \uf9d0\u540d \u6cb3\uf9ca \u6e56\u6cca \u5c71\uf9ab * \u5c71 \u8108 \u5cf6 \u5dbc \u76c6 \u5730 \u6c99 \u6f20 \u5e73 \u539f \u9ad8 \u539f \u8349 \u539f \u4e18 \uf959 \u5408\u8a08 \u5f8c\u7db4 \u6c5f,\u6cb3 * , \u6eaa, \u6c34 * \u6e56, \u932f, \u6c60, \u6d77 * \u5c71, \u5cf0, \u5c71 \u8108 \u5cf6 * \u76c6 \u5730 \u6c99 \u6f20 * \u5e73 \u539f \u9ad8 \u539f \u8349 \u539f \u4e18 \uf959 \uf9ab",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "xq\uff0c\u5f80\u5f80\u662f\u7576\u524d\u8a5e\u76ee\u672c\u8eab\uff0c\u4e5f\u53ef\u80fd\u662f\u7576\u524d\u8a5e\u76ee\u6240\u4ee3\u8868\u7684\ufa08\u653f\u55ae\u4f4d\u7684\u4e0a\u7d1a\u55ae \u4f4d\u3002 \u8a5e\u76ee\u66ff\u4ee3\u8a5e td\uff0c\u5305\u62ec\"\uf96d\u5883\"\u3001\"\u5168\uf96d\"\u3001\"\u5e02\u5883\"\u3001\"\u5168\u5e02\"\u3001\"\u5340\u5883\"\u3001\"\u5168\u5340\"\u3001\"\u7e23\u5883\"\u3001 \"\u5168\u7e23\"\u3002 \ufa08\u653f\u5340\u5283\u7684\u5206\u90e8 fb\uff0c\u5305\u62ec\"\u5e02\u5340\"\u3001\"\u57ce\u5340\"\u3001\"\u90ca\u5340\"\u3001\"\u6d77\u57df\"\u3001\"\uf9d3\u57df\"\u3001\"\uf9d3\u5730\"\uff0c\u9084\u5305 \u62ec\u65b9\u4f4d\u5206\u90e8\u5982\"\u6771\u90e8\"\u3001\"\u897f\uf963\u90e8\"\u3002 \u5177\u6709\u9762\u7a4d\u5c6c\u6027\u7684\u540d\u8a5e\u6027\u8a5e\u8a9e mc\uff0c\u5305\u62ec\"\u8349\u539f\"\u3001\"\u5e73\u539f\"\u3001\"\u8015\u5730\"\u3001\"\u571f\u5730\"\u3001\"\uf9d3\u5730\"\u3001\"\u68ee \uf9f4\"\u3001\"\u8352\u5730\"\u3001\"\u8352\u5c71\"\u3001\"\u5580\u65af\u7279\u5730\u8c8c\"\u3001\"\uf9fe\u5712\"\u3001\"\u6851\u5712\"\u3001\"\u679c\u5712\"\u7b49\u3002(\u6ce8\uff1a\"\u5c71\uf9ab\"\u3001\"\u5c71 \u8108\"\u3001\"\u6cb3\uf9ca\"\u7b49\uf967\u5177\u6709\u9762\u7a4d\u5c6c\u6027\u3002) \u5177\u6709\u9762\u7a4d\u5c6c\u6027\u7684\u52d5\u8a5e\u6027\u8a5e\u8a9e dc\uff0c\u5305\u62ec\"\u7a2e\u690d\"\u3001\"\u64ad\u7a2e\"\u3001\"\u990a\u6b96\"\u3001\"\u6de1\u6c34\u990a\u6b96\"\u7b49\u3002 \u8207\u5177\u6709\u9762\u7a4d\u5c6c\u6027\u7684\u52d5\u8a5e\u95dc\uf997\u7684\u540d\u8a5e\u6027\u8a5e\u8a9e md\uff0c\u5982\u8207\u7a2e\u690d\u548c\u64ad\u7a2e\u95dc\uf997\u7684\u6709\"\u4f5c\u7269\"\u3001\"\u7d93 \u6fdf\u4f5c\u7269\"\u3001\"\uf97b\u98df\u4f5c\u7269\"\uff0c\u4ee5\u53ca\u5177\u9ad4\u7684\u4f5c\u7269\u540d\u7a31\"\u6c34\u7a3b\"\u3001\"\u5c0f\u9ea5\"\u3001\"\u68c9\u82b1\"\u3001\"\uf9fe\uf96e\"\u3001\"\u83f8\u8349\"\u3001 \"\u751c\u83dc\"\u3001\"\u6a61\u81a0\"\u7b49\uff1b\u8207\u990a\u6b96\u6709\u95dc\u7684\u6709\"\u9b5a\"\u3001\"\u8766\"\u7b49\u3002 \u5177\u6709\u9762\u7a4d\u5c6c\u6027\u7684\u5c08\u540d zm\uff0c\u5176\uf9d0\u578b\u5305\u62ec\"\u8fb2\u5834\"\u3001\"\uf9f4\u5834\"\u3001\"\u98a8\u666f\u5340\"\u3001\"\u81ea\u7136\u4fdd\u8b77\u5340\"\u4ee5 \u53ca\u5404\u7a2e\u5efa\u7bc9\u7269\u7b49\u3002 \ufa08\u653f\u5340\u5283\uf9d0\u578b xl\uff0c\u5305\u62ec\"\uf96d\"\u3001\"\u81ea\u6cbb\u5340\"\u3001\"\u5e02\"\u3001\"\u5730\u5340\"\u3001\"\u81ea\u6cbb\u5dde\"\u3001\"\u5340\"\u3001\"\u7e23\"\u3001\"\u93ae\"\u3002 \u6620\u5c04\uff1a {td\u2192xq}\uff0c\u7531\u8a5e\u76ee\u66ff\u4ee3\u8a5e\u5230\u8a5e\u76ee\u672c\u8eab\uff0c\u5982\u5728\"\u6c5f\u8607\uf96d\"\u91cb\u6587\u4e2d\uff0c\"\u5168\uf96d\"\u6620\u5c04\u70ba\"\u6c5f\u8607\uf96d\"\u3002 {xq\u2192xl}\uff0c\u7531\ufa08\u653f\u5340\u5283\u540d\u5230\u5b83\u672c\u8eab\u7684\ufa08\u653f\u5340\u5283\uf9d0\u578b\uff0c\u5982\u7531\"\u6c5f\u8607\uf96d\"\u6620\u5c04\u70ba\"\uf96d\"\u3002 {xq\u2192xq}\uff0c\u7531\ufa08\u653f\u5340\u5283\u540d\u5230\u5b83\u7684\u4e0a\u7d1a\ufa08\u653f\u5340\u5283\u540d\uff0c\u5982\u7531\"\u8607\u5dde\u5e02\"\u6620\u5c04\u70ba\"\u6c5f\u8607\uf96d\"\u3002 {md\u2192dc}\uff0c\u7531\u540d\u8a5e\u5230\u8207\u5b83\u95dc\uf997\u7684\u5177\u6709\u9762\u7a4d\u5c6c\u6027\u7684\u52d5\u8a5e\uff0c\u5982\u7531\"\u68c9\u82b1\"\u6620\u5c04\u70ba\"\u7a2e\u690d\"\u3002 {mc\u2192mc}\uff0c\u7531\u540d\u8a5e\u5230\u5b83\u7684\u4e0a\u7d1a\u8a9e\u7fa9\u540d\u8a5e\uff0c\u5982\u7531\"\uf97b\u98df\u4f5c\u7269\"\u6620\u5c04\u70ba\"\u4f5c\u7269\"\u3002 \u95dc\u4fc2\uff1a \uf969\uf97e\u95dc\u4fc2\uff1asl(time, body, number, area-unit, modifier)\uff0c\u5373\u6642\u9593\u3001\u4e3b\u9ad4\u3001\uf969\u503c\u3001\u9762\u7a4d\u55ae \u4f4d\u3001\u4fee\u98fe\u6210\u5206\u6eff\u8db3\uf969\uf97e\u95dc\u4fc2\u3002 \u6bd4 \uf9b5 \u95dc \u4fc2 \uff1a bl(time, body-numerator, body-denominator, ratio, modifier-before, modifier-after)\uff0c\u5373\u6642\u9593\u3001\u5206\u5b50\u4e3b\u9ad4\u3001\u5206\u6bcd\u4e3b\u9ad4\u3001\u6bd4\u503c\u3001\u524d\u4fee\u98fe\u6210\u5206\u3001\u5f8c\u4fee\u98fe\u6210\u5206\u6eff\u8db3\u6bd4 \uf9b5\u95dc\u4fc2\u3002 \u8b8a \u5316\uf969 \uf97e\u95dc\u4fc2 \uff1absl(time-before,time-after, body, extend-reduce, number, area-unit, mordify)\uff0c\u5373\u8b8a\u5316\u524d\u6642\u9593\u3001\u8b8a\u5316\u5f8c\u6642\u9593\u3001\u4e3b\u9ad4\u3001\u64f4\u7e2e\u6a19\u8a18\u3001\uf969\u503c\u3001\u9762\u7a4d\u55ae\u4f4d\u3001\u4fee\u98fe\u6210\u5206\u6eff \u8db3\u8b8a\u5316\uf969\uf97e\u95dc\u4fc2\u3002 \u8b8a\u5316\u6bd4\uf9b5\u95dc\u4fc2\uff1abbl(time-before,time-after, body, extend-reduce, ratio, mordify) \uff0c\u5373\u8b8a \u5316\u524d\u6642\u9593\u3001\u8b8a\u5316\u5f8c\u6642\u9593\u3001\u4e3b\u9ad4\u3001\u64f4\u7e2e\u6a19\u8a18\u3001\u6bd4\u503c\u3001\u4fee\u98fe\u6210\u5206\u6eff\u8db3\u8b8a\u5316\u6bd4\uf9b5\u95dc\u4fc2\u3002 \u9019\u500b\u95dc\u4fc2\u7684 5 \u500b\uf969\u64da\"20 \u4e16\u7d00 50 \uf98e\u4ee3\u4ee5\u524d\"\u3001\"\u5b89\u9806\u5e02\u57ce\u5340\"\u3001\"1.4\"\u3001\"\u5e73\u65b9\u516c\uf9e9\"\u3001\"\u50c5\" \u5206\u5225\u5b58\u653e\u5728 sl \uf969\u64da\u5eab\u7684 5 \u500b\u5b57\u6bb5 time\u3001body\u3001number\u3001area-unit\u3001modifier \u4e0b\uff0c\u8868\u793a\u5728 \u9019\u500b\u95dc\u4fc2\u7684 6 \u500b\uf969\u64da\"nil\"\u3001\"\u5b89\u9054\u5e02\u5e02\u5883\u8349\u539f\"\u3001\"\u5b89\u9054\u5e02\u5e02\u5883\"\u3001\"51.5\uff05\"\u3001\"nil\"\u3001\"\u4ee5 \u4e0a\"\u5206\u5225\u5b58\u653e\u5728 bl \uf969\u64da\u5eab\u7684 6 \u500b\u5b57\u6bb5 time\u3001body-numerator\u3001body-denominator\u3001ratio\u3001 (2) \u70ba\uf9ba\u69cb\u9020\u9019\u4e00\u5f62\u5f0f\u7cfb\u7d71\u9700\u8981\u505a\u5927\uf97e\u7684\u4eba\u5de5\u8abf\u67e5\u3001\u5206\u6790\u3001\u6a19\u6ce8\u5de5\u4f5c\u3002\uf974\u4fe1\u606f \u63d0\u53d6\u7684\u7126\u9ede\uf967\u8b8a\u800c\u50c5\u50c5\u63db\u6389\u6587\u672c (\u628a\u4e2d\u570b\u5927\u767e\u79d1\u5168\u66f8\u63db\u6210\u5176\u4ed6\u767e\u79d1\u8fad\u5178\u6587",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Generic NLP Technologies: Language, Knowledge and Information Extraction",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Tsujii",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2000,
+                "venue": "Proc. of ACL2000",
+                "volume": "",
+                "issue": "",
+                "pages": "11--18",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Tsujii, J., \"Generic NLP Technologies: Language, Knowledge and Information Extraction\", Proc. of ACL2000, 2000, pp.11-18.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Automatic acquisition of biographic knowledge from encyclopedic texts",
+                "authors": [
+                    {
+                        "first": "R",
+                        "middle": [],
+                        "last": "Hull",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Gomez",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "Expert Systems with Applications",
+                "volume": "16",
+                "issue": "",
+                "pages": "261--270",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Hull, R., and Gomez, F., \"Automatic acquisition of biographic knowledge from encyclopedic texts\", Expert Systems with Applications 16(1999), pp.261-270.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "CRYSTAL: Inducing a Conceptual Dictionary",
+                "authors": [
+                    {
+                        "first": "W",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Soderland",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Fisher",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [],
+                        "last": "Aseltine",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Lehnert",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1995,
+                "venue": "Proc. of the International Joint Conference on Artificial Intelligence",
+                "volume": "",
+                "issue": "",
+                "pages": "1314--1319",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Soderland, W. D., Fisher, J. Aseltine, and W. Lehnert, \"CRYSTAL: Inducing a Conceptual Dictionary\", Proc. of the International Joint Conference on Artificial Intelligence, Montreal, Canada, 1995, pp. 1314-1319.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "Biological Knowledge Acquisition From the Electronic Encyclopedia of China",
+                "authors": [
+                    {
+                        "first": "F",
+                        "middle": [],
+                        "last": "Gu",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "Cao",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2001,
+                "venue": "Proc. of ICYCS",
+                "volume": "",
+                "issue": "",
+                "pages": "1199--1203",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Gu, F. and Cao, C., \"Biological Knowledge Acquisition From the Electronic Encyclopedia of China\", Proc. of ICYCS'2001, 2001, pp.1199-1203.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "TABREF3": {
+                "text": "\u500b\uf969\u64da\"nil\uff02\u3001\"\u963f\u514b\u8607\u5e02\uff02\u3001\"1.83 \u842c\uff02\u3001\"\u5e73\u65b9\u516c\uf9e9\uff02\u3001\"nil\uff02 \u5206\u5225\u5b58\u653e\u5728 sl \uf969\u64da\u5eab\u7684 5 \u500b\u5b57\u6bb5 time\u3001body\u3001number\u3001area-unit\u3001modifier \u4e0b\uff0c\u8868\u793a\u5728 \u8a72\u767e\u79d1\u8fad\u5178\u7de8\u5236\u6642\u963f\u514b\u8607\u5e02\u9762\u7a4d\u6070\u70ba 1.83 \u842c\u5e73\u65b9\u516c\uf9e9\u3002 dot-comma time string fb \u9762\u7a4d [modifier] [\u70ba] number area-unit \u2192sl(time, xq td, number, area-unit, modifier) \uf9b5\u5982\uff1a\"\u5b89\u9806\u5e02\"\u91cb\u6587\u4e2d\u6709\uff1a 20 \u4e16\u7d00 50 \uf98e\u4ee3\u4ee5\u524d\uff0c\u57ce\u5340\u9762\u7a4d\u50c5 1.4 \u5e73\u65b9\u516c\uf9e9\uff0c\u4eba\u53e3 2.4 \u842c\u4eba\u3002 \u5339\u914d\u898f\u5247\u7684\u689d\u4ef6\u90e8\u5206\u5f8c\uff0c\u5f97\u5230\u7684\uf969\uf97e\u95dc\u4fc2\u662f\uff1a sl(20 \u4e16\u7d00 50 \uf98e\u4ee3\u4ee5\u524d\uff0c\u5b89\u9806\u5e02\u57ce\u5340\uff0c1.4\uff0c\u5e73\u65b9\u516c\uf9e9\uff0c\u50c5)",
+                "num": null,
+                "content": "<table><tr><td>\u5176\u4e2d\u9762\u7a4d\u4e3b\u9ad4 body \u7684\u69cb\u6210\u65b9\u5f0f\u70ba\uff1a</td></tr><tr><td>xq [fb[fb]] [(zm | mc | md {md\u2192dc}md)]</td></tr><tr><td>\u5f0f\u4e2d ( | ) \u8868\u793a\u9078\u64c7\uff0c[ ] \u8868\u793a\u53ef\u6709\u53ef\u7121\u3002</td></tr><tr><td>\u898f\u5247\uff1a</td></tr><tr><td>\u898f\u5247\u7684\u4f5c\u7528\u5c31\u662f\u5f9e\u6587\u672c\u4e2d\u7684\u9069\u7576\u4f4d\u7f6e\u62bd\u53d6\u95dc\u4fc2\u4e2d\uf941\u5143\u6240\u6307\u7684\u5167\u5bb9\u3002\u898f\u5247\u7684\u5f62\u5f0f\u662f\uff1a</td></tr><tr><td>\u6587\u672c\u6a21\u5f0f\u2192\u95dc\u4fc2\uff0c\u5176\u4e2d\u6587\u672c\u6a21\u5f0f\uf99c\u51fa\u95dc\u4fc2\u4e2d\u5404\uf941\u5143\u6240\u6307\u5167\u5bb9\u5728\u6587\u672c\u4e2d\u7684\u76f8\u5c0d\u65bc\"\u9762\u7a4d\"\u7684</td></tr><tr><td>\u4f4d\u7f6e\u3002\u540c\u4e00\u500b\u898f\u5247\u4e2d\u7684\u540c\u4e00\u500b\u8b8a\u5143\uf974\u91cd\u8907\u51fa\u73fe\uff0c\u5247\u4ee3\u8868\u540c\u4e00\u500b\u5167\u5bb9\u3002</td></tr><tr><td>\u4e0b\u9762\uf99c\u51fa\u4e00\u4e9b\u5e38\u7528\u7684\u898f\u5247\u3002\u5176\u4e2d\uff0ctext-begin \u8868\u793a\u7bc7\u9996\uff0cdot-comma \u8868\u793a\uf906\u865f\u6216\u9017\u865f\uff0c</td></tr><tr><td>no-area-string \u8868\u793a\u4e00\u500b\uf906\uf905\uff0c\u5176\u4e2d\uf967\u51fa\u73fe\"\u9762\u7a4d\"\u3002string \u8868\u793a\u4e00\u500b\uf906\uf905\uff0c\u5176\u4e2d\u6bcf\u500b\u6a19\u9ede\uf906</td></tr><tr><td>\u7684\u9996\u8a5e\uf967\u5e36\u6709 xq\u3001fb\u3001mc\u3001md\u3001zm\u3001time \u5c6c\u6027\uff0c\u800c\u4e14\uf906\uf905\u4e2d\u5305\u542b\u7684\u6a19\u9ede\uf906\uf967\u8d85\u904e 6 \uf906\u3002</td></tr><tr><td>\u6211\u5011\u9019\u88cf\u6240\uf96f\u7684\uf906\uf905\u5c31\u662f\u4e00\uf905\u6a19\u9ede\uf906\uff0c\u800c\u6a19\u9ede\uf906\u5c31\u662f\u6587\u672c\u4e2d\u4ee5\u9017\u865f\u3001\uf906\u865f\u3001\u5206\u865f\u3001\u5606\u865f\u3001</td></tr><tr><td>\u554f\u865f\u5206\u9694\u7684\u5b57\uf905\u3002</td></tr><tr><td>text-begin no-area-string dot-comma [\u7e3d] \u9762\u7a4d [modifier] [\u70ba] number area-unit</td></tr><tr><td>\u2192sl(nil, xq, number, area-unit, modifier)</td></tr><tr><td>\uf9b5\u5982\uff1a\"\u963f\u514b\u8607\u5e02\"\u91cb\u6587\u7684\u958b\u59cb\u5e7e\uf906\u662f\uff1a</td></tr><tr><td>\u65b0\u7586\u963f\u514b\u8607\u5730\u5340\u8f44\u5e02\u548c\ufa08\u7f72\u99d0\u5730,\u65b0\u7586\u91cd\u9ede\u58be\u5340\u3002\u4f4d\u65bc\u5854\uf9e9\u6728\u76c6\u5730\u897f\uf963\u90e8\u3002\u9762\u7a4d 1.83</td></tr><tr><td>\u842c\u5e73\u65b9\u516c\uf9e9,\u4eba\u53e3 38.13 \u842c\u3002</td></tr><tr><td>\u5339\u914d\u898f\u5247\u7684\u689d\u4ef6\u90e8\u5206\u5f8c\uff0c\u5f97\u5230\u7684\uf969\uf97e\u95dc\u4fc2\u662f\uff1a</td></tr><tr><td>sl(nil\uff0c\u963f\u514b\u8607\u5e02\uff0c1.83 \u842c\uff0c\u5e73\u65b9\u516c\uf9e9\uff0cnil)</td></tr><tr><td>\u9019\u500b\u95dc\u4fc2\u7684 5</td></tr></table>",
+                "type_str": "table",
+                "html": null
+            },
+            "TABREF4": {
+                "text": "20 \u4e16\u7d00 50 \uf98e\u4ee3\u4ee5\u524d\u5b89\u9806\u5e02\u57ce\u5340\u9762\u7a4d\u50c5 1.4 \u5e73\u65b9\u516c\uf9e9\u3002 dot-comma td string mc \u9762 \u7a4d [modifer-before] \u5360 [td][[ \u7e3d ] \u9762 \u7a4d ] [ \u7684 ] ratio [modifier-after] \u2192bl(nil, {td\u2192xq}td mc, {td\u2192xq}td , ratio, modifier-before\uff0cmodifier-after) \uf9b5\u5982\uff1a\"\u5b89\u9054\u5e02\"\u91cb\u6587\u4e2d\u6709\uff1a \u5e02\u5883\u5730\u5f62\u5e73\u5766,\u5e73\u5747\u6d77\u62d4 150 \u7c73\u3002\u8349\u539f\u9762\u7a4d\u5360 51.5\uff05\u4ee5\u4e0a\uff0c\u5b9c\u767c\u5c55\u755c\u7267\u3002",
+                "num": null,
+                "content": "<table><tr><td>\u5339\u914d\u898f\u5247\u7684\u689d\u4ef6\u90e8\u5206\u5f8c\uff0c\u5f97\u5230\u7684\u6bd4\uf9b5\u95dc\u4fc2\u662f\uff1a</td></tr><tr><td>bl(nil\uff0c\u5b89\u9054\u5e02\u5e02\u5883\u8349\u539f\uff0c\u5b89\u9054\u5e02\u5e02\u5883\uff0c51.5\uff05\uff0cnil\uff0c\u4ee5\u4e0a)</td></tr></table>",
+                "type_str": "table",
+                "html": null
+            },
+            "TABREF5": {
+                "text": "\u689d\u898f\u5247\u4f7f\u7528 94 \u6b21\uff0c\u7b2c 2 \u689d\u898f\u5247\u4f7f\u7528 22 \u6b21\uff0c\u5168\u90e8\u6b63\u78ba\u3002\u7279\u5225\u662f\uff0c 107 \u500b\u8a5e\u76ee\u91cb\u6587\u4e2d\uff0c\u6709 103 \u500b\u63d0\u5230\uf9ba\u8a72\u8a5e\u76ee\u6240\u4ee3\u8868\u7684\ufa08\u653f\u5340\u5283\u7684\u9762\u7a4d\uff0c\u5b83\u5011\u90fd\u51fa\u73fe\u5728\u9760 \u8fd1\u7bc7\u9996\u7684\u4f4d\u7f6e\uff0c\u5176\u4e2d 102 \u689d\u53ef\u4ee5\u7528\u898f\u5247\u5c07\u9762\u7a4d\u4fe1\u606f\u63d0\u53d6\u51fa\uf92d\uff0c94 \u689d\u7528\u4e0a\u8ff0\u7b2c 1 \u689d\u898f\u5247\uff0c",
+                "num": null,
+                "content": "<table><tr><td>modifier-before\uff0cmodifier-after \u4e0b\uff0c\u8868\u793a\u5728\u8a72\u767e\u79d1\u8fad\u5178\u7de8\u5236\u6642\u5b89\u9054\u5e02\u5e02\u5883\u8349\u539f\u9762\u7a4d\u5360\u5b89\u9054</td></tr><tr><td>\u5e02\u5e02\u5883\u9762\u7a4d 51.5\uff05\u4ee5\u4e0a\u3002</td></tr><tr><td>\u7531\u65bc\u9019\u4e9b\u88ab\u63d0\u53d6\u51fa\uf92d\u7684\u4fe1\u606f\u4ee5\u95dc\u4fc2\uf969\u64da\u5eab\u7684\u5f62\u5f0f\u5b58\u653e\uff0c\u6240\u4ee5\u53ef\u4ee5\u501f\u52a9\uf969\u64da\u5eab\u6aa2\uf96a\u5de5</td></tr><tr><td>\u5177\uf92d\u6aa2\uf96a\u3002</td></tr><tr><td>4. \u6e2c\u8a66\u8207\u8a0e\uf941</td></tr><tr><td>\u6211\u5011\u6aa2\u67e5\uf9ba\u4e2d\u570b\ufa08\u653f\u5730\u540d\u8a5e\u76ee\u6309\u6f22\u8a9e\u62fc\u97f3\u6392\u5e8f a-d \u7684 107 \u500b\u8a5e\u76ee\u7684\u91cb\u6587\uff0c\u9019\u88cf\u9762\u51fa\u73fe\"\u9762</td></tr><tr><td>\u7a4d\"176 \u6b21\uff0c\u5e36\u6709\uf969\uf97e\u7684\u9762\u7a4d\u4fe1\u606f 153 \u689d\uff0c\u5176\u4e2d\u6709\u4e9b\u662f\ufa08\u653f\u5340\u5283\u672c\u8eab\u7684\u9762\u7a4d\uff0c\u6709\u4e9b\u662f\ufa08\u653f</td></tr><tr><td>\u5340\u5283\u5167\u90e8\u67d0\u500b\u5206\u5340\u7684\u9762\u7a4d\u3002\u4f7f\u7528\u8a72\u7cfb\u7d71\u7684\u898f\u5247\u80fd\u5920\u6b63\u78ba\u63d0\u53d6\u4fe1\u606f\u7684 141 \u689d\uff0c\u6e96\u78ba\uf961\u7d04\u70ba</td></tr><tr><td>92%\u3002\u5176\u4e2d\uff0c\u4e0a\u8ff0\u7b2c 1 8 \u689d\u7528\u7b2c 2 \u689d\u898f\u5247\u3002\u767c\u751f\u932f\u8aa4\u7684\u5927\u90fd\u662f\ufa08\u653f\u5340\u5283\u5167\u67d0\u4e00\u90e8\u5206\u4e2d\u67d0\u7a2e\u7279\u5b9a\u5730\u57df\u7684\u9762\u7a4d\uff0c\u4e3b</td></tr><tr><td>\u8981\u554f\u984c\u662f\u9762\u7a4d\u4e3b\u9ad4\u904e\u65bc\u8907\u96dc\u3002\u5982\"\u5b89\u5fbd\uf96d\"\u91cb\u6587\u4e2d\u6709\uff1a</td></tr><tr><td>\u7696\u4e2d\u4e18\uf959\u6c34\u65f1\u4f5c\u7269\u904e\u6e21\u5340\u3002\u4ee5\u6c34\u7a3b\u3001\u5c0f\u9ea5\u70ba\u4e3b\u7684\u6c34\u65f1\u517c\u4f5c\u3001\u4e00\uf98e\uf978\u719f\u5340\u3002\u4f4d\u65bc\u6dee\u6cb3</td></tr><tr><td>\u4ee5\u5357\u3001\u6c5f\u6dee\u5206\u6c34\uf9ab-\u6ec1\u6cb3\u4e00\u7dda\u4ee5\uf963\uff0c\u571f\u5730\u9762\u7a4d\u5360\u5168\uf96d 23.7\uff05\uff0c\u2026\u2026</td></tr><tr><td>\u8a72\uf9b5\u4e2d\uff0c\u7b2c\u4e00\uf906\u662f\u500b\u5c0f\u6a19\u984c\uff0c\u5f8c\u9762\u5e7e\uf906\u662f\u5c0d\u8a72\u6a19\u984c\u6240\u6d89\u5730\u5340\u7684\uf96f\u660e\u3002\u6700\u5f8c\u4e00\uf906\u4e2d\u7684</td></tr><tr><td>\"\u9762\u7a4d\"\u7684\u4e3b\u9ad4\u662f\"\u5b89\u5fbd\uf96d\u7696\u4e2d\u4e18\uf959\u6c34\u65f1\u4f5c\u7269\u904e\u6e21\u5340\u571f\u5730\"\u3002\u9019\u4e00\u4e3b\u9ad4\u7684\u69cb\u6210\u65b9\u5f0f\u904e\u65bc\u8907</td></tr><tr><td>\u96dc\uff0c\u96e3\u4ee5\uf9fc\u5225\u3002</td></tr><tr><td>\u5f9e\u9019\u4e00\u5be6\u9a57\u4e2d\u53ef\ufa0a\uff0c</td></tr><tr><td>(1) \u7531\u6982\uf9a3\u3001\u6620\u5c04\u3001\u95dc\u4fc2\u548c\u898f\u5247\u7d44\u6210\u7684\u5f62\u5f0f\u7cfb\u7d71\u53ef\u4ee5\u6bd4\u8f03\u5168\u9762\u6e96\u78ba\u5730\u8868\u793a\u4e00\u4e9b</td></tr><tr><td>\u7c21\u55ae\u77e5\uf9fc\u5728\u767e\u79d1\u8fad\u5178\u6587\u672c\u4e2d\u7684\u5f62\u5f0f\uff0c\u9019\u4e00\u500b\u57fa\u65bc\u8a9e\u7fa9\u5c6c\u6027\u7684\u5f62\u5f0f\u7cfb\u7d71\u7684\u6846</td></tr><tr><td>\u67b6\u8a2d\u8a08\u662f\u6210\u529f\u7684\u3002</td></tr></table>",
+                "type_str": "table",
+                "html": null
+            }
+        }
+    }
+}

Full_text_JSON/prefixO/json/O03/O03-1001.json

@@ -0,0 +1,1473 @@
+{
+    "paper_id": "O03-1001",
+    "header": {
+        "generated_with": "S2ORC 1.0.0",
+        "date_generated": "2023-01-19T08:01:29.753223Z"
+    },
+    "title": "Word-Transliteration Alignment",
+    "authors": [
+        {
+            "first": "Tracy",
+            "middle": [],
+            "last": "Lin",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Chiao Tung University",
+                "location": {
+                    "addrLine": "Ta Hsueh Road",
+                    "postCode": "1001, 300",
+                    "settlement": "Hsinchu",
+                    "country": "Taiwan"
+                }
+            },
+            "email": "tracylin@cm.nctu.edu.tw"
+        },
+        {
+            "first": "Chien-Cheng",
+            "middle": [],
+            "last": "Wu",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Tsing Hua University",
+                "location": {
+                    "addrLine": "101, Kuangfu Road, Hsinchu, 300",
+                    "country": "Taiwan"
+                }
+            },
+            "email": ""
+        },
+        {
+            "first": "Jason",
+            "middle": [
+                "S"
+            ],
+            "last": "Chang",
+            "suffix": "",
+            "affiliation": {
+                "laboratory": "",
+                "institution": "National Tsing Hua University",
+                "location": {
+                    "addrLine": "101, Kuangfu Road, Hsinchu, 300",
+                    "country": "Taiwan"
+                }
+            },
+            "email": "jschang@cs.nthu.edu.tw"
+        }
+    ],
+    "year": "",
+    "venue": null,
+    "identifiers": {},
+    "abstract": "The named-entity phrases in free text represent a formidable challenge to text analysis. Translating a named-entity is important for the task of Cross Language Information Retrieval and Question Answering. However, both tasks are not easy to handle because named-entities found in free text are often not listed in a monolingual or bilingual dictionary. Although it is possible to identify and translate named-entities on the fly without a list of proper names and transliterations, an extensive list certainly will ensure the high accuracy rate of text analysis. We use a list of proper names and transliterations to train a Machine Transliteration Model. With the model it is possible to extract proper names and their transliterations in a bilingual corpus with high average precision and recall rates.",
+    "pdf_parse": {
+        "paper_id": "O03-1001",
+        "_pdf_hash": "",
+        "abstract": [
+            {
+                "text": "The named-entity phrases in free text represent a formidable challenge to text analysis. Translating a named-entity is important for the task of Cross Language Information Retrieval and Question Answering. However, both tasks are not easy to handle because named-entities found in free text are often not listed in a monolingual or bilingual dictionary. Although it is possible to identify and translate named-entities on the fly without a list of proper names and transliterations, an extensive list certainly will ensure the high accuracy rate of text analysis. We use a list of proper names and transliterations to train a Machine Transliteration Model. With the model it is possible to extract proper names and their transliterations in a bilingual corpus with high average precision and recall rates.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Abstract",
+                "sec_num": null
+            }
+        ],
+        "body_text": [
+            {
+                "text": "Multilingual named entity identification and (back) transliteration has been increasingly recognized as an important research area for many applications, including machine translation (MT), cross language information retrieval (CLIR), and question answering (QA). These transliterated words are often domainspecific and many of them are not found in existing bilingual dictionaries. Thus, it is difficult to handle transliteration only via simple dictionary lookup. For CLIR, the accuracy of transliteration highly affects the performance of retrieval.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "Transliteration of proper names tends to be varied from translator to translator. Consensus on transliteration of celebrated place and person names emerges over a short period of inconsistency and stays unique and unchanged thereafter. But for less known persons and unfamiliar places, the transliterations of names may vary a great deal. That is exacerbated by different systems used for Ramanizing Chinese or Japanese person and place names. For back transliteration task of converting many transliterations back to the unique original name, there is one and only solution. So back transliteration is considered more difficult than transliteration. Knight and Graehl (1998) pioneered the study of machine transliteration and proposed a statistical transliteration model from English to Japanese to experiment on back transliteration of Japanese named entities. Most previous approaches to machine transliteration (Al-Onaizan and Knight, 2002; Chen et al., 1998; Lin and Chen, 2002) ; English/Japanese (Knight and Graehl, 1998; Lee and Choi, 1997; Oh and Choi, 2002) focused on the tasks of transliteration and back-transliteration. Very little has been touched upon for the issue of aligning and acquiring words and transliterations in a parallel corpus.",
+                "cite_spans": [
+                    {
+                        "start": 651,
+                        "end": 675,
+                        "text": "Knight and Graehl (1998)",
+                        "ref_id": "BIBREF8"
+                    },
+                    {
+                        "start": 931,
+                        "end": 944,
+                        "text": "Knight, 2002;",
+                        "ref_id": "BIBREF0"
+                    },
+                    {
+                        "start": 945,
+                        "end": 963,
+                        "text": "Chen et al., 1998;",
+                        "ref_id": "BIBREF1"
+                    },
+                    {
+                        "start": 964,
+                        "end": 983,
+                        "text": "Lin and Chen, 2002)",
+                        "ref_id": "BIBREF10"
+                    },
+                    {
+                        "start": 1003,
+                        "end": 1028,
+                        "text": "(Knight and Graehl, 1998;",
+                        "ref_id": "BIBREF8"
+                    },
+                    {
+                        "start": 1029,
+                        "end": 1048,
+                        "text": "Lee and Choi, 1997;",
+                        "ref_id": "BIBREF9"
+                    },
+                    {
+                        "start": 1049,
+                        "end": 1067,
+                        "text": "Oh and Choi, 2002)",
+                        "ref_id": "BIBREF12"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "The alternative to on-the-fly (back) machine transliteration is simple lookup in an extensive list automatically acquired from parallel corpora. Most instances of (back) transliteration of proper names can often be found in a parallel corpus of substantial size and relevant to the task. For instance, fifty topics of the CLIR task in the NTCIR 3 evaluation conference contain many named entities (NEs) that require (back) transliteration. The CLIR task involves document retrieval from a collection of late 1990s news articles published in Taiwan. Most of those NEs and transliterations can be found in the articles from the Sinorama Corpus of parallel Chinese-English articles dated from 1990 to 2001, including \"Bill Clinton,\" \"Chernobyl,\" \"Chiayi,\" \"Han dynasty,\" \"James Soong,\" \"Kosovo,\" \"Mount Ali,\" \"Nobel Prize,\" \"Oscar,\" \"Titanic,\" and \"Zhu Rong Ji.\" Therefore it is important for CLIR research that we align and extract words and transliterations in a parallel corpus.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "In this paper, we propose a new machine transliteration method based on a statistical model trained automatically on a bilingual proper name list via unsupervised learning. We also describe how the parameters in the model can be estimated and smoothed for best results. Moreover, we show how the model can be applied to align and extract words and their transliterations in a parallel corpus.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "The remainder of the paper is organized as follows: Section 2 lays out the model and describes how to apply the model to align word and transliteration. Section 3 describes how the model is trained on a set of proper names and transliterations. Section 4 describes experiments and evaluation. Section 5 contains discussion and we conclude in Section 6.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Introduction",
+                "sec_num": "1."
+            },
+            {
+                "text": "We will first illustrate our approach with examples. A formal treatment of the approach will follow in Section 2.2.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Machine Transliteration Model",
+                "sec_num": "2."
+            },
+            {
+                "text": "Consider the case where one is to convert a word in English into another language, says Chinese, based on its phonemes rather than meaning. For instance, consider transliteration of the word \"Stanford,\" into Chinese. The most common transliteration of \"Stanford\" is \"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Examples",
+                "sec_num": "2.1"
+            },
+            {
+                "text": ".\" (Ramanization: [shi-dan-fo]). We assume that transliteration is a piecemeal, statistical process, converting one to six letters at a time to a Chinese character. For instance, to transliterate \"Stanford,\" the word is broken into \"s,\" \"tan,\" \"for,\" and \"d,\" which are converted into zero to two Chinese characters independently. Those fragments of the word in question are called transliteration units (TUs). In this case, the TU \"s\" is converted to the Chinese character \" ,\" \"tan\" to \" ,\" \"for\" to \" ,\" and \"d\" to the empty string \u03bb. In other words, we model the transliteration process based on independence of conversion of TUs. Therefore, we have the transliteration probability of getting the transliteration \" \" given \"Stanford,\" P( | Stanford),",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Examples",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "P( | Stanford) = P( | s) P( | tan) P( | for) P( \u03bb | d)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Examples",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "There are several ways such a machine transliteration model (MTM) can be applied, including (1) transliteration of proper names (2) back transliteration to the original proper name (3) wordtransliteration alignment in a parallel corpus. We formulate those three problems based on the probabilistic function under MTM:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Examples",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "Transliteration problem (TP)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Examples",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "Given a word w (usually a proper noun) in a language (L1), produce automatically the transliteration t in another language (L2). For instance, the transliterations in (2) are the results of solving the TP for four given words in (1).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Examples",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "(",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Examples",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "1) Berg, Stanford, Nobel, (2) , ,",
+                        "eq_num": ", Tsing Hua"
+                    }
+                ],
+                "section": "Examples",
+                "sec_num": "2.1"
+            },
+            {
+                "text": "Given a transliteration t in a language (L2), produce automatically the original word w in (L1) that gives rise to t. For instance, the words in (4) are the results of solving the BTP for two given transliterations in",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Back transliteration Problem (BTP)",
+                "sec_num": null
+            },
+            {
+                "text": "(3).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Back transliteration Problem (BTP)",
+                "sec_num": null
+            },
+            {
+                "text": "(",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Back transliteration Problem (BTP)",
+                "sec_num": null
+            },
+            {
+                "text": "3) , Lin Ku-fang (4) Michelangelo,",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Back transliteration Problem (BTP)",
+                "sec_num": null
+            },
+            {
+                "text": "Given a pair of sentence and translation counterpart, align the words and transliterations therein. For instance, given (5a) and (5b), the alignment results are the three word-transliteration pairs in (6), while the two pairs of word and back transliteration in (8) are the results of solving WTAP for (7a) and (7b) (5a) Paul Berg, professor emeritus of biology at Stanford University and a Nobel laureate, \u2026",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Word Transliteration Alignment Problem (WTAP)",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "1 (6) (Stanford, ), (Nobel, ), (Berg, )",
+                        "eq_num": "(5b)"
+                    }
+                ],
+                "section": "Word Transliteration Alignment Problem (WTAP)",
+                "sec_num": null
+            },
+            {
+                "text": "PRC premier Zhu Rongji's saber-rattling speech on the eve of the election is also seen as having aroused resentment among Taiwan's electorate, and thus given Chen Shui-bian a last-minute boost.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Word Transliteration Alignment Problem (WTAP)",
+                "sec_num": null
+            },
+            {
+                "text": "(7b) 2 (8) (Zhu Rongji, ), (Chen Shui-bian, )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Word Transliteration Alignment Problem (WTAP)",
+                "sec_num": null
+            },
+            {
+                "text": "Both transliteration and back transliteration are important for machine translation and cross language information retrieval. For instance, the person and place names are likely not listed in a dictionary, therefore should be mapped to the target language via run-time transliteration. Similarly, a large percentage of keywords in a cross language query are person and place names. It is important for an information system to produce appropriate counterpart names in the language of documents being searched. Those counterparts can be obtained via direct transliteration based on the machine transliteration and language models (of proper names in the target language).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Word Transliteration Alignment Problem (WTAP)",
+                "sec_num": null
+            },
+            {
+                "text": "The memory-based alternative is to find those word-transliteration in the aligned sentences in a parallel corpus (Chuang, You, and Chang 2002) . Word-transliteration alignment problem certainly can be dealt with based on lexical statistics (Gale and Church 1992; Melamed 2000). However, lexical statistics is known to be very ineffective for low-frequency words (Dunning 1993). We propose to attack WTAP at the sub-lexical, phoneme level.",
+                "cite_spans": [
+                    {
+                        "start": 113,
+                        "end": 142,
+                        "text": "(Chuang, You, and Chang 2002)",
+                        "ref_id": "BIBREF2"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Word Transliteration Alignment Problem (WTAP)",
+                "sec_num": null
+            },
+            {
+                "text": "We propose a new way for modeling transliteration of an English word w into Chinese t via a Machine Transliteration Model. We assume that transliteration is carried out by decomposing w into k translation units (TUs), \u03c9 1 , \u03c9 2 , \u2026, \u03c9 k which are subsequently converted independently into \u03c4 1 , \u03c4 2 , \u2026, \u03c4 k respectively.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "Finally, \u03c4 1 , \u03c4 2 , \u2026, \u03c4 k are put together, forming t as output. Therefore, the probability of converting w into t can be expressed as P( Figure 1 for more details.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 140,
+                        "end": 148,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "t | w) = ) | ( max , 1 ... , ... , k 1 k 1 i i k i k P \u03c9 \u03c4 \u03c4 \u03c4 \u03c9 \u03c9 = \u03a0 , where w = \u03c9 1 \u03c9 2 \u2026\u03c9 k , t = \u03c4 1 \u03c4 2 \u2026\u03c4 k , |t| \u2264 k \u2264 |t|+|w|, \u03c4 i \u03c9 i \u2260 \u03bb. See Equation (1) in",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "Based on MTM, we can formulate the solution to the Transliteration Problem by optimizing P(t | w)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "for the given w. On the other hand, we can formulate the solution to the Back Transliteration Problem by optimizing P(t | w) P( w) for the given t. See Equations (2) through (4) in Figure 1 for more details.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 181,
+                        "end": 189,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "The word-transliteration alignment process may be handled by first finding the proper names in English and matching up with the transliteration for each proper name. For instance, consider the following sentences in the Sinorama Corpus:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "(9c) \u4f60 \u4f60 (9e) \"When you understand all about the sun and all about the atmosphere and all about the rotation of the earth, you may still miss the radiance of the sunset.\" So wrote English philosopher Alfred North Whitehead.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "It is not difficult to build part of speech tagger or named entity recognizer for finding the following proper names (PN):",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "(10a) Alfred, (10b) North, (10c) Whitehead.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "We use Equation 5in Figure 1 to model the alignment of a word w and its transliteration t in s based on the alignment probability P(s, w) which is the product of transliteration probability P(\u03c3 | \u03c9) and a trigram match probability, P(m",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 20,
+                        "end": 28,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "i | m i-2 , m i-1 )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": ", where m i is the type of the i-th match in the alignment path.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "We define three match types based on lengths a and b, a ",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "= | \u03c4 |, b = | \u03c9 |: match(a, b) = H if a = 0, match(a, b) = V",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "The Model",
+                "sec_num": "2.2"
+            },
+            {
+                "text": "The probability of transliteration t of the word w",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "P(t | w) = ,",
+                        "eq_num": "(1)"
+                    }
+                ],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": ") | ( , 1 ... , ... , max k 1 k 1 i i k i k P \u03c9 \u03c4 \u03c4 \u03c4 \u03c9 \u03c9 \u03a0 = where w = \u03c9 1 \u03c9 2 \u2026 \u03c9 k , t = \u03c4 1 \u03c4 2 \u2026\u03c4 k , | t | \u2264 k \u2264 | t | + | w |, | \u03c4 i \u03c9 i | \u2265 1.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "TRANSLITERATION: Produce the phonetic translation equivalent t for the given word w t = arg",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "P(t | w)",
+                        "eq_num": "(2)"
+                    }
+                ],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "t max BACK TRANSLITERATION: Produce the original word w for the given transliteration t",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "P(w | t) = ) P( ) P( ) | P( t w w t (3) w = ) P( ) | P( max arg ) P( ) P( ) | P( max arg w w t t w w t t t = (4)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "WORD-TRANSLITERATION ALIGNMENT: Align a word w with its transliteration t in a sentence s",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "P(s, w) = P(\u03c3 \u03a0 = k i k , 1 ... , ... , max k 1 k 1 \u03c3 \u03c3 \u03c9 \u03c9 i | \u03c9 i ) P(m i | m i-2 , m i-1 ),",
+                        "eq_num": "(5)"
+                    }
+                ],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "where w = \u03c9 1 \u03c9 2 ...\u03c9 \u03ba , s = \u03c3 1 \u03c3 2 ...\u03c3 \u03ba , (both \u03c9 i and \u03c3 i can be empty) To compute the alignment probability efficiently, we need to define and calculate the forward probability \u03b1(i, j) of P(s, w) via dynamic programming (Manning and Schutze 1999) , \u03b1(i, j) denotes the probability of aligning the first i Chinese characters of s and the first j English letters of w. For the match type trigram in Equation 5and 8, we need also compute \u00b5(i, j), the types of the last two matches in the Viterbi alignment path. See Equations (5) through (9) in Figure 1 for more details.",
+                "cite_spans": [
+                    {
+                        "start": 229,
+                        "end": 255,
+                        "text": "(Manning and Schutze 1999)",
+                        "ref_id": "BIBREF11"
+                    }
+                ],
+                "ref_spans": [
+                    {
+                        "start": 551,
+                        "end": 559,
+                        "text": "Figure 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "| s | \u2264 k \u2264 | w | + | s |, |\u03c9 i \u03c3 i | \u2265 1, m i is the type of the (\u03c9 i , \u03c3 i ) match, m i = match (|\u03c9 i |, | \u03c3 i | ),",
+                        "eq_num": "match"
+                    }
+                ],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "For instance, given w = \"Whitehead\" and s = \" \u4f60 \u4f60 ,\" the best Viterbi path indicates a decomposition of word \"Whitehead\" into four TUs, \"whi,\" \"te,\" \"hea,\" and \"d\" matching \" ,\" \u03bb, \" ,\"",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "\" \" respectively. By extracting the sequence of Dand V-matches, we generate the result of wordtransliteration alignment. For instance, we will have ( , Whitehead) as the output. See Figure 2 for more details.",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 182,
+                        "end": 190,
+                        "text": "Figure 2",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "MACHINE TRANSLITERATION MODEL:",
+                "sec_num": null
+            },
+            {
+                "text": "In the training phase, we estimate the transliteration probability function P(\u03c4 | \u03c9), for any given TU \u03c9 and transliteration character \u03c4, based on a given list of word-transliterations. Based on the Expectation Maximization (EM) algorithm with Viterbi decoding (Forney, 1973) , the iterative parameter estimation procedure on a training data of word-transliteration list, (E k , C k ), k = 1 to n is described as follows:",
+                "cite_spans": [
+                    {
+                        "start": 261,
+                        "end": 275,
+                        "text": "(Forney, 1973)",
+                        "ref_id": "BIBREF7"
+                    }
+                ],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Estimation of Model Parameters",
+                "sec_num": "3."
+            },
+            {
+                "text": "Initialization Step:",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Estimation of Model Parameters",
+                "sec_num": "3."
+            },
+            {
+                "text": "Initially, we have a simple model P 0 (\u03c4 | \u03c9) ,' we have and R(\u03c4 1 ) = 'na' and R(\u03c4 2 ) = 'ya' under Yanyu Pinyin Romanization System. Therefore, breaking up w into two TUs, \u03c9 1 = 'nay' \u03c9 2 = 'yar' is most probable, since that maximizes P 0 (\u03c4 1 | \u03c9 1 ) \u00d7 P 0 (\u03c4 2 | \u03c9 2 ) P 0 (\u03c4 1 | \u03c9 1 )= sim( na | nay) = 2 \u00d7 2 / (2+3) = 0.8 P 0 (\u03c4 2 | \u03c9 2 )= sim( ya | yar) = 2 \u00d7 2 / (2+3) = 0.8",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Estimation of Model Parameters",
+                "sec_num": "3."
+            },
+            {
+                "text": "EQUATION",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [
+                    {
+                        "start": 0,
+                        "end": 8,
+                        "text": "EQUATION",
+                        "ref_id": "EQREF",
+                        "raw_str": "P 0 (\u03c4 | \u03c9) = sim( R(\u03c4) | \u03c9) = dice(t 1 t 2 \u2026t a , w 1 w 2 \u2026w b )",
+                        "eq_num": "(8)"
+                    }
+                ],
+                "section": "Estimation of Model Parameters",
+                "sec_num": "3."
+            },
+            {
+                "text": "In the Expectation Step, we find the best way to describe how a word get transliterated via decomposition into TUs which amounts to finding the best Viterbi path aligning TUs in E k and characters in C k for all pairs (E k , C k ), k = 1 to n, in the training set. This can be done using Equations (5) through (9). In the training phase, we have slightly different situation of s = t. Table 1 . The results of using P 0 (\u03c4 |\u03c9) to align TUs and transliteration characters w s=t \u03c9-\u03c4 match on Viterbi path",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 385,
+                        "end": 392,
+                        "text": "Table 1",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Expectation Step:",
+                "sec_num": null
+            },
+            {
+                "text": "The Viterbi path can be found via a dynamic programming process of calculating the forward probability function \u03b1(i, j) of the transliteration alignment probability P(E k , C k ) for 0 < i < | C k | and 0 < j < | E k |. After calculating P(C k , E k ) via dynamic programming, we also obtain the TU matches (\u03c4, \u03c9) on the Viterbi path. After all pairs are processed and TUs and translation characters are found, we then reestimate the transliteration probability P(\u03c4 | \u03c9) in the Maximization Step",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Expectation Step:",
+                "sec_num": null
+            },
+            {
+                "text": "Step: Based on all the TU alignment pairs obtained in the Expectation Step, we update the maximum likelihood estimates (MLE) of model parameters using Equation 9.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Maximization",
+                "sec_num": null
+            },
+            {
+                "text": "\u2211 \u2211 \u2211 \u2211 = = = n i C E n i C E MLE 1 ) , ( in matches ' 1 ) , ( in matches ) count( ) , count( ) | ( P i i i i \u03c9 \u03c9 \u03c4 \u03c9 \u03c4 \u03c9 \u03c4 \u03c9 \u03c4 (9)",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Maximization",
+                "sec_num": null
+            },
+            {
+                "text": "The Viterbi EM algorithm iterates between the Expectation Step and Maximization Step, until a stopping criterion is reached or after a predefined number of iterations. Re-estimation of P(\u03c4 | \u03c9) leads to convergence under the Viterbi EM algorithm.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Maximization",
+                "sec_num": null
+            },
+            {
+                "text": "The maximum likelihood estimate is generally not suitable for statistical inference of parameters in the proposed machine transliteration model due to data sparseness (even if we use a longer list of names for training, the problem still exists). MLE is not capturing the fact that there are other transliteration possibilities that we may have not encountered. For instance, consider the task of aligning the word \"Michelangelo\" and the transliteration \" \" in Example (11):",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Parameter Smoothing",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "(11) (Michelangelo, )",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Parameter Smoothing",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "It turns out in the model trained on some word-transliteration data provides the MLE parameters in the MTM in Table 2 . Understandably, the MLE-based model assigns 0 probability to a lot of cases not seen in the training data and that could lead to problems in word-transliteration alignment. For instance, relevant parameters for Example (11) such as P( | che) and P( | lan) are given 0 probability. Good Turing estimation is one of the most commonly used approaches to deal with the problems caused by data sparseness and zero probability. However, GTE assigns identical probabilistic values to all unseen events, which might lead to problem in our case. We observed that although there is great variation in Chinese transliteration characters for any given English word, the initial, mostly consonants, tend to be consistent. See Table 3 for ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 110,
+                        "end": 117,
+                        "text": "Table 2",
+                        "ref_id": "TABREF0"
+                    },
+                    {
+                        "start": 833,
+                        "end": 844,
+                        "text": "Table 3 for",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Parameter Smoothing",
+                "sec_num": "3.1"
+            },
+            {
+                "text": "We have carried out rigorous evaluation on an implementation of the method proposed in this paper.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiments and evaluation",
+                "sec_num": "4"
+            },
+            {
+                "text": "Close examination of the experimental results reveal that the machine transliteration is general effective in aligning and extracting proper names and their transliterations from a parallel corpus.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiments and evaluation",
+                "sec_num": "4"
+            },
+            {
+                "text": "The parameters of the transliteration model were trained on some 1,700 proper names and transliterations from Scientific American Magazine. We place 10 H-matches before and after the Viterbi alignment (1) 200 bilingual examples in Longman Dictionary of Comtemporary Dictionary, English-Chinese Edition. (2) 200 aligned sentences from Scientific American, US and Taiwan Editions.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Experiments and evaluation",
+                "sec_num": "4"
+            },
+            {
+                "text": "(3) 200 aligned sentences from the Sinorama Corpus. Table 5 shows that on the average the precision rate of exact match is between 75-90%, while the precision rate for character level partial match is from 90-95%. The average recall rates are about the same as the precision rates. ",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 52,
+                        "end": 59,
+                        "text": "Table 5",
+                        "ref_id": "TABREF1"
+                    }
+                ],
+                "eq_spans": [],
+                "section": "Experiments and evaluation",
+                "sec_num": "4"
+            },
+            {
+                "text": "The success of the proposed method for the most part has to do with the capability to balance the conflicting needs of capturing lexical preference of transliteration and smoothing to cope with data sparseness and generality. Although we experimented with a model trained on English to Chinese transliteration, the model seemed to perform reasonably well even with situations in the opposite direction, Chinese to English transliteration. This indicates that the model with the parameter estimation method is very general in terms of dealing with unseen events and bi-directionality.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Discussion",
+                "sec_num": "5."
+            },
+            {
+                "text": "We have restricted our discussion and experiments to transliteration of proper names. While it is commonplace for Japanese to have transliteration of common nouns, transliteration of Chinese common nouns into English is rare. It seems that is so only when the term is culture-specific and there is no counterparts in the West. For instance, most instances \" \" and \" \" found in the Sinorama corpus are mapped into lower case transliterations as shown in Example (11) and (12):",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Discussion",
+                "sec_num": "5."
+            },
+            {
+                "text": "(11a) (11b) Are ch'i-p'aos--the national dress of China--really out of fashion?",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Discussion",
+                "sec_num": "5."
+            },
+            {
+                "text": "(12a) (12b) a scroll of shou chin ti calligraphy Without capitalized transliterations, it remains to be seen how word-transliteration alignment related to common nouns should be handled.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Discussion",
+                "sec_num": "5."
+            },
+            {
+                "text": "In this paper, we propose a new statistical machine transliteration model and describe how to apply the model to extract words and transliterations in a parallel corpus. The model was first trained on a modest list of names and transliteration. The training resulted in a set of 'syllabus' to character transliteration probabilities, which are subsequently used to extract proper names and transliterations in a parallel corpus.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "6."
+            },
+            {
+                "text": "These named entities are crucial for the development of named entity identification module in CLIR and QA.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "6."
+            },
+            {
+                "text": "We carried out experiments on an implementation of the word-transliteration alignment algorithms and tested on three sets of test data. The evaluation showed that very high precision rates were achieved.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "6."
+            },
+            {
+                "text": "A number of interesting future directions present themselves. First, it would be interesting to see how effectively we can port and apply the method to other language pairs such as English-Japanese and English-Korean. We are also investigating the advantages of incorporate a machine transliteration module in sentence and word alignment of parallel corpora.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Conclusion",
+                "sec_num": "6."
+            },
+            {
+                "text": "Scientific American, US and Taiwan editions. What Clones? Were claims of the first human embryo premature? Gary Stix and (Trans.) December 24, 2001.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            },
+            {
+                "text": "Sinorama Chinese-English Magazine, A New Leader for the New Century--Chen Elected President, April 2000, p. 13.",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "",
+                "sec_num": null
+            }
+        ],
+        "back_matter": [
+            {
+                "text": "We acknowledge the support for this study through grants from National Science Council and Ministry of Education, Taiwan (NSC 90-2411-H-007-033-MC and MOE EX-91-E-FA06-4-4).",
+                "cite_spans": [],
+                "ref_spans": [],
+                "eq_spans": [],
+                "section": "Acknowledgement",
+                "sec_num": null
+            },
+            {
+                "text": "path to simulate the word-transliteration situation and trained the trigram match type probability. Table 4 shows the estimates of the trigram model. The model was then tested on three sets of test data:",
+                "cite_spans": [],
+                "ref_spans": [
+                    {
+                        "start": 100,
+                        "end": 107,
+                        "text": "Table 4",
+                        "ref_id": null
+                    }
+                ],
+                "eq_spans": [],
+                "section": "annex",
+                "sec_num": null
+            }
+        ],
+        "bib_entries": {
+            "BIBREF0": {
+                "ref_id": "b0",
+                "title": "Translating named entities using monolingual and bilingual resources",
+                "authors": [
+                    {
+                        "first": "Y",
+                        "middle": [],
+                        "last": "Al-Onaizan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Knight",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2002,
+                "venue": "Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics (ACL)",
+                "volume": "",
+                "issue": "",
+                "pages": "400--408",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Al-Onaizan, Y. and K. Knight. 2002. Translating named entities using monolingual and bilingual re- sources. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics (ACL), pages 400-408.",
+                "links": null
+            },
+            "BIBREF1": {
+                "ref_id": "b1",
+                "title": "Proper name translation in cross-language information retrieval",
+                "authors": [
+                    {
+                        "first": "H",
+                        "middle": [
+                            "H"
+                        ],
+                        "last": "Chen",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S-J",
+                        "middle": [],
+                        "last": "Huang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Y-W",
+                        "middle": [],
+                        "last": "Ding",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "S-C",
+                        "middle": [],
+                        "last": "Tsai",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of 17th COLING and 36th ACL",
+                "volume": "",
+                "issue": "",
+                "pages": "232--236",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chen, H.H., S-J Huang, Y-W Ding, and S-C Tsai. 1998. Proper name translation in cross-language infor- mation retrieval. In Proceedings of 17th COLING and 36th ACL, pages 232-236.",
+                "links": null
+            },
+            "BIBREF2": {
+                "ref_id": "b2",
+                "title": "Adaptive Bilingual Sentence Alignment",
+                "authors": [
+                    {
+                        "first": "T",
+                        "middle": [],
+                        "last": "Chuang",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "G",
+                        "middle": [
+                            "N"
+                        ],
+                        "last": "You",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Chang",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2002,
+                "venue": "Lecture Notes in Artificial Intelligence",
+                "volume": "2499",
+                "issue": "",
+                "pages": "21--30",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Chuang, T., G.N. You, J.S. Chang (2002) Adaptive Bilingual Sentence Alignment, Lecture Notes in Arti- ficial Intelligence 2499, 21-30.",
+                "links": null
+            },
+            "BIBREF3": {
+                "ref_id": "b3",
+                "title": "What Clones? SCIENTIFIC AMERICAN, Inc",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [
+                            "B R P"
+                        ],
+                        "last": "Cibelli",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "M",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Lanza",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "C",
+                        "middle": [],
+                        "last": "West",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Ezzell",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2002,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Cibelli, J.B. R.P. Lanza, M.D. West, and C. Ezzell. 2002. What Clones? SCIENTIFIC AMERICAN, Inc., New York, January. http://www.sciam.com.",
+                "links": null
+            },
+            "BIBREF4": {
+                "ref_id": "b4",
+                "title": "Robust bilingual word alignment for machine aided translation",
+                "authors": [
+                    {
+                        "first": "I",
+                        "middle": [],
+                        "last": "Dagan",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [
+                            "W"
+                        ],
+                        "last": "Church",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "W",
+                        "middle": [
+                            "A"
+                        ],
+                        "last": "Gale",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1993,
+                "venue": "Proceedings of the Workshop on Very Large Corpora: Academic and Industrial Perspectives",
+                "volume": "",
+                "issue": "",
+                "pages": "1--8",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Dagan, I., Church, K. W., and Gale, W. A. 1993. Robust bilingual word alignment for machine aided translation. In Proceedings of the Workshop on Very Large Corpora: Academic and Industrial Per- spectives, pages 1-8, Columbus Ohio.",
+                "links": null
+            },
+            "BIBREF5": {
+                "ref_id": "b5",
+                "title": "Maximum likelihood from incomplete data via the EM algorithm",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [],
+                        "last": "Dempster",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "N",
+                        "middle": [],
+                        "last": "Laird",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "Rubin",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [],
+                        "last": "",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1977,
+                "venue": "Journal of the Royal Statistical Society, Series B",
+                "volume": "39",
+                "issue": "1",
+                "pages": "1--38",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Dempster, A., Laird, N., and Rubin, D. (1977). Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, Series B, 39(1):1-38.",
+                "links": null
+            },
+            "BIBREF6": {
+                "ref_id": "b6",
+                "title": "Maximum likelihood from incomplete data via the EM algorithm",
+                "authors": [
+                    {
+                        "first": "A",
+                        "middle": [
+                            "P"
+                        ],
+                        "last": "Dempster",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "N",
+                        "middle": [
+                            "M"
+                        ],
+                        "last": "Laird",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "D",
+                        "middle": [
+                            "B"
+                        ],
+                        "last": "Rubin",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1977,
+                "venue": "Journal of the Royal Statistical Society",
+                "volume": "39",
+                "issue": "1",
+                "pages": "1--38",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Dempster, A.P., N.M. Laird, and D.B. Rubin. 1977. Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, 39(1):1-38.",
+                "links": null
+            },
+            "BIBREF7": {
+                "ref_id": "b7",
+                "title": "The Viterbi algorithm",
+                "authors": [
+                    {
+                        "first": "G",
+                        "middle": [
+                            "D"
+                        ],
+                        "last": "Forney",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1973,
+                "venue": "Proceedings of IEEE",
+                "volume": "61",
+                "issue": "",
+                "pages": "268--278",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Forney, G.D. 1973. The Viterbi algorithm. Proceedings of IEEE, 61:268-278, March.",
+                "links": null
+            },
+            "BIBREF8": {
+                "ref_id": "b8",
+                "title": "Machine transliteration",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Knight",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "J",
+                        "middle": [],
+                        "last": "Graehl",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Computational Linguistics",
+                "volume": "24",
+                "issue": "4",
+                "pages": "599--612",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Knight, K. and J. Graehl. 1998. Machine transliteration. Computational Linguistics, 24(4):599-612.",
+                "links": null
+            },
+            "BIBREF9": {
+                "ref_id": "b9",
+                "title": "A statistical method to generate various foreign word transliterations in multilingual information retrieval system",
+                "authors": [
+                    {
+                        "first": "J",
+                        "middle": [
+                            "S"
+                        ],
+                        "last": "Lee",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K-S",
+                        "middle": [],
+                        "last": "Choi",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1997,
+                "venue": "Proceedings of the 2nd International Workshop on Information Retrieval with Asian Languages (IRAL'97)",
+                "volume": "",
+                "issue": "",
+                "pages": "123--128",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lee, J.S. and K-S Choi. 1997. A statistical method to generate various foreign word transliterations in multilingual information retrieval system. In Proceedings of the 2nd International Workshop on Infor- mation Retrieval with Asian Languages (IRAL'97), pages 123-128, Tsukuba, Japan.",
+                "links": null
+            },
+            "BIBREF10": {
+                "ref_id": "b10",
+                "title": "Backward transliteration by learning phonetic similarity",
+                "authors": [
+                    {
+                        "first": "W-H",
+                        "middle": [],
+                        "last": "Lin",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "H-H",
+                        "middle": [],
+                        "last": "Lin",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Chen",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2002,
+                "venue": "CoNLL-2002, Sixth Conference on Natural Language Learning",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Lin, W-H Lin and H-H Chen. 2002. Backward transliteration by learning phonetic similarity. In CoNLL- 2002, Sixth Conference on Natural Language Learning, Taipei, Taiwan.",
+                "links": null
+            },
+            "BIBREF11": {
+                "ref_id": "b11",
+                "title": "Foundations of Statistical Natural Language Processing",
+                "authors": [
+                    {
+                        "first": "Ch",
+                        "middle": [],
+                        "last": "Manning",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "H",
+                        "middle": [],
+                        "last": "Schutze",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1999,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Manning, Ch. and H. Schutze. 1999. Foundations of Statistical Natural Language Processing, MIT Press; 1st edition.",
+                "links": null
+            },
+            "BIBREF12": {
+                "ref_id": "b12",
+                "title": "An English-Korean transliteration model using pronunciation and contextual rules",
+                "authors": [
+                    {
+                        "first": "J-H",
+                        "middle": [],
+                        "last": "Oh",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K-S",
+                        "middle": [],
+                        "last": "Choi",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2002,
+                "venue": "Proceedings of the 19th International Conference on Computational Linguistics (COLING)",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Oh, J-H and K-S Choi. 2002. An English-Korean transliteration model using pronunciation and contex- tual rules. In Proceedings of the 19th International Conference on Computational Linguistics (COLING), Taipei, Taiwan.",
+                "links": null
+            },
+            "BIBREF14": {
+                "ref_id": "b14",
+                "title": "Sinorama Magazine",
+                "authors": [
+                    {
+                        "first": "",
+                        "middle": [],
+                        "last": "Sinorama",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2002,
+                "venue": "",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Sinorama. 2002. Sinorama Magazine. http://www.greatman.com.tw/sinorama.htm.",
+                "links": null
+            },
+            "BIBREF15": {
+                "ref_id": "b15",
+                "title": "Translating names and technical terms in Arabic text",
+                "authors": [
+                    {
+                        "first": "B",
+                        "middle": [
+                            "G"
+                        ],
+                        "last": "Stalls",
+                        "suffix": ""
+                    },
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Knight",
+                        "suffix": ""
+                    }
+                ],
+                "year": 1998,
+                "venue": "Proceedings of the COLING/ACL Workshop on Computational Approaches to Semitic Languages",
+                "volume": "",
+                "issue": "",
+                "pages": "",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Stalls, B.G. and K. Knight. 1998. Translating names and technical terms in Arabic text. In Proceedings of the COLING/ACL Workshop on Computational Approaches to Semitic Languages.",
+                "links": null
+            },
+            "BIBREF16": {
+                "ref_id": "b16",
+                "title": "Automatic extraction of translational Japanese-KATAKANA and English word pairs from bilingual corpora",
+                "authors": [
+                    {
+                        "first": "K",
+                        "middle": [],
+                        "last": "Tsujii",
+                        "suffix": ""
+                    }
+                ],
+                "year": 2002,
+                "venue": "International Journal of Computer Processing of Oriental Languages",
+                "volume": "15",
+                "issue": "3",
+                "pages": "261--279",
+                "other_ids": {},
+                "num": null,
+                "urls": [],
+                "raw_text": "Tsujii, K. 2002. Automatic extraction of translational Japanese-KATAKANA and English word pairs from bilingual corpora. International Journal of Computer Processing of Oriental Languages, 15(3):261-279.",
+                "links": null
+            }
+        },
+        "ref_entries": {
+            "FIGREF0": {
+                "text": "if b = 0, and match(a, b) = D if a > 0 and b > 0. The D-match represents a non-empty TU \u03c9 matching a transliteration character \u03c4, while the V-match represents the English letters omitted in the transliteration process.",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF1": {
+                "text": "(a, b) = H, if b = 0, match(a, b) = V, if a = 0, match(a, b) = D, if a > 0 and b > 0, P(m i | m i-2 , m i-1 ) is trigram Markov model probabiltiy of match types. \u03b1(i, j ) = P(s 1:i-1 , w 1:j-1 :j-1 | w i-b:i-1 ) P( match(a, b) | \u00b5(i-a, j-b) ). (8) \u00b5(i, j) = (m, match(a*, b*)), where \u00b5(i-a*, j-b*) = (x, m),(9)where (a*, b*) = \u03b1(i-a, j-b) :j-1 | w i-b:i-1 ) P( match(a, b) | \u00b5(i-a, j-b) ).",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF2": {
+                "text": "The equations for finding the Viterbi path of matching a proper name and its translation in a sentence The Viterbi alignment path for Example (9c) and the proper name \"Whitehead\" (10c) in the sentence (9e), consisting of one V-match (te-\u03bb), three D-matches (whi\u2212 , hea\u2212 , d\u2212 ), and many H-matches.",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF3": {
+                "text": "\u03c4) = Romanization of Chinese character \u03c4 R(\u03c4) = t 1 t 2 \u2026t a \u03c9 = w 1 w 2 \u2026w b c = # of common letters between R(\u03c4) and \u03c9For instance, given w = 'Nayyar' and t = '",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "FIGREF4": {
+                "text": "more details. Based on that observation, we use the linear interpolation of the Good-Turing estimation of TU-to-TU and the class-based initial-to-initial function to approximate the parameters in MTM. Therefore, we have",
+                "type_str": "figure",
+                "num": null,
+                "uris": null
+            },
+            "TABREF0": {
+                "type_str": "table",
+                "html": null,
+                "text": "P MLE (t | n) value relevant toExample (11)",
+                "content": "<table><tr><td>English TU \u03c9 Transliteration \u03c4</td><td>P MLE (\u03c4 | \u03c9)</td></tr><tr><td>mi</td><td>0.00394</td></tr><tr><td>mi</td><td>0.00360</td></tr><tr><td>mi</td><td>0.00034</td></tr><tr><td>mi</td><td>0.00034</td></tr><tr><td>mi</td><td>0.00017</td></tr><tr><td>che</td><td>0.00034</td></tr><tr><td>che</td><td>0.00017</td></tr><tr><td>che</td><td>0.00017</td></tr><tr><td>che</td><td>0.00017</td></tr><tr><td>che</td><td>0.00017</td></tr><tr><td>che</td><td>0.00017</td></tr><tr><td>che</td><td>0</td></tr><tr><td>lan</td><td>0.00394</td></tr><tr><td>lan</td><td>0.00051</td></tr><tr><td>lan</td><td>0.00017</td></tr><tr><td>lan</td><td>0</td></tr><tr><td>ge</td><td>0.00102</td></tr><tr><td>ge</td><td>0.00085</td></tr><tr><td>ge</td><td>0.00068</td></tr><tr><td>ge</td><td>0.00017</td></tr><tr><td>ge</td><td>0.00017</td></tr><tr><td>lo</td><td>0.00342</td></tr><tr><td>lo</td><td>0.00171</td></tr><tr><td>lo</td><td>0.00017</td></tr></table>",
+                "num": null
+            },
+            "TABREF1": {
+                "type_str": "table",
+                "html": null,
+                "text": "The experimental results of word-transliteration alignement",
+                "content": "<table><tr><td>Test</td><td># of words</td><td># of matches</td><td>Word precision</td></tr><tr><td>Data</td><td>( # of characters)</td><td>(# of characters)</td><td>(Characters)</td></tr><tr><td>LODCE</td><td>200</td><td>179</td><td>89.5%</td></tr><tr><td/><td>(496)</td><td>(470)</td><td>(94.8%)</td></tr><tr><td>Sinorama</td><td>200</td><td>151</td><td>75.5%</td></tr><tr><td/><td>(512)</td><td>(457)</td><td>(89.3%)</td></tr><tr><td>Sci. Am.</td><td>200</td><td>180</td><td>90.0%</td></tr><tr><td/><td>(602)</td><td>(580)</td><td>(96.3%)</td></tr></table>",
+                "num": null
+            }
+        }
+    }
+}