File size: 93,791 Bytes
6fa4bc9 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 |
{
"paper_id": "I13-1024",
"header": {
"generated_with": "S2ORC 1.0.0",
"date_generated": "2023-01-19T07:14:13.621224Z"
},
"title": "Dependency Parsing for Identifying Hungarian Light Verb Constructions",
"authors": [
{
"first": "Veronika",
"middle": [],
"last": "Vincze",
"suffix": "",
"affiliation": {
"laboratory": "Research Group on Artificial Intelligence",
"institution": "Hungarian Academy of Sciences",
"location": {}
},
"email": "vinczev@inf.u-szeged.hu"
},
{
"first": "J\u00e1nos",
"middle": [],
"last": "Zsibrita",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "University of Szeged",
"location": {}
},
"email": "zsibrita@inf.u-szeged.hu"
},
{
"first": "Istv\u00e1n",
"middle": [],
"last": "Nagy",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "University of Szeged",
"location": {}
},
"email": ""
}
],
"year": "",
"venue": null,
"identifiers": {},
"abstract": "Light verb constructions (LVCs) are verb and noun combinations in which the verb has lost its meaning to some degree and the noun is used in one of its original senses. They often share their syntactic pattern with other constructions (e.g. verbobject pairs) thus LVC detection can be viewed as classifying certain syntactic patterns as light verb constructions or not. In this paper, we explore a novel way to detect LVCs in texts: we apply a dependency parser to carry out the task. We present our experiments on a Hungarian treebank, which has been manually annotated for dependency relations and light verb constructions. Our results outperformed those achieved by state-of-the-art techniques for Hungarian LVC detection, especially due to the high precision and the treatment of long-distance dependencies.",
"pdf_parse": {
"paper_id": "I13-1024",
"_pdf_hash": "",
"abstract": [
{
"text": "Light verb constructions (LVCs) are verb and noun combinations in which the verb has lost its meaning to some degree and the noun is used in one of its original senses. They often share their syntactic pattern with other constructions (e.g. verbobject pairs) thus LVC detection can be viewed as classifying certain syntactic patterns as light verb constructions or not. In this paper, we explore a novel way to detect LVCs in texts: we apply a dependency parser to carry out the task. We present our experiments on a Hungarian treebank, which has been manually annotated for dependency relations and light verb constructions. Our results outperformed those achieved by state-of-the-art techniques for Hungarian LVC detection, especially due to the high precision and the treatment of long-distance dependencies.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Abstract",
"sec_num": null
}
],
"body_text": [
{
"text": "Multiword expressions (MWEs) are lexical items that can be decomposed into single words and display lexical, syntactic, semantic, pragmatic and/or statistical idiosyncrasy (Kim, 2008) . Light verb constructions (LVCs) form a subtype of MWEs: they are verb and noun combinations in which the verb has lost its meaning to some degree and the noun is used in one of its original senses (e.g. make a decision or take a walk). In several NLP applications like information retrieval or machine translation it is important to identify LVCs in context since they require special treatment, particularly because of their semantic features. Thus, LVCs should be identified to help these applications.",
"cite_spans": [
{
"start": 172,
"end": 183,
"text": "(Kim, 2008)",
"ref_id": "BIBREF12"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Light verb constructions (e.g. make a mistake) often share their syntactic pattern with literal verb + noun combinations (e.g. make a cake). Thus, specific syntactic constructions -e.g. verb-object pairs -can be separated into two classes: one where the noun behaves as a real object (cake) and one where the noun functions as the light verb object (mistake). Thus, LVC detection can be viewed as classifying certain syntactic patterns as LVCs or not and assigning a specific syntactic label to the argument of the light verb.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "In this paper, we explore a novel way to LVC detection: we apply a dependency parser to carry out the task. Although the usability of identified multiword expressions has been investigated in the literature (see Section 4), and many MWE detection systems rely on syntactic information, we are not aware of any approach that aimed at applying a dependency parser for the dedicated task of identifying LVCs. Our approach requires a treebank annotated for syntactic and LVC information at the same time. Due to the availability of annotated resources, we focus on light verb constructions in Hungarian, a morphologically rich language. Thus, we present our experiments on the legal subcorpus of the Szeged Dependency Treebank annotated for LVCs as well as dependency relations . We will pay special attention to noncontiguous LVCs in our investigations as there are quite a few non-contiguous LVCs in Hungarian due to the free word order. Our results empirically prove that LVCs can be detected as a \"side effect\" of dependency parsing.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Hungarian is an agglutinative language, which means that a word can have hundreds of word forms due to inflectional or derivational morphology (\u00c9. Kiss, 2002) . Hungarian word order is related to information structure, e.g. new (or emphatic) information (focus) always precedes the verb and old information (topic) precedes the focus position. Thus, the position relative to the verb has no predictive force as regards the syntactic function of the given argument. In English, the noun phrase before the verb is most typically the subject whereas in Hungarian, it is the focus of the sentence, which itself can be the subject, object or any other argument. The grammatical function of words is determined by case suffixes. Hungarian nouns can have about 20 cases, which mark the relationship between the verb and its arguments (subject, object, dative etc.) and adjuncts (mostly adverbial modifiers). Although there are postpositions in Hungarian, case suffixes can also express relations that are expressed by prepositions in English. Verbs are inflected for person and number and the definiteness of the object. There are several other linguistic phenomena that are syntactic in nature in English but they are encoded morphologically in Hungarian. For instance, causation and modality are expressed by derivational suffixes.",
"cite_spans": [
{
"start": 147,
"end": 158,
"text": "Kiss, 2002)",
"ref_id": "BIBREF5"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions in Hungarian",
"sec_num": "2"
},
{
"text": "The canonical form of a Hungarian light verb construction is a bare noun + third person singular verb, for instance, tan\u00e1csot ad advice-ACC give \"to give advice\". Due to the above features, they may occur in non-canonical versions as well: the verb may precede the noun, or they may be not adjacent, moreover, the verb may occur in different surface forms inflected for tense, mood, person and number.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions in Hungarian",
"sec_num": "2"
},
{
"text": "LVCs may occur in several forms due to their syntactic flexibility. Besides the prototypical verbal form in Hungarian, they can have a participial form (e.g. figyelembe vev\u0151 account-INE taking \"taking into account\") and they may also undergo nominalization, yielding a nominal compound (e.g.\u00e9letbe l\u00e9p\u00e9s life-INE step \"entering into force\"). 1 From a morphological perspective, LVCs can also be divided into groups. First, the nominal component is the object of the verb, i.e. it bears an accusative case in Hungarian (e.g. d\u00f6nt\u00e9st hoz decision-ACC bring \"to make a decision\" or tan\u00e1csot ad advice-ACC give \"to give advice\"). Second, the nominal component can bear other (oblique) cases as well (e.g. zavarba hoz embarrassment-ILL bring \"to embarrass\" or figyelemmel k\u00eds\u00e9r attention-INS follow \"to pay attention\"). Third, -although rarely -a postpositional phrase can also occur in the construction (e.g. uralom al\u00e1 jut rule under get \"to get under rule\" or hat\u00e1s alatt\u00e1ll effect under stand \"to be under effect\").",
"cite_spans": [
{
"start": 342,
"end": 343,
"text": "1",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions in Hungarian",
"sec_num": "2"
},
{
"text": "Although light verb constructions are made of two parts, namely, the nominal component and the verb, thus, they show phrasal properties, it can be argued that from a semantic point of view they form one unit. First, many light verb constructions have a verbal counterpart with the same meaning (e.g. d\u00f6nt\u00e9st hoz decision-ACC bring \"to make a decision\" -d\u00f6nt \"to decide\"). Second, there are meanings that can only be expressed through a light verb construction (e.g. h\u00e1zkutat\u00e1st tart (search.of.premises-ACC hold) 'to conduct search of premises' in Hungarian). Third, there are languages that abound in verb + noun constructions or multiword verbs (such as Estonian (Muischnek and Kaalep, 2010) or Persian (Mansoory and Bijankhan, 2008) ): verbal concepts are mostly expressed by combining a noun with a light verb (Mansoory and Bijankhan, 2008) . On the other hand, there are views that the relationship between the verbal and the nominal component is not that of a normal argument. For instance, Meyers et al. (2004) assume that support verbs (a term related to light verbs) share their arguments with a noun. Chomsky (1981, p.37) calls advantage a quasi-argument of take in the idiom take advantage of. 2 Alonso Ramos (1998) proposes the role of quasi-object: this relationship holds between parts of idiomatic constructions, which is in accordance with Chomsky's usage of the term idiom. In this spirit, the term quasiargument might be extended to signal the relationship between the verbal and the nominal components of light verb constructions as well since they behave as a semantic unit, forming one complex predicate.",
"cite_spans": [
{
"start": 665,
"end": 693,
"text": "(Muischnek and Kaalep, 2010)",
"ref_id": "BIBREF16"
},
{
"start": 705,
"end": 735,
"text": "(Mansoory and Bijankhan, 2008)",
"ref_id": "BIBREF14"
},
{
"start": 814,
"end": 844,
"text": "(Mansoory and Bijankhan, 2008)",
"ref_id": "BIBREF14"
},
{
"start": 997,
"end": 1017,
"text": "Meyers et al. (2004)",
"ref_id": "BIBREF15"
},
{
"start": 1111,
"end": 1131,
"text": "Chomsky (1981, p.37)",
"ref_id": null
},
{
"start": 1214,
"end": 1226,
"text": "Ramos (1998)",
"ref_id": "BIBREF0"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions as Complex Predicates",
"sec_num": "3"
},
{
"text": "Higher-level NLP applications can also profit from this solution because the identification of light verb constructions can be enhanced in this way, which has impact on e.g. information extrac-tion (IE). For instance, in event extraction the parser should recognize the special status of the quasi-argument and treat it in a specific way as in the following sentence:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions as Complex Predicates",
"sec_num": "3"
},
{
"text": "Pete made a decision on his future.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions as Complex Predicates",
"sec_num": "3"
},
{
"text": "Thus, the following data can be yielded by the IE algorithm:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions as Complex Predicates",
"sec_num": "3"
},
{
"text": "EVENT: decision-making ARGUMENT 1 : Pete ARGUMENT 2 : his future Instead of:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions as Complex Predicates",
"sec_num": "3"
},
{
"text": "*EVENT: making ARGUMENT 1 : Pete ARGUMENT 2 : decision ARGUMENT 3 : his future",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions as Complex Predicates",
"sec_num": "3"
},
{
"text": "Thus, there is an event of decision-making, Pete is its subject and it is about his future (and not an event of making with the arguments decision, Pete and his future as it would be assumed if decision was not marked as a quasi-argument of the verb).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions as Complex Predicates",
"sec_num": "3"
},
{
"text": "In order to reach this way of representation, there are two possibilities. First, we employ linguistic preprocessing of the data (including dependency parsing), then an LVC detector is used and in a post-processing step after syntactic parsing, the special relation of the nominal and the verbal component should be marked, i.e. certain syntactic labels are overwritten. Second, we execute parsing in a way that the training dataset already contains LVC-specific syntactic labels, that is, it is the dependency parser that carries out LVC detection. In this paper, we experiment with both ways and present and evaluate our results.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Light Verb Constructions as Complex Predicates",
"sec_num": "3"
},
{
"text": "There have been a considerable number of studies on LVC detection for several languages. They have been automatically identified in several languages such as English (Cook et al., 2007; Tu and Roth, 2011) , Dutch (Van de Cruys and Moir\u00f3n, 2007) , Basque (Gurrutxaga and Alegria, 2011) and German (Evert and Kermes, 2003) just to mention a few.",
"cite_spans": [
{
"start": 166,
"end": 185,
"text": "(Cook et al., 2007;",
"ref_id": "BIBREF3"
},
{
"start": 186,
"end": 204,
"text": "Tu and Roth, 2011)",
"ref_id": "BIBREF21"
},
{
"start": 221,
"end": 244,
"text": "Cruys and Moir\u00f3n, 2007)",
"ref_id": "BIBREF22"
},
{
"start": 254,
"end": 284,
"text": "(Gurrutxaga and Alegria, 2011)",
"ref_id": "BIBREF11"
},
{
"start": 296,
"end": 320,
"text": "(Evert and Kermes, 2003)",
"ref_id": "BIBREF7"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "4"
},
{
"text": "We are aware of one machine learning system that identifies Hungarian LVCs in texts: the system described in selects LVC candidates from texts on the basis of syntactic information, then in a second step it classifies them as genuine LVCs or not, using morphological, lexical, syntactic and semantic features.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "4"
},
{
"text": "Regarding the methods they use, , Van de Cruys and Moir\u00f3n (2007) and Gurrutxaga and Alegria (2011) used statistical features for identifying LVCs. Others employed rule-based systems (Diab and Bhutada, 2009; Nagy T. et al., 2011) , which usually make use of (shallow) linguistic information. Some hybrid systems integrated both statistical and linguistic information as well (Tan et al., 2006; Tu and Roth, 2011) .",
"cite_spans": [
{
"start": 41,
"end": 64,
"text": "Cruys and Moir\u00f3n (2007)",
"ref_id": "BIBREF22"
},
{
"start": 69,
"end": 98,
"text": "Gurrutxaga and Alegria (2011)",
"ref_id": "BIBREF11"
},
{
"start": 182,
"end": 206,
"text": "(Diab and Bhutada, 2009;",
"ref_id": "BIBREF4"
},
{
"start": 207,
"end": 228,
"text": "Nagy T. et al., 2011)",
"ref_id": "BIBREF17"
},
{
"start": 374,
"end": 392,
"text": "(Tan et al., 2006;",
"ref_id": "BIBREF20"
},
{
"start": 393,
"end": 411,
"text": "Tu and Roth, 2011)",
"ref_id": "BIBREF21"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "4"
},
{
"text": "As we aim at identifying LVCs by applying a dependency parser, next we concentrate on studies that are based on syntactic information and are related to MWE extraction. Seretan (2011) developed a method for collocation extraction based on syntactic constraints. Wehrli et al. (2010) argued that collocations can highly contribute to the performance of the parser since many parsing ambiguities can be excluded if collocations are known and treated as one syntactic unit. Nivre and Nilsson (2004) analyzed the influence of (previous) MWE recognition on dependency parsing and showed that known MWEs have a beneficial effect on parsing results. Korkontzelos and Manandhar (2010) investigated whether known MWEs improve the performance of statistical shallow parsers and found that they can significantly contribute to the efficiency of parsing. Eryigit et al. (2011) analysed the impact of extracting MWEs on improving the accuracy of a dependency parser in Turkish. They found that the integration of compound verb and noun formations (which concept is similar to the one of light verb constructions applied here) has a detrimental effect on parsing accuracy since it increases lexical sparsity.",
"cite_spans": [
{
"start": 262,
"end": 282,
"text": "Wehrli et al. (2010)",
"ref_id": "BIBREF27"
},
{
"start": 471,
"end": 495,
"text": "Nivre and Nilsson (2004)",
"ref_id": "BIBREF18"
},
{
"start": 843,
"end": 864,
"text": "Eryigit et al. (2011)",
"ref_id": "BIBREF6"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "4"
},
{
"text": "As can be seen, many previous studies examined the effects of already identified MWEs on the efficiency of parsing. On the other hand, there have been some current studies that aim at experimenting in the other direction, namely, using parsers for identifying MWEs: constituency parsing models are employed in identifying contiguous MWEs in French and Arabic (Green et al., 2013) . Their method relied on a syntactic treebank, an MWE list and a morphological analyzer.",
"cite_spans": [
{
"start": 359,
"end": 379,
"text": "(Green et al., 2013)",
"ref_id": "BIBREF10"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "4"
},
{
"text": "In this paper, we also experiment in this area: we employ a dependency parser for identifying LVCs in Hungarian texts as a \"side effect\" of parsing sentences. Our dependency parser based method for identifying Hungarian LVCs is novel since to the best of our knowledge, dependency parsers have not been directly applied to identify LVCs. Moreover, it requires only a syntactic treebank enhanced with LVC annotation, in other words, there is no need to implement a separate LVC detector from scratch. In the following, we present our experiments and discuss our results.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "4"
},
{
"text": "In this section, we will present our corpus, our methodology for detecting light verb constructions and we will show our results.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Experiments",
"sec_num": "5"
},
{
"text": "The Szeged Constituency Treebank has been manually annotated for light verb constructions . This treebank exists in another manually annotated version, namely, with dependency annotation . Thus, manual annotations for LVCs and dependency structures are available for the same bunch of texts, which made it possible to map the two manual annotations. Thus, dependency relations were enhanced with LVC-specific relations that can be found between the two members of the constructions. For instance, instead of the traditional OBJ (object) relation, which occurred in the original version of the Szeged Dependency Treebank, the relation OBJ-LVC can be found between the words d\u00f6nt\u00e9st (decision-ACC) and hoz \"bring\", members of the LVC d\u00f6nt\u00e9st hoz \"to make a decision\" in the version used in this experiment. Here we provide a list of LVC-specific relations that occurred in our data (neglecting a handful of cases which were mislabeled due to some annotation errors in the dependency treebank):",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "\u2022 ATT-LVC -relation between a noun and a participial occurrence of a light verb:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "(a tegnap) adott tan\u00e1cs (the yesterday) given advice \"(the) advice that was given (yesterday)\"",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "\u2022 OBJ-LVC -relation between a light verb and its object:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "bejelent\u00e9st tesz announcement-ACC makes \"to make an announcement\"",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "\u2022 OBL-LVC -relation between a light verb and its nominal argument (which is not the subject or object or dative):",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "eletbe l\u00e9p life-ILL step \"to take effect\"",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "\u2022 SUBJ-LVC -relation between a light verb and its subject:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "sor ker\u00fcl (vmire)",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "turn get sg-SUB \"the time has come for sg\"",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "When mapping the LVC annotations and the dependency structures, we paid attention to the fact that it is only LVCs spelt as two tokens that could be identified with our methodology since no internal structure of compound words are marked in the Hungarian treebank and thus no dependency relation can be found among the members of the compound. So, we neglect LVCs spelt as one word and focus only on verbal and participial LVCs that consist of two members (cf. Footnote 1). Figure 1 shows an example of a sentence with and without LVC-specific dependency labels. As can be seen, we have the light verb construction d\u00f6nt\u00e9st hoz decision-ACC bring \"to make a decision\" in the sentence. However, it is parsed as a \"normal\" object of the verb in the first case (OBJ) and as a light verb object (OBJ-LVC) in the second case. Moreover, it is also seen that the two components of the LVC are not adjacent hence there are crossing branches in the dependency graph.",
"cite_spans": [],
"ref_spans": [
{
"start": 474,
"end": 482,
"text": "Figure 1",
"ref_id": null
}
],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "Although the entire Szeged Corpus contains manual LVC and dependency annotation, for the purpose of our study, we just selected texts from the law domain since they contain the biggest number of LVCs. Sentences in the law subcorpus were further filtered due to the fact that state-ofthe-art dependency parsers cannot adequately treat verbless sentences, hence verbless sentences were ignored (see Farkas et al. (2012) for a detailed discussion of the problem). After this filtering step, we experimented with 6173 sentences, which consist of 156,744 tokens and contain 1101 LVCs. We present statistical data on the frequency of the LVC-specific relations in Table 1 .",
"cite_spans": [
{
"start": 397,
"end": 417,
"text": "Farkas et al. (2012)",
"ref_id": "BIBREF8"
}
],
"ref_spans": [
{
"start": 658,
"end": 665,
"text": "Table 1",
"ref_id": null
}
],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "As Hungarian is a free word order language, the two components of LVCs, namely, the noun and the light verb, may not be adjacent in all cases, Figure 1 : Dependency graph of the sentence Holnap nagyon fontos d\u00f6nt\u00e9st kell hoznunk \"Tomorrow we will have to make a very important decision\" with or without LVC-specific dependency relations. which has a potentially detrimental effect on their identification in texts. Thus, we investigated the frequency of such cases in the data. Table 1 reveals that it is a quite frequent phenomenon in the corpus: almost one third of LVCs are non-contiguous. The largest distance between the noun and the verb is 21 tokens and the average distance between the two non-adjacent components is 4.28 tokens. All this suggests that sequence labeling approaches for LVC detection may not be as effective on the data as expected, however, a dependency parser that is able to identify long-distance dependencies may deal with the problem of non-adjacent but grammatically dependent elements in a more accurate way, which we will test below.",
"cite_spans": [],
"ref_spans": [
{
"start": 143,
"end": 151,
"text": "Figure 1",
"ref_id": null
},
{
"start": 478,
"end": 485,
"text": "Table 1",
"ref_id": null
}
],
"eq_spans": [],
"section": "The Corpus",
"sec_num": "5.1"
},
{
"text": "Farkas et al. (2012) carried out the first experiments on Hungarian dependency parsing. They empirically showed that state-of-the-art dependency parsers achieve similar results -in terms of attachment scores -on Hungarian and English.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Dependency Parsing for LVC Detection",
"sec_num": "5.2"
},
{
"text": "Although the results are not directly comparable due to domain differences and annotation schema divergences, they concluded that the difficulty of parsing Hungarian is very similar to parsing English and statistical dependency parsing is a viable way of parsing Hungarian, a morphologically rich language with free word order.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Dependency Parsing for LVC Detection",
"sec_num": "5.2"
},
{
"text": "As their results indicated, the Bohnet dependency parser (Bohnet, 2010) proved to be the most effective on Hungarian data (Farkas et al., 2012 ), thus we applied it in our experiments too. It is an efficient second order dependency parser that models the interaction between siblings as well as grandchildren. Its decoder works on labeled edges, i.e. it uses a single-step approach for obtaining labeled dependency trees. It uses a rich and well-engineered feature set and it is enhanced by a Hash Kernel, which leads to higher accuracy.",
"cite_spans": [
{
"start": 57,
"end": 71,
"text": "(Bohnet, 2010)",
"ref_id": "BIBREF1"
},
{
"start": 122,
"end": 142,
"text": "(Farkas et al., 2012",
"ref_id": "BIBREF8"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Dependency Parsing for LVC Detection",
"sec_num": "5.2"
},
{
"text": "Due to the free word order, there are quite many long-distance dependencies in Hungarian sentences, where a word and its parent are not adjacent (see also Figure 1 ). However, these linguistic phenomena are reasonably well-treated by dependency parsers. Furthermore, there seem to be quite a lot of non-contiguous LVCs in Hungarian. Hence, we think that these facts justify our experiments on applying a dependency parser for identifying LVCs.",
"cite_spans": [],
"ref_spans": [
{
"start": 155,
"end": 163,
"text": "Figure 1",
"ref_id": null
}
],
"eq_spans": [],
"section": "Dependency Parsing for LVC Detection",
"sec_num": "5.2"
},
{
"text": "We trained and evaluated the Bohnet parser on the data in a ten-fold cross validation manner. To evaluate the quality of the dependency parsing, we applied the Labeled Attachment Score (LAS) and Unlabeled Attachment Score (ULA) metrics, taking into account punctuation as well. On the other hand, we also employed F \u03b2=1 scores inter- preted on the LVC-specific relations to evaluate the performance of detecting LVCs in the corpus and we evaluated our system on contiguous and non-contiguous LVCs as well.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Methodology",
"sec_num": "5.3"
},
{
"text": "As baselines, we made use of the methods described in . They first employed dictionary matching, where LVCs collected from a parallel corpus annotated for Hungarian LVCs (Vincze, 2012) were mapped to the lemmatized texts.",
"cite_spans": [
{
"start": 170,
"end": 184,
"text": "(Vincze, 2012)",
"ref_id": "BIBREF26"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Methodology",
"sec_num": "5.3"
},
{
"text": "We also applied dictionary matching as one of our baselines. The main method of first parsed each sentence and extracted potential LVCs on the basis of the dependency relations found between verb-object, verb-subject, verbprepositional object, verb-other argument and noun-modifier pairs. The dependency labels were provided by magyarlanc (Zsibrita et al., 2013) . Later, C4.5 decision trees were applied to classify candidate LVCs, which exploits a rich feature set. For instance, morphological features exploited the fact that the nominal component of LVCs is typically derived from a verbal stem or coincides with a verb, on the other hand, the POS tags of the words and surrounding words were also used as features. As for semantic features, the activity or event semantic senses were looked for among the upper level hyperonyms of the head of the noun phrase in the Hungarian WordNet 3 . As lexical features, fifteen typical light verbs were selected from the list of the most frequent verbs taken from the Szeged ParalellFX corpus (Vincze, 2012) and it was checked whether the lemmatised verbal component of the candidate was one of these fifteen verbs. The lemma of the noun was also applied as a lexical feature.",
"cite_spans": [
{
"start": 339,
"end": 362,
"text": "(Zsibrita et al., 2013)",
"ref_id": "BIBREF28"
},
{
"start": 1037,
"end": 1051,
"text": "(Vincze, 2012)",
"ref_id": "BIBREF26"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Methodology",
"sec_num": "5.3"
},
{
"text": "We evaluated our database with this system too in a ten-fold cross validation manner (using the same data splits as previously) and as evaluation metrics, we employed F \u03b2=1 scores. The results of our experiments are shown in Tables 2 and 3 . Table 3 : Results on detecting contiguous and noncontiguous LVCs.",
"cite_spans": [],
"ref_spans": [
{
"start": 225,
"end": 239,
"text": "Tables 2 and 3",
"ref_id": "TABREF2"
},
{
"start": 242,
"end": 249,
"text": "Table 3",
"ref_id": null
}
],
"eq_spans": [],
"section": "Methodology",
"sec_num": "5.3"
},
{
"text": "As Table 2 shows, the dependency parser with the LVC-specific relations achieved an F-score of 0.7563 (recall: 0.6712, precision: 0.8660) interpreted on the LVC-specific relations. This result exceeds the ones obtained by the baselines: it outperforms the dictionary matching method by 54.38% in terms of F-score with a considerably better recall value, and, on the other hand, it also performs better than the classification method with a 1.18% gain in F-score -the results are significant (ANOVA, p = 0.012). In the latter case, the improvement is due to the higher precision value. The identification of non-contiguous LVCs proved to be more difficult for both methods than that of contiguous LVCs.",
"cite_spans": [],
"ref_spans": [
{
"start": 3,
"end": 10,
"text": "Table 2",
"ref_id": "TABREF2"
}
],
"eq_spans": [],
"section": "Results",
"sec_num": "6"
},
{
"text": "The classification approach significantly outperforms the dependency parser on the contiguous LVC class (ANOVA, p = 0.0455) but on the non-contiguous class the dependency parser performs significantly better with an F-score of 0.6401 (ANOVA, p = 0.0343).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Results",
"sec_num": "6"
},
{
"text": "In order to analyze the performance in more detail, we compared the precision, recall and Fscores for each LVC-specific label. Data in Table 4 reveal that SUBJ-LVCs are the easiest ones to predict (with both high precision and recall values) and participial uses of LVCs are the most difficult to identify (ATT-LVC) relation between the noun and the participle, mostly due to the low recall value. Although the precision value is rather low in the case of objects (OBJ-LVC), objects and other arguments (OBL-LVC) can be detected reasonably well. Table 5 shows results for (non-)contiguous LVC classes. It is revealed that for OBL-LVCs, there is no substantial difference between contiguous and non-contiguous LVCs but for objects and participial LVCs, the dis- Table 5 : Results in terms of precision, recall and F-score as predicted by the dependency parser for (non-)contiguous (NC/C) LVC classes.",
"cite_spans": [],
"ref_spans": [
{
"start": 135,
"end": 143,
"text": "Table 4",
"ref_id": "TABREF5"
},
{
"start": 547,
"end": 554,
"text": "Table 5",
"ref_id": null
},
{
"start": 762,
"end": 769,
"text": "Table 5",
"ref_id": null
}
],
"eq_spans": [],
"section": "Results",
"sec_num": "6"
},
{
"text": "tance between the two components of the LVC has an essential effect on the efficiency. 4 As for the performance on dependency parsing, we got 90.38 (LAS) and 92.12 (ULA) when training with LVC-specific relations. If these results are compared to those achieved with traditional (i.e. non-LVC-specific) relations, then it is revealed that in the latter case LAS is 90.63, i.e. 0.25 percentage point higher, which can be considered negligible.",
"cite_spans": [
{
"start": 87,
"end": 88,
"text": "4",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Results",
"sec_num": "6"
},
{
"text": "As the results show, the dependency parsing approach achieved the best results on LVC detection, especially due to the high precision score. This is probably due to the rich feature set applied by the Bohnet parser. Furthermore, our approach to solve the problem of LVC detection as a classification of syntactic constructions by using a dependency parser is also justified by these results.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Discussion",
"sec_num": "7"
},
{
"text": "A comparison with previous parser-based approach to MWE detection might also prove use-ful. Green et al. (2013) employed constituency parsers to identify contiguous MWEs in French and Arabic. As a main difference between our approach and theirs, we applied a dependency parser for the task of LVC detection, which proved especially effective since we worked with a free word order language, thus we had to deal with non-contiguous LVCs as well. Our dependency parser approach could adequately identify them as well, however, experimenting with a constituency parser will be a possible way to continue our work.",
"cite_spans": [
{
"start": 92,
"end": 111,
"text": "Green et al. (2013)",
"ref_id": "BIBREF10"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Discussion",
"sec_num": "7"
},
{
"text": "In Hungarian, it sometimes happens that a sequence that looks like an LVC is actually not an LVC in the specific context as in A d\u00e9k\u00e1n\u00fajabb el\u0151ad\u00e1st tartott sz\u00fcks\u00e9gesnek the dean new-COMP presentation-ACC hold-PAST-3SG necessary-DAT \"The dean thought that another presentation was necessary\". In other contexts, el\u0151ad\u00e1st tart presentation-ACC hold \"to have a presentation\" would most probably function as an LVC. However, in this case we encounter with another fixed grammatical construction of Hungarian, namely, valamilyennek tart valamit somewhat-DAT hold something-ACC \"to regard something as something\", e.g. sz\u00e9pnek tartja a l\u00e1nyt beautiful-DAT hold-3SG-OBJ the girl-ACC \"he thinks that the girl is beautiful\". Thus, there is no LVC in the above example, but approaches that heavily build on MWE lexicons may falsely identify this verbobject pair as a light verb object-light verb pair since they hardly consider contextual information. In contrast, dependency parsers have access to information about other dependents of the verb hence they may learn that in such cases the presence of a dative dependent argues against the identification of the verb-object pair as an LVC.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Discussion",
"sec_num": "7"
},
{
"text": "As for the specific LVC-relations, our approach was most successful on LVCs where the noun fulfilled the role of the subject (i.e. it had the relation SUBJ-LVC). This may be attributed to the fact that these LVCs are the least diverse in the corpus: there are only a handful of such types, and each LVC type has several occurrences in the data thus they can be easily identified. On the other hand, participial uses of LVCs (ATT-LVC) were the hardest to detect, which is partly due to their lexical divergence and partly due to the fact that currently adjectives and participles are not distinguished in Hungarian morphological parsing, i.e. they have the same morphological codes. Thus, the parser, which heavily builds on morpho-logical information, has no chance to learn that it is only participles that tend to occur as parts of LVCs but adjectives do not. A distinction of participles and adjectives in the Hungarian computational morphology would most probably have beneficial effects on identifying LVCs.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Discussion",
"sec_num": "7"
},
{
"text": "Our results empirically prove that a dependency parser may be effectively applied to identify LVCs in free texts, provided that we have a dependency model trained on LVC-specific relations, which itself requires a treebank manually annotated for dependency relations and LVCs. Although the LAS scores are somewhat lower than in the case of LVC-less dependency relations, the task of LVC detection can be also performed by the parser. On the other hand, the classification approach needs a trained dependency model since it classifies LVC candidates selected on the basis of syntactic information. It also uses LVC lists gathered from annotated corpora and in order to denote LVC-specific relations (i.e. quasi-arguments) in the case of complex predicates, an extra postprocessing step is needed in the workflow. Thus, the resources needed by the two approaches are the same but with the dependency parsing approach, the implementation of a new LVC-detector from scratch might be saved and complex predicates are provided immediately by the parser. Moreover, another advantage of the dependency parser is that it performs better on non-contiguous LVCs, which are frequent in Hungarian.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Discussion",
"sec_num": "7"
},
{
"text": "We also carried out an error analysis in order to compare the two methods. It was difficult for both the dependency parser and the classifier to recognize rare LVCs or those that included a nonfrequent light verb. A typical source of error for the dependency parser was that sometimes an LVC-specific relation was proposed for non-nouns (e.g. adverbs or conjunctions) as well, like in ak\u00e1r\u00edrnia (either write-INF.3SG) \"either he should write\", where ak\u00e1r was labeled as an LVC-object of the verb instead of a conjunction. Furthermore, the classifier often made an error in cases where the sentence included an LVC but another argument of the verb was labeled as part of the LVC, e.g. filmet forgalomba hoz (film-ACC circulation-INE bring) \"to put a film into circulation\", where the gold standard LVC is forgalomba hoz \"to put something into circulation\" but filmet hoz \"to bring a film\" was labeled as a false positive LVC. Since different phenomena proved to be difficult for the two systems, a possible direction for future work may be to combine the two approaches in order to minimize prediction errors.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Discussion",
"sec_num": "7"
},
{
"text": "Here we experimented with Hungarian, a morphologically rich language. Nevertheless, we believe that the method of applying a dependency parser for LVC detection is not specific to this typological class of languages and it can be employed for any language that has a dependency treebank which contains annotation for LVCs.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Discussion",
"sec_num": "7"
},
{
"text": "In this paper, we empirically showed that a dependency parser can be employed to detect LVCs in free texts. For this, we used a Hungarian treebank, which has been manually annotated for dependency relations and light verb constructions. Our results outperformed those achieved by state-ofthe-art techniques for Hungarian LVC detection and the main advantages of our system is its high precision on the one hand and the adequate treatment of non-contiguous LVCs on the other hand.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusions",
"sec_num": "8"
},
{
"text": "The error analysis of the systems applied suggests that since the two systems make errors in different cases, combining them may lead to more precise results. Another possible way of improving the system is to explore methods for the treatment of participial LVCs. Furthermore, as future work we aim at experimenting with the dependency parser in other scenarios (e.g. the newspaper subcorpus of the Szeged Dependency Treebank) in order to make further generalizations on the role of dependency parsing in LVC detection.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusions",
"sec_num": "8"
},
{
"text": "Due to some orthographical rules, certain nominal or participial occurrences of LVCs should be spelt as one word in Hungarian (such as tan\u00e1csad\u00f3 advice.giver \"consultant\"). These latter cases are not identifiable with syntax-based methods, only with morphological methods, thus we omit them from our investigations.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "In our view, take advantage of is a light verb construction rather than an idiom.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "http://www.inf.u-szeged.hu/rgai/HuWN",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "As there were hardly any non-contiguous SUBJ-LVCs in the dataset, we cannot draw any conclusions on the difficulty level of identifying non-contiguous light verb subjects.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
}
],
"back_matter": [
{
"text": "This work was supported in part by the European Union and the European Social Fund through the project FuturICT.hu (grant no.: T\u00c1MOP-4.2.2.C-11/1/KONV-2012-0013) and by the COST Action PARSEME (IC1207).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Acknowledgments",
"sec_num": null
}
],
"bib_entries": {
"BIBREF0": {
"ref_id": "b0",
"title": "Etude s\u00e9manticosyntaxique des constructions\u00e0 verbe support",
"authors": [
{
"first": "Margarita",
"middle": [
"Alonso"
],
"last": "Ramos",
"suffix": ""
}
],
"year": 1998,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Margarita Alonso Ramos. 1998. Etude s\u00e9mantico- syntaxique des constructions\u00e0 verbe support. Ph.D. thesis, Universit\u00e9 de Montr\u00e9al, Montreal, Canada.",
"links": null
},
"BIBREF1": {
"ref_id": "b1",
"title": "Top accuracy and fast dependency parsing is not a contradiction",
"authors": [
{
"first": "Bernd",
"middle": [],
"last": "Bohnet",
"suffix": ""
}
],
"year": 2010,
"venue": "Proceedings of Coling 2010",
"volume": "",
"issue": "",
"pages": "89--97",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Bernd Bohnet. 2010. Top accuracy and fast depen- dency parsing is not a contradiction. In Proceedings of Coling 2010, pages 89-97.",
"links": null
},
"BIBREF2": {
"ref_id": "b2",
"title": "Lectures on Government and Binding",
"authors": [
{
"first": "Noam",
"middle": [],
"last": "Chomsky",
"suffix": ""
}
],
"year": 1981,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Noam Chomsky. 1981. Lectures on Government and Binding. Foris, Dordrecht.",
"links": null
},
"BIBREF3": {
"ref_id": "b3",
"title": "Pulling their weight: exploiting syntactic forms for the automatic identification of idiomatic expressions in context",
"authors": [
{
"first": "Paul",
"middle": [],
"last": "Cook",
"suffix": ""
},
{
"first": "Afsaneh",
"middle": [],
"last": "Fazly",
"suffix": ""
},
{
"first": "Suzanne",
"middle": [],
"last": "Stevenson",
"suffix": ""
}
],
"year": 2007,
"venue": "Proceedings of MWE 2007",
"volume": "",
"issue": "",
"pages": "41--48",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Paul Cook, Afsaneh Fazly, and Suzanne Stevenson. 2007. Pulling their weight: exploiting syntactic forms for the automatic identification of idiomatic expressions in context. In Proceedings of MWE 2007, pages 41-48, Morristown, NJ, USA. ACL.",
"links": null
},
"BIBREF4": {
"ref_id": "b4",
"title": "Verb Noun Construction MWE Token Classification",
"authors": [
{
"first": "Mona",
"middle": [],
"last": "Diab",
"suffix": ""
},
{
"first": "Pravin",
"middle": [],
"last": "Bhutada",
"suffix": ""
}
],
"year": 2009,
"venue": "Proceedings of MWE 2009",
"volume": "",
"issue": "",
"pages": "17--22",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Mona Diab and Pravin Bhutada. 2009. Verb Noun Construction MWE Token Classification. In Pro- ceedings of MWE 2009, pages 17-22, Singapore, August. ACL.",
"links": null
},
"BIBREF5": {
"ref_id": "b5",
"title": "The Syntax of Hungarian",
"authors": [
{
"first": "",
"middle": [],
"last": "Katalin\u00e9",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Kiss",
"suffix": ""
}
],
"year": 2002,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Katalin\u00c9. Kiss. 2002. The Syntax of Hungarian. Cambridge University Press, Cambridge.",
"links": null
},
"BIBREF6": {
"ref_id": "b6",
"title": "Multiword expressions in statistical dependency parsing",
"authors": [
{
"first": "G\u00fcl\u015fen",
"middle": [],
"last": "Eryigit",
"suffix": ""
},
{
"first": "Tugay",
"middle": [],
"last": "Ilbay",
"suffix": ""
},
{
"first": "Ozan Arkan",
"middle": [],
"last": "Can",
"suffix": ""
}
],
"year": 2011,
"venue": "Proceedings of SPMRL 2011",
"volume": "",
"issue": "",
"pages": "45--55",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "G\u00fcl\u015fen Eryigit, Tugay Ilbay, and Ozan Arkan Can. 2011. Multiword expressions in statistical depen- dency parsing. In Proceedings of SPMRL 2011, pages 45-55, Dublin, Ireland. ACL.",
"links": null
},
"BIBREF7": {
"ref_id": "b7",
"title": "Experiments on candidate data for collocation extraction",
"authors": [
{
"first": "Stefan",
"middle": [],
"last": "Evert",
"suffix": ""
},
{
"first": "Hannah",
"middle": [],
"last": "Kermes",
"suffix": ""
}
],
"year": 2003,
"venue": "Proceedings of EACL 2003",
"volume": "",
"issue": "",
"pages": "83--86",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Stefan Evert and Hannah Kermes. 2003. Experiments on candidate data for collocation extraction. In Pro- ceedings of EACL 2003, pages 83-86.",
"links": null
},
"BIBREF8": {
"ref_id": "b8",
"title": "Dependency Parsing of Hungarian: Baseline Results and Challenges",
"authors": [
{
"first": "Rich\u00e1rd",
"middle": [],
"last": "Farkas",
"suffix": ""
},
{
"first": "Veronika",
"middle": [],
"last": "Vincze",
"suffix": ""
},
{
"first": "Helmut",
"middle": [],
"last": "Schmid",
"suffix": ""
}
],
"year": 2012,
"venue": "Proceedings of EACL 2012",
"volume": "",
"issue": "",
"pages": "55--65",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Rich\u00e1rd Farkas, Veronika Vincze, and Helmut Schmid. 2012. Dependency Parsing of Hungarian: Baseline Results and Challenges. In Proceedings of EACL 2012, pages 55-65, Avignon, France, April. ACL.",
"links": null
},
"BIBREF9": {
"ref_id": "b9",
"title": "Distinguishing Subtypes of Multiword Expressions Using Linguistically-Motivated Statistical Measures",
"authors": [
{
"first": "Afsaneh",
"middle": [],
"last": "Fazly",
"suffix": ""
},
{
"first": "Suzanne",
"middle": [],
"last": "Stevenson",
"suffix": ""
}
],
"year": 2007,
"venue": "Proceedings of MWE 2007",
"volume": "",
"issue": "",
"pages": "9--16",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Afsaneh Fazly and Suzanne Stevenson. 2007. Distin- guishing Subtypes of Multiword Expressions Using Linguistically-Motivated Statistical Measures. In Proceedings of MWE 2007, pages 9-16, Prague, Czech Republic, June. ACL.",
"links": null
},
"BIBREF10": {
"ref_id": "b10",
"title": "Parsing models for identifying multiword expressions",
"authors": [
{
"first": "Spence",
"middle": [],
"last": "Green",
"suffix": ""
},
{
"first": "Marie-Catherine",
"middle": [],
"last": "De Marneffe",
"suffix": ""
},
{
"first": "Christopher",
"middle": [
"D"
],
"last": "Manning",
"suffix": ""
}
],
"year": 2013,
"venue": "Computational Linguistics",
"volume": "39",
"issue": "1",
"pages": "195--227",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Spence Green, Marie-Catherine de Marneffe, and Christopher D. Manning. 2013. Parsing models for identifying multiword expressions. Computational Linguistics, 39(1):195-227.",
"links": null
},
"BIBREF11": {
"ref_id": "b11",
"title": "Automatic Extraction of NV Expressions in Basque: Basic Issues on Cooccurrence Techniques",
"authors": [
{
"first": "Antton",
"middle": [],
"last": "Gurrutxaga",
"suffix": ""
},
{
"first": "I\u00f1aki",
"middle": [],
"last": "Alegria",
"suffix": ""
}
],
"year": 2011,
"venue": "Proceedings of MWE 2011",
"volume": "",
"issue": "",
"pages": "2--7",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Antton Gurrutxaga and I\u00f1aki Alegria. 2011. Auto- matic Extraction of NV Expressions in Basque: Basic Issues on Cooccurrence Techniques. In Pro- ceedings of MWE 2011, pages 2-7, Portland, Ore- gon, USA, June. ACL.",
"links": null
},
"BIBREF12": {
"ref_id": "b12",
"title": "Statistical Modeling of Multiword Expressions",
"authors": [
{
"first": "Kim",
"middle": [],
"last": "Su Nam",
"suffix": ""
}
],
"year": 2008,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Su Nam Kim. 2008. Statistical Modeling of Multiword Expressions. Ph.D. thesis, University of Melbourne, Melbourne.",
"links": null
},
"BIBREF13": {
"ref_id": "b13",
"title": "Can recognising multiword expressions improve shallow parsing?",
"authors": [
{
"first": "Ioannis",
"middle": [],
"last": "Korkontzelos",
"suffix": ""
},
{
"first": "Suresh",
"middle": [],
"last": "Manandhar",
"suffix": ""
}
],
"year": 2010,
"venue": "NAACL HLT 2010",
"volume": "",
"issue": "",
"pages": "636--644",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Ioannis Korkontzelos and Suresh Manandhar. 2010. Can recognising multiword expressions improve shallow parsing? In NAACL HLT 2010, pages 636- 644, Los Angeles, California, June. ACL.",
"links": null
},
"BIBREF14": {
"ref_id": "b14",
"title": "The possible effects of Persian light verb constructions on Persian WordNet",
"authors": [
{
"first": "Niloofar",
"middle": [],
"last": "Mansoory",
"suffix": ""
},
{
"first": "Mahmood",
"middle": [],
"last": "Bijankhan",
"suffix": ""
}
],
"year": 2008,
"venue": "Proceedings of GWC 2008",
"volume": "",
"issue": "",
"pages": "297--303",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Niloofar Mansoory and Mahmood Bijankhan. 2008. The possible effects of Persian light verb construc- tions on Persian WordNet. In Proceedings of GWC 2008, pages 297-303, Szeged, Hungary, January. University of Szeged.",
"links": null
},
"BIBREF15": {
"ref_id": "b15",
"title": "The NomBank Project: An Interim Report",
"authors": [
{
"first": "Adam",
"middle": [],
"last": "Meyers",
"suffix": ""
},
{
"first": "Ruth",
"middle": [],
"last": "Reeves",
"suffix": ""
},
{
"first": "Catherine",
"middle": [],
"last": "Macleod",
"suffix": ""
},
{
"first": "Rachel",
"middle": [],
"last": "Szekely",
"suffix": ""
},
{
"first": "Veronika",
"middle": [],
"last": "Zielinska",
"suffix": ""
},
{
"first": "Brian",
"middle": [],
"last": "Young",
"suffix": ""
},
{
"first": "Ralph",
"middle": [],
"last": "Grishman",
"suffix": ""
}
],
"year": 2004,
"venue": "HLT-NAACL 2004 Workshop: Frontiers in Corpus Annotation",
"volume": "",
"issue": "",
"pages": "24--31",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Adam Meyers, Ruth Reeves, Catherine Macleod, Rachel Szekely, Veronika Zielinska, Brian Young, and Ralph Grishman. 2004. The NomBank Project: An Interim Report. In HLT-NAACL 2004 Work- shop: Frontiers in Corpus Annotation, pages 24-31, Boston, Massachusetts, USA. ACL.",
"links": null
},
"BIBREF16": {
"ref_id": "b16",
"title": "The variability of multi-word verbal expressions in Estonian. Language Resources and Evaluation",
"authors": [
{
"first": "Kadri",
"middle": [],
"last": "Muischnek",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Heiki Jaan Kaalep",
"suffix": ""
}
],
"year": 2010,
"venue": "",
"volume": "44",
"issue": "",
"pages": "115--135",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Kadri Muischnek and Heiki Jaan Kaalep. 2010. The variability of multi-word verbal expressions in Esto- nian. Language Resources and Evaluation, 44(1- 2):115-135.",
"links": null
},
"BIBREF17": {
"ref_id": "b17",
"title": "Domain-dependent identification of multiword expressions",
"authors": [
{
"first": "Istv\u00e1n",
"middle": [],
"last": "Nagy",
"suffix": ""
},
{
"first": "T",
"middle": [],
"last": "",
"suffix": ""
},
{
"first": "Veronika",
"middle": [],
"last": "Vincze",
"suffix": ""
},
{
"first": "G\u00e1bor",
"middle": [],
"last": "Berend",
"suffix": ""
}
],
"year": 2011,
"venue": "Proceedings of RANLP 2011",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Istv\u00e1n Nagy T., Veronika Vincze, and G\u00e1bor Berend. 2011. Domain-dependent identification of multi- word expressions. In Proceedings of RANLP 2011, Hissar, Bulgaria.",
"links": null
},
"BIBREF18": {
"ref_id": "b18",
"title": "Multiword units in syntactic parsing",
"authors": [
{
"first": "Joakim",
"middle": [],
"last": "Nivre",
"suffix": ""
},
{
"first": "Jens",
"middle": [],
"last": "Nilsson",
"suffix": ""
}
],
"year": 2004,
"venue": "Proceedings of MEMURA",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Joakim Nivre and Jens Nilsson. 2004. Multiword units in syntactic parsing. In Proceedings of MEMURA 2004.",
"links": null
},
"BIBREF19": {
"ref_id": "b19",
"title": "Syntax-Based Collocation Extraction",
"authors": [
{
"first": "Violeta",
"middle": [],
"last": "Seretan",
"suffix": ""
}
],
"year": 2011,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Violeta Seretan. 2011. Syntax-Based Collocation Extraction. TSD. Springer, Dordrecht.",
"links": null
},
"BIBREF20": {
"ref_id": "b20",
"title": "Extending corpus-based identification of light verb constructions using a supervised learning framework",
"authors": [
{
"first": "Min-Yen",
"middle": [],
"last": "Yee Fan Tan",
"suffix": ""
},
{
"first": "Hang",
"middle": [],
"last": "Kan",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Cui",
"suffix": ""
}
],
"year": 2006,
"venue": "Proceedings of MWE 2006",
"volume": "",
"issue": "",
"pages": "49--56",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Yee Fan Tan, Min-Yen Kan, and Hang Cui. 2006. Extending corpus-based identification of light verb constructions using a supervised learning frame- work. In Proceedings of MWE 2006, pages 49-56, Trento, Italy, April. ACL.",
"links": null
},
"BIBREF21": {
"ref_id": "b21",
"title": "Learning English Light Verb Constructions: Contextual or Statistical",
"authors": [
{
"first": "Yuancheng",
"middle": [],
"last": "Tu",
"suffix": ""
},
{
"first": "Dan",
"middle": [],
"last": "Roth",
"suffix": ""
}
],
"year": 2011,
"venue": "Proceedings of MWE 2011",
"volume": "",
"issue": "",
"pages": "31--39",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Yuancheng Tu and Dan Roth. 2011. Learning English Light Verb Constructions: Contextual or Statistical. In Proceedings of MWE 2011, pages 31-39, Port- land, Oregon, USA, June. ACL.",
"links": null
},
"BIBREF22": {
"ref_id": "b22",
"title": "Semantics-based multiword expression extraction",
"authors": [
{
"first": "Tim",
"middle": [],
"last": "Van De Cruys",
"suffix": ""
},
{
"first": "Bego\u00f1a",
"middle": [],
"last": "Villada Moir\u00f3n",
"suffix": ""
}
],
"year": 2007,
"venue": "Proceedings of MWE 2007",
"volume": "",
"issue": "",
"pages": "25--32",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Tim Van de Cruys and Bego\u00f1a Villada Moir\u00f3n. 2007. Semantics-based multiword expression extraction. In Proceedings of MWE 2007, pages 25-32, Mor- ristown, NJ, USA. ACL.",
"links": null
},
"BIBREF23": {
"ref_id": "b23",
"title": "Hungarian corpus of light verb constructions",
"authors": [
{
"first": "Veronika",
"middle": [],
"last": "Vincze",
"suffix": ""
},
{
"first": "J\u00e1nos",
"middle": [],
"last": "Csirik",
"suffix": ""
}
],
"year": 2010,
"venue": "Proceedings of Coling 2010",
"volume": "",
"issue": "",
"pages": "1110--1118",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Veronika Vincze and J\u00e1nos Csirik. 2010. Hungarian corpus of light verb constructions. In Proceedings of Coling 2010, pages 1110-1118, Beijing, China, August. Coling 2010 Organizing Committee.",
"links": null
},
"BIBREF24": {
"ref_id": "b24",
"title": "Hungarian Dependency Treebank",
"authors": [
{
"first": "Veronika",
"middle": [],
"last": "Vincze",
"suffix": ""
},
{
"first": "D\u00f3ra",
"middle": [],
"last": "Szauter",
"suffix": ""
},
{
"first": "Attila",
"middle": [],
"last": "Alm\u00e1si",
"suffix": ""
},
{
"first": "Gy\u00f6rgy",
"middle": [],
"last": "M\u00f3ra",
"suffix": ""
},
{
"first": "J\u00e1nos",
"middle": [],
"last": "Zolt\u00e1n Alexin",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Csirik",
"suffix": ""
}
],
"year": 2010,
"venue": "Proceedings of LREC 2010",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Veronika Vincze, D\u00f3ra Szauter, Attila Alm\u00e1si, Gy\u00f6rgy M\u00f3ra, Zolt\u00e1n Alexin, and J\u00e1nos Csirik. 2010. Hun- garian Dependency Treebank. In Proceedings of LREC 2010, Valletta, Malta, May. ELRA.",
"links": null
},
"BIBREF25": {
"ref_id": "b25",
"title": "Identifying English and Hungarian Light Verb Constructions: A Contrastive Approach",
"authors": [
{
"first": "Veronika",
"middle": [],
"last": "Vincze",
"suffix": ""
},
{
"first": "Istv\u00e1n",
"middle": [],
"last": "Nagy",
"suffix": ""
},
{
"first": "T",
"middle": [],
"last": "",
"suffix": ""
},
{
"first": "Rich\u00e1rd",
"middle": [],
"last": "Farkas",
"suffix": ""
}
],
"year": 2013,
"venue": "Proceedings of ACL-2013: Short Papers",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Veronika Vincze, Istv\u00e1n Nagy T., and Rich\u00e1rd Farkas. 2013. Identifying English and Hungarian Light Verb Constructions: A Contrastive Approach. In Pro- ceedings of ACL-2013: Short Papers, Sofia. ACL.",
"links": null
},
"BIBREF26": {
"ref_id": "b26",
"title": "Light Verb Constructions in the SzegedParalellFX English-Hungarian Parallel Corpus",
"authors": [
{
"first": "Veronika",
"middle": [],
"last": "Vincze",
"suffix": ""
}
],
"year": 2012,
"venue": "Proceedings of LREC 2012",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Veronika Vincze. 2012. Light Verb Constructions in the SzegedParalellFX English-Hungarian Paral- lel Corpus. In Proceedings of LREC 2012, Istanbul, Turkey.",
"links": null
},
"BIBREF27": {
"ref_id": "b27",
"title": "Sentence analysis and collocation identification",
"authors": [
{
"first": "Eric",
"middle": [],
"last": "Wehrli",
"suffix": ""
},
{
"first": "Violeta",
"middle": [],
"last": "Seretan",
"suffix": ""
},
{
"first": "Luka",
"middle": [],
"last": "Nerima",
"suffix": ""
}
],
"year": 2010,
"venue": "Proceedings of MWE 2010",
"volume": "",
"issue": "",
"pages": "28--36",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Eric Wehrli, Violeta Seretan, and Luka Nerima. 2010. Sentence analysis and collocation identification. In Proceedings of MWE 2010, pages 28-36, Beijing, China, August. Coling 2010 Organizing Committee.",
"links": null
},
"BIBREF28": {
"ref_id": "b28",
"title": "magyarlanc: A Toolkit for Morphological and Dependency Parsing of Hungarian",
"authors": [
{
"first": "J\u00e1nos",
"middle": [],
"last": "Zsibrita",
"suffix": ""
},
{
"first": "Veronika",
"middle": [],
"last": "Vincze",
"suffix": ""
},
{
"first": "Rich\u00e1rd",
"middle": [],
"last": "Farkas",
"suffix": ""
}
],
"year": 2013,
"venue": "Proceedings of RANLP-2013",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "J\u00e1nos Zsibrita, Veronika Vincze, and Rich\u00e1rd Farkas. 2013. magyarlanc: A Toolkit for Morphological and Dependency Parsing of Hungarian. In Proceedings of RANLP-2013, Hissar, Bulgaria.",
"links": null
}
},
"ref_entries": {
"TABREF2": {
"text": "Results on LVC detection.",
"type_str": "table",
"html": null,
"content": "<table/>",
"num": null
},
"TABREF5": {
"text": "Distribution of relations in the gold standard data and results in terms of precision, recall and F-score as predicted by the dependency parser.",
"type_str": "table",
"html": null,
"content": "<table><tr><td colspan=\"3\">Relation & type Precision Recall F-score</td></tr><tr><td>ATT-LVC C</td><td>0.9524</td><td>0.4878 0.6452</td></tr><tr><td>ATT-LVC NC</td><td>0.6667</td><td>0.3667 0.4731</td></tr><tr><td>OBJ-LVC C</td><td>0.8535</td><td>0.7507 0.7988</td></tr><tr><td>OBJ-LVC NC</td><td>0.8025</td><td>0.5478 0.6512</td></tr><tr><td>OBL-LVC C</td><td>0.9226</td><td>0.7176 0.8073</td></tr><tr><td>OBL-LVC NC</td><td>0.8947</td><td>0.6800 0.7727</td></tr><tr><td>SUBJ-LVC C</td><td>0.9785</td><td>0.9286 0.9529</td></tr><tr><td>SUBJ-LVC NC</td><td>0.6000</td><td>0.7500 0.6667</td></tr></table>",
"num": null
}
}
}
} |