File size: 110,607 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 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 |
{
"paper_id": "2019",
"header": {
"generated_with": "S2ORC 1.0.0",
"date_generated": "2023-01-19T07:29:29.004796Z"
},
"title": "A Survey on Ontology Enrichment from Text",
"authors": [
{
"first": "Vivek",
"middle": [],
"last": "Iyer",
"suffix": "",
"affiliation": {},
"email": ""
},
{
"first": "Y",
"middle": [],
"last": "Raghu",
"suffix": "",
"affiliation": {},
"email": ""
},
{
"first": "Reddy",
"middle": [],
"last": "Iiit",
"suffix": "",
"affiliation": {},
"email": ""
},
{
"first": "Gachibowli",
"middle": [],
"last": "Hyderabad",
"suffix": "",
"affiliation": {},
"email": ""
},
{
"first": "Mehar",
"middle": [],
"last": "Bhatia",
"suffix": "",
"affiliation": {},
"email": ""
}
],
"year": "",
"venue": null,
"identifiers": {},
"abstract": "Increased internet bandwidth at low cost is leading to the creation of large volumes of unstructured data. This data explosion opens up opportunities for the creation of a variety of data-driven intelligent systems, such as the Semantic Web. Ontologies form one of the most crucial layers of semantic web, and the extraction and enrichment of ontologies given this data explosion becomes an inevitable research problem. In this paper, we survey the literature on semi-automatic and automatic ontology extraction and enrichment and classify them into four broad categories based on the approach. Then, we proceed to narrow down four algorithms from each of these categories, implement and analytically compare them based on parameters like context relevance, efficiency and precision. Lastly, we propose a Long Short Term Memory Networks (LSTM) based deep learning approach to try and overcome the gaps identified in these approaches.",
"pdf_parse": {
"paper_id": "2019",
"_pdf_hash": "",
"abstract": [
{
"text": "Increased internet bandwidth at low cost is leading to the creation of large volumes of unstructured data. This data explosion opens up opportunities for the creation of a variety of data-driven intelligent systems, such as the Semantic Web. Ontologies form one of the most crucial layers of semantic web, and the extraction and enrichment of ontologies given this data explosion becomes an inevitable research problem. In this paper, we survey the literature on semi-automatic and automatic ontology extraction and enrichment and classify them into four broad categories based on the approach. Then, we proceed to narrow down four algorithms from each of these categories, implement and analytically compare them based on parameters like context relevance, efficiency and precision. Lastly, we propose a Long Short Term Memory Networks (LSTM) based deep learning approach to try and overcome the gaps identified in these approaches.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Abstract",
"sec_num": null
}
],
"body_text": [
{
"text": "There has been an explosion of data on the Internet in the past few years, primarily caused by the drastic increase in the number of internet users over the years. About 90% of the data on internet has been created since 2016, mainly because of the massive increase in the user base and machine to machine communication. Data is defined as unprocessed facts and figures that do not contain any added interpretation or analysis. Information is interpretation of structured or unstructured data so that it holds meaning. Knowledge is processed information, experience, and insight combined such that it is beneficial to the end user 1 .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Web pages, the primary source of knowledge on the World Wide Web (WWW) are primarily 1 https://tinyurl.com/datainfknowledge text documents annotated using Hypertext Markup Language (HTML). Lack of semantic markup of pages can result in irrelevant search results. The semantic web 2 provides a format or structure to machines to understand the meaning of the web page data rather relying on HTML markup, to make web intelligent and intuitive to user's queries.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "The semantic web includes data-centric publishing languages, including RDF (Resource Description Framework -the data modeling language for the semantic web), SPARQL (SPARQL protocol and RDF query language for semantic web) and OWL (Web Ontology Language -schema language, or knowledge representation language, of the semantic web), which allows meaning and structure to be added to content in a machine-readable format.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "OWL 3 allows definition of concepts composably, i.e. in such a way that it allows the reuse of concepts and relationships. Given the amount of information being extracted from the data generated on a regular basis in various domains, it becomes essential for it to be stored in the form of knowledge in ontologies.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "However, the knowledge stored in ontologies is rarely static. Like all other knowledge structures, its vital for ontologies to be enriched with time so as to improve the quality of search results.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Given recent advances in the fields of artificial intelligence and machine learning, as well as increased data processing capabilities with increase in compute power, newer, better and more accurate ways of extracting and enriching ontologies from text are now possible. Ontology extraction from text has primarily been at lower layers in ontology \"layer cake\" (Buitelaar et al., 2005) . A pre-existing seed ontology created manually or through learning needs enrichment. Ontology enrichment (Faatz and Steinmetz, 2002) is population, updation, and adaptation (Noy and Klein, 2004) of concepts, relations and rules. In the context of this paper, we assume a pre-existing seed ontology that is enriched by learning (semi-automatic or automatic) from text. The survey in this paper attempts to address the following important research questions:",
"cite_spans": [
{
"start": 361,
"end": 385,
"text": "(Buitelaar et al., 2005)",
"ref_id": "BIBREF6"
},
{
"start": 492,
"end": 519,
"text": "(Faatz and Steinmetz, 2002)",
"ref_id": "BIBREF19"
},
{
"start": 560,
"end": 581,
"text": "(Noy and Klein, 2004)",
"ref_id": "BIBREF33"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "\u2022 How are ontologies enriched by learning from unstructured text, and which algorithms are considered seminal?",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "\u2022 How do seminal algorithms compare with each other, in regards to context relevance, algorithmic efficiency and precision? \u2022 What gaps are identified in these algorithms, and how can they be potentially addressed?",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "We did not focus on the other knowledge representation methods such as knowledge graphs, frames, semantic nets and others in the survey considering the extensivity of ontology research and the generalizability of research trends to other knowledge representation methods. The further sections of the document contain our literature survey approach for identifying the state-of-the-art in section 2; we explain the broad genres identified in the ontology extraction in section 3; we proceed with a critical analysis of the major approaches through the years, by analyzing the algorithms on context relevance, efficiency and precision in section 4; we propose a deep learning based methodology (LSTM -Long Short Term Memory) to possibly overcome the gaps in the ontology enrichment in section 5 and finally end with a conclusion summarizing our observations.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "We started the review on ontology learning from text before focusing on enrichment. Research on ontology learning from text started in 1995 (Mahesh et al., 1995) but still continues to be an area of interest. In the last two decades, there have been 20 survey papers on ontology tools, learning, evolution, construction, enrichment, change, generation, population, and matching with text. The large count of survey papers indicates the growing interest among researchers and changing research approaches in ontology learning. We classified these survey papers 4 on the basis of text format (structured or unstructured), evaluation methods, ontology layer cake, AI techniques, level of automation, etc. Most survey papers recommended human intervention, continued automation, gold standards and graphical interfaces for improved quality, expressiveness and scalability. While the survey papers were thorough, there weren't any papers that follow the systematic literature review (SLR) or systematic mapping process (Kitchenham, 2004) , or any that discussed seminal papers that led to change of approaches.",
"cite_spans": [
{
"start": 140,
"end": 161,
"text": "(Mahesh et al., 1995)",
"ref_id": "BIBREF29"
},
{
"start": 1014,
"end": 1032,
"text": "(Kitchenham, 2004)",
"ref_id": "BIBREF25"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Approach for Literature Review",
"sec_num": "2"
},
{
"text": "Based on our study of the survey papers' classification methods and future directions, the keywords for search from digital libraries were \"Extraction\", \"Evolution\", \"Enrichment\", \"Maintain\", and \"Learning\" along with \"Ontology\" as keyword. We did not follow SLR process as our objective was to analyze the seminal papers based on context relevance, precision and algorithm efficiency. The input to our survey process consisted of 166 research papers extracted from ScienceDirect, Springer, IEEE, and ACM digital libraries from 1990-2018 time period. After reviewing the abstract and conclusion, 65 papers were eliminated from the list as they were thesis, patents, grey material, non-English, position or tutorial papers and others. The papers related to construction of data, text summarization using ontologies, machine translation, Information Retrieval, etc, of the extracted research papers were also excluded from further analysis. While there were about 23 domains for validation, Medical and Education domains were the most referred domains in the shortlisted papers. The Figure 1 (Y-axis is the count of papers and X-axis is the year of publication) on ontology learning depicts the ongoing interest of researchers. The study on approaches of the shortlisted papers stated that although natural language processing and description logic continue to be used; Word2Vec, a step towards deep learning is being more leveraged for ontology learning. The shortlisted papers were categorized after reading the abstract, introduction and conclusion, as shown in Figure 2 . The papers on \"create\" were related to ontology construction or population or generation. The papers on \"update\" were related to ontology evolution, enrichment, updation, refinement, maintain, etc. The papers on \"CRUD\" operation dealt with creation, updation and deletion of redundant concepts and relations as well. For further analysis, the papers on \"update\" and \"CRUD\" on ontology were clustered into 4 categories based on the approach used for enrichment. ",
"cite_spans": [],
"ref_spans": [
{
"start": 1081,
"end": 1089,
"text": "Figure 1",
"ref_id": "FIGREF0"
},
{
"start": 1563,
"end": 1572,
"text": "Figure 2",
"ref_id": "FIGREF1"
}
],
"eq_spans": [],
"section": "Approach for Literature Review",
"sec_num": "2"
},
{
"text": "We proceeded with a review of the 23 shortlisted papers of the 4 categories in ontology enrichment.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Categories in Ontology Enrichment",
"sec_num": "3"
},
{
"text": "Some of the earliest papers in the field of ontology enrichment from text, adopted similarity-based clustering approach. A hierarchical clustering algorithm to classify ontology-based metadata (Maedche and Zacharias, 2002) was proposed in 2002. Later, a similarity-based clustering approach was proposed to identify concepts in a gene ontology (Cheng et al., 2004) . The unsupervised guided hierarchical clustering algorithm (Cimiano and Staab, 2005) uses an oracle of hypernyms derived from WordNet, text and WWW corpora for clustering concepts in a hierarchy. The fuzzy inference mechanism (Lee et al., 2007) uses fuzzy numbers that calculate the conceptual similarity between concepts to obtain new learning instances.",
"cite_spans": [
{
"start": 193,
"end": 222,
"text": "(Maedche and Zacharias, 2002)",
"ref_id": "BIBREF28"
},
{
"start": 344,
"end": 364,
"text": "(Cheng et al., 2004)",
"ref_id": "BIBREF8"
},
{
"start": 425,
"end": 450,
"text": "(Cimiano and Staab, 2005)",
"ref_id": "BIBREF12"
},
{
"start": 592,
"end": 610,
"text": "(Lee et al., 2007)",
"ref_id": "BIBREF27"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Similarity Based Clustering Algorithms:",
"sec_num": "3.1"
},
{
"text": "These algorithms used a set-theoretic approach to order concepts. Harris's distributional hypothesis (Sahlgren, 2008) modeled the context of a certain word with its dependencies, and on the basis of this information, Formal Concept Analysis (FCA) (Cimiano et al., 2005a ) outputs a concept lattice which is then converted into a concept hierarchy. Also, algorithms and transformations that combine FCA and the Horn model (Ben-Khalifa and Motameny, 2007) of a concept lattice have been proposed (Haav, 2004) . A fuzzy extension of FCA (De Maio et al., 2009) described an approach for automatic elicitation of ontologies by web analysis. It also formalized a method that generated an OWL-based representation of concepts, individuals and properties.",
"cite_spans": [
{
"start": 101,
"end": 117,
"text": "(Sahlgren, 2008)",
"ref_id": "BIBREF34"
},
{
"start": 247,
"end": 269,
"text": "(Cimiano et al., 2005a",
"ref_id": "BIBREF10"
},
{
"start": 494,
"end": 506,
"text": "(Haav, 2004)",
"ref_id": "BIBREF22"
},
{
"start": 534,
"end": 556,
"text": "(De Maio et al., 2009)",
"ref_id": "BIBREF16"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Set Theoretic Based Algorithms",
"sec_num": "3.2"
},
{
"text": "Relational Concept Analysis (RCA) (Hacene et al., 2008) constructs ontologies in a semi-automated manner by translating concept lattices with interrelated elements to concepts and relations in the ontology. RCA is an extension of FCA that allows for the processing of multi-relational datasets, each with its own set of attributes and relationships amongst themselves.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Set Theoretic Based Algorithms",
"sec_num": "3.2"
},
{
"text": "Web corpus Based Algorithms used web as a big data corpus to overcome problems of data sparsity.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Web Corpus Based Algorithms",
"sec_num": "3.3"
},
{
"text": "The categories and labels from Wikipedia were used to classify concepts (Cui et al., 2009; Ahmed et al., 2012; Medelyan et al., 2009) leveraging N-grams and other related NLP algorithms. The Open Linked Data (Booshehri and Luksch, 2014) , a freely available source of semantic knowledge is used as a skeleton to construct ontologies (Tiddi et al., 2012) . DBPedia, another crowd-sourced Linked Data dataset that extracts structured information from Wikipedia is used to enrich ontology (Booshehri and Luksch, 2015) .",
"cite_spans": [
{
"start": 72,
"end": 90,
"text": "(Cui et al., 2009;",
"ref_id": "BIBREF15"
},
{
"start": 91,
"end": 110,
"text": "Ahmed et al., 2012;",
"ref_id": "BIBREF0"
},
{
"start": 111,
"end": 133,
"text": "Medelyan et al., 2009)",
"ref_id": "BIBREF30"
},
{
"start": 208,
"end": 236,
"text": "(Booshehri and Luksch, 2014)",
"ref_id": "BIBREF4"
},
{
"start": 333,
"end": 353,
"text": "(Tiddi et al., 2012)",
"ref_id": "BIBREF37"
},
{
"start": 486,
"end": 514,
"text": "(Booshehri and Luksch, 2015)",
"ref_id": "BIBREF5"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Web Corpus Based Algorithms",
"sec_num": "3.3"
},
{
"text": "An automatic and unsupervised methodology that uses the Web to learn ontological concept properties, or attributes, and attribute restrictions, was proposed (S\u00e1nchez, 2010) . In the \"Self Annotating Web\", globally available knowledge, or syntactic resources, were used for the creation of metadata, the basic idea being that the statistical distribution of syntactic structures on the web can be used to approximate semantics. One such algorithm that implemented this paradigm is called PANKOW (Pattern-based Annotation through Knowledge On the Web) (Cimiano et al., 2004) , in which patterns were instantiated from schemata and the number of hits of related entities for each concept were counted. C-PANKOW (Cimiano et al., 2005b) , or Context-driven PANKOW that outperforms its predecessor, PANKOW by downloading abstracts offline, performing linguistic analysis and using the context to resolve ambiguity.",
"cite_spans": [
{
"start": 157,
"end": 172,
"text": "(S\u00e1nchez, 2010)",
"ref_id": "BIBREF35"
},
{
"start": 550,
"end": 572,
"text": "(Cimiano et al., 2004)",
"ref_id": "BIBREF9"
},
{
"start": 708,
"end": 731,
"text": "(Cimiano et al., 2005b)",
"ref_id": "BIBREF11"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Web Corpus Based Algorithms",
"sec_num": "3.3"
},
{
"text": "In recent years, learning algorithms driven by feedback from domain experts have gained popularity. OntoAMAS (Benomrane et al., 2016) tool is based on adaptive multi-agent system (AMAS) for ontology enrichment and makes proposals based on ontologists' feedback. Also noteworth, is the Probabilistic Relational Hierarchy Extraction technique based on Probabilistic Relational Concept Extraction (Drumond and Girardi, 2010) to extract concepts and the taxonomic relationships from inference on Markov Logic Networks. Group storytelling technique has been used (Confort et al., 2015) to gather knowledge from those involved in the field in the first phase, which makes the system learn the concepts for an ontology automatically. OntoHarvester system (Mousavi et al., 2014) used deep NLP-based algorithms to mine text and extract domain-specific ontologies by iteratively extracting ontological relations that link the concepts in the ontology to the terms in the text, out of which strongly connected concepts were added to the ontology. The Automated Ontology Generation Framework (Alobaidi et al., 2018) , used Linked Biomedical Ontologies, various NLP techniques (in text processing based on \"Compute on Demand\" method, N-Grams, ontology linking and classification), semantic enrichment (using RDF mining), syntactic pattern and graph-based techniques (to extract relations), and domain inference engine (to build the formal ontology).",
"cite_spans": [
{
"start": 109,
"end": 133,
"text": "(Benomrane et al., 2016)",
"ref_id": "BIBREF3"
},
{
"start": 394,
"end": 421,
"text": "(Drumond and Girardi, 2010)",
"ref_id": "BIBREF17"
},
{
"start": 575,
"end": 580,
"text": "2015)",
"ref_id": "BIBREF14"
},
{
"start": 748,
"end": 770,
"text": "(Mousavi et al., 2014)",
"ref_id": "BIBREF31"
},
{
"start": 1080,
"end": 1103,
"text": "(Alobaidi et al., 2018)",
"ref_id": "BIBREF1"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Learning based Algorithms",
"sec_num": "3.4"
},
{
"text": "They also proposed Linked Biomedical Ontologies as a promising solution towards automating the ontology generation process in the disease-drug domain.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Learning based Algorithms",
"sec_num": "3.4"
},
{
"text": "Word2Vec was used (Wohlgenannt and Minic, 2016) to extract similar meaning terms or concepts and to get certain semantic and syntactic relations based on simple vector operations.",
"cite_spans": [
{
"start": 18,
"end": 47,
"text": "(Wohlgenannt and Minic, 2016)",
"ref_id": "BIBREF38"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Learning based Algorithms",
"sec_num": "3.4"
},
{
"text": "The (Cimiano et al., 2005b) 3. Set Theoretic Based Algorithms: Constructing a concept hierarchy using Formal Concept Analysis (Cimiano et al., 2005a) 4. Deep Learning Based Algorithms: The Word2Vec-based algorithm (Wohlgenannt and Minic, 2016) In this section, we performed an in-depth analysis of these algorithms and compared their performance based on ontology evaluation (Netzer et al., 2009) methods like contextual relevance, precision and algorithmic efficiency.",
"cite_spans": [
{
"start": 4,
"end": 27,
"text": "(Cimiano et al., 2005b)",
"ref_id": "BIBREF11"
},
{
"start": 126,
"end": 149,
"text": "(Cimiano et al., 2005a)",
"ref_id": "BIBREF10"
},
{
"start": 214,
"end": 243,
"text": "(Wohlgenannt and Minic, 2016)",
"ref_id": "BIBREF38"
},
{
"start": 375,
"end": 396,
"text": "(Netzer et al., 2009)",
"ref_id": "BIBREF32"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Learning based Algorithms",
"sec_num": "3.4"
},
{
"text": "The guided agglomerative clustering algorithm (Cimiano and Staab, 2005 ) has a citation count of 99 and published in 2005. The paper is based on Harris's distributional hypothesis and works by clustering concepts based on their similarities. Hypernym oracle extracted using different methods is the driving factor in the clustering process. Hypernyms oracle is constructed with Figure 3 : Hearst Patterns (Cimiano and Staab, 2005) the help of Hearst Patterns (Hearst, 1992) . Hearst Patterns 3, used Noun Phrases (NPs) consisting of a determiner, an optional adjective sequence and a common noun sequence which constitutes the NP head. The hypernym oracle H(t) is constructed using the following three sources: 1. WordNet: Uses synsets from WordNet for extracting hypernyms 2. Text Corpus using Hearst Patterns: Hearst Patterns were matched against the underlying text corpus, by using a regular expression comprising of POS tags to match Noun Phrases, thus constructing an is-a relation between the two terms. 3. WWW Corpus using Hearst Patterns: Every concept of interest is instantiated in a Hearst Pattern to form queries to Google API, and the abstracts from the results were downloaded offline. Hearst patterns were matched against these abstracts similar to how they were matched in the text corpus, and is-a relations were extracted accordingly.",
"cite_spans": [
{
"start": 46,
"end": 70,
"text": "(Cimiano and Staab, 2005",
"ref_id": "BIBREF12"
},
{
"start": 405,
"end": 430,
"text": "(Cimiano and Staab, 2005)",
"ref_id": "BIBREF12"
},
{
"start": 459,
"end": 473,
"text": "(Hearst, 1992)",
"ref_id": "BIBREF24"
}
],
"ref_spans": [
{
"start": 378,
"end": 386,
"text": "Figure 3",
"ref_id": null
}
],
"eq_spans": [],
"section": "Guided Agglomerative Clustering",
"sec_num": "4.1"
},
{
"text": "The algorithm takes a list of words to be clustered as input.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Guided Agglomerative Clustering",
"sec_num": "4.1"
},
{
"text": "Once the hypernym oracle was constructed, each of these terms were paired up and sorted in the descending order of similarity. The clustering algorithm used the oracle to construct parent-child or sibling relationships between these terms.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Guided Agglomerative Clustering",
"sec_num": "4.1"
},
{
"text": "After this step, the unclassified terms were classified using the r-matches relation.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Guided Agglomerative Clustering",
"sec_num": "4.1"
},
{
"text": "Though WordNet provides easy and accurate hypernyms, it is not extensive and has a very limited scope. It does not classify proper nouns or infrequently occurring terms, leading to most instances remaining unclassified leading to sparsity and scalability issues.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Guided Agglomerative Clustering",
"sec_num": "4.1"
},
{
"text": "Moreover, matching Hearst Patterns had very bad precision ( 13%), as shown in Figure 4 and outputs a lot of noisy data. This is due to the algorithm paying no attention to context relevance and extracting hypernyms that were irrelevant to a domain. The same word that had different meanings in different contexts (for instance, bank -which could refer to a river bank or a blood bank or a financial bank) were clustered together. In addition, this approach disregarded a lot of relevant relations because it relied on an exact syntactic pattern match that pays no attention to semantics.",
"cite_spans": [],
"ref_spans": [
{
"start": 78,
"end": 86,
"text": "Figure 4",
"ref_id": "FIGREF2"
}
],
"eq_spans": [],
"section": "Guided Agglomerative Clustering",
"sec_num": "4.1"
},
{
"text": "The C-PANKOW algorithm (Cimiano et al., 2005b) again by Cimiano et al. has a citation count of 246. The algorithm was based on the paradigm of \"Learning by Googling\". In this paradigm, given an instance, evidence was collected from the internet for the possible concepts. Then, either the instance was mapped to the concept with maximum evidence, or alternatively, an engineer with domain-specific knowledge does mapping manually. The PANKOW (Pattern-based Annotation through Knowledge on the Web) algorithm (Cimiano et al., 2004) , the predecessor of the C-PANKOW algorithm, instantiated a query using pre-defined patterns or regular expressions. A one-to-one mapping was done between each concept and instance to generate a query from these patterns. This query, similar to how the hypernym oracle was extracted using the WWW corpus in Guided Clustering, was made available to the Google API and the number of hits for this query were counted. Based on the statistical web fingerprint, or the total number of search results for each entity, the instance were mapped to the concept to get disambiguation by maximal evidence. The statistical web fingerprint were presented to the knowledge engineer to review and take the final decision.",
"cite_spans": [
{
"start": 23,
"end": 46,
"text": "(Cimiano et al., 2005b)",
"ref_id": "BIBREF11"
},
{
"start": 508,
"end": 530,
"text": "(Cimiano et al., 2004)",
"ref_id": "BIBREF9"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "However, PANKOW had a few disadvantages. Firstly, it issued a large number of requests to the Google Web API, which is proportional to the number of ontology concepts, so it does not scale well for large ontologies. Also, because of the restrictions inherent in the generation of patterns, many actual instances were not found.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "C-PANKOW addresses some issues by downloading results of queries, or the abstracts, and then doing the pattern matching locally by linguistic analysis.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "Downloading web pages reduced the number of requests made to Google Web API and the network traffic by issuing a constant number of queries per instance. In addition, it factors context into consideration and calculates the contextual similarity between two pages before doing concept-instance mapping, which reduced ambiguity especially in cases where a word has multiple meanings and its meaning depends on context. C-PANKOW presented a novel idea to concept extraction by combining the approaches of maximum frequency-based mapping and document similarity-based filtering.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "Frequency-based mapping reduces noise and gives only the most relevant relations, whereas similarity-based filtering using Doc2Vec (Lau and Baldwin, 2016) helps partially address the issue of context relevance by preemptively filtering out irrelevant abstracts.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "These two approaches augmented C-PANKOW's precision ( 36%) to be more than that of Guided Clustering. The filtering also increased algorithmic efficiency as, unlike Guided Clustering, it does not look for matches in irrelevant documents.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "However, despite its advantages, since C-PANKOW (like Guided Clustering) uses naive syntactic pattern matching to extract hypernymy relations, it does yield noisy data as well, whilst ignoring relevant results. This is because: a) The pattern matching fails to take semantics and language structure into consideration. b) It is also ineffective in situations where the concept being referred were already defined in an earlier sentence c) Though Doc2Vec does partially address the issue of context relevance at the document level, it does not check the relevance at the sentence or paragraph level, resulting in noisy data as well.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "To address the concerns of disambiguation in concepts or relations, agent based models have been proposed for the enrichment of ontologies (Sellami et al., 2013) .",
"cite_spans": [
{
"start": 139,
"end": 161,
"text": "(Sellami et al., 2013)",
"ref_id": "BIBREF36"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "An agent has local knowledge about itself and other neighbour agents, as well as about the lexical terms and concepts extracted from the corpus. It uses this knowledge to evaluate its own relevance in the ontology and manage its relationships with other agents. When new documents are added to the corpus, or when the ontologist suggests changes to the ontology proposed by the MAS, there were perturbations or disturbances caused in the system. Each agent in the MAS reacts to these perturbations by modifying its relations with other agents, updating its knowledge on and/or communicating with other agents in order to reach a stable state. On reaching this stable state, the MAS proposes a new version of the ontology which is once again presented to the ontologist. The ontologist suggests changes again and this whole process continues iteratively till the MAS reaches a final state where the ontology is not challenged by him anymore. In DYNAMO-MAS (Sellami et al., 2013) word disambiguation is handled by the Teminological Ontological Resource (TOR) model which comprises of a conceptual component (the ontology) and a lexical component (the terminology). Terms were attached to concepts by denotation links and contain a confidence score. These denotation links can be changed by the agents if a request with a higher confidence score is made. Thus, any term is attached by a denotation link to the concept with the highest confidence score. Since the same term can have different meanings in different context, the TOR model is able to disambiguate the meaning using these confidence scores. However, the confidence score is partly generated from a pattern score, which in turn has to be manually defined from empirical evaluations. Moreover, the ontologist has to manually verify the annotations proposed by the MAS which in turn means the text corpus has to be limited to a few hundred documents and cannot work on the larger web corpus. Thus while this approach makes a massive progress towards solving the issue of context relevance, it suffers from scalability issues.",
"cite_spans": [
{
"start": 955,
"end": 977,
"text": "(Sellami et al., 2013)",
"ref_id": "BIBREF36"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "4.3 Constructing concepts using FCA (Cimiano et al., 2005a) has 693 citations and is the primary source for research on FCA from text corpora.",
"cite_spans": [
{
"start": 36,
"end": 59,
"text": "(Cimiano et al., 2005a)",
"ref_id": "BIBREF10"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "The algorithm was based on Set-Theoretic approach that uses FCA to convert a partial order to a concept hierarchy on the basis of syntactic dependencies taken as features. With NLTK, the Part of Speech (POS) tags are extracted, separated into chunks, reduced to base-form (lemmatized), smoothed to overcome data sparseness, weighted, and only those terms with values above a threshold are converted into a formal context (Ganter and Wille, 1999) . FCA (Ganter and Wille, 1996) is then applied to this context to transform into a partial order, which is then compacted to remove abstract concepts and get the final concept hierarchy. This algorithmic approach uses pseudo-syntactic dependencies to extract concepts from the parse tree. Hence, it significantly outperforms Guided Clustering and C-PANKOW in terms of precision. This algorithm forms clusters and also provides an intentional description for them, leading to better understanding. However, this algorithm does not identify labels that describe the intention of a specific cluster, resulting in sparsely populated concepts.",
"cite_spans": [
{
"start": 421,
"end": 445,
"text": "(Ganter and Wille, 1999)",
"ref_id": "BIBREF21"
},
{
"start": 452,
"end": 476,
"text": "(Ganter and Wille, 1996)",
"ref_id": "BIBREF20"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "In addition, it is inefficient as construction of a separate concept lattice for every document is time expensive. Thus while it is more efficient than Clustering, it loses out to C-PANKOW in efficiency. However, the greater precision does shows that enriching contextual features using pseudo-syntactic dependencies is a viable alternative that outperforms enriching from parse trees.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "C-PANKOW algorithm",
"sec_num": "4.2"
},
{
"text": "Word2Vec, a 2-layer neural network has also been used to build a sample ontology learning system (Wohlgenannt and Minic, 2016) . The neural nets are trained on the linguistic context by Word2Vec, using two methods: Continuous Bag of Words (CBoW) and skip grams. CBoW is used to predict the context of a word, given the word, while skip grams predict the context given the word as input.",
"cite_spans": [
{
"start": 97,
"end": 126,
"text": "(Wohlgenannt and Minic, 2016)",
"ref_id": "BIBREF38"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Word2Vec-based algorithm",
"sec_num": "4.4"
},
{
"text": "Word2Vec allows vector operations, and is trained to output high quality similar terms given any input term. The Word2Vec model can be trained on the Google News corpus on any other large corpus.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Word2Vec-based algorithm",
"sec_num": "4.4"
},
{
"text": "The algorithm provided higher percentage of relevant concepts that can be used to enrich the ontology. In addition to having greater precision (60%) and efficiency than the previous algorithms, this algorithm makes headway in solving the issue of contextual relevance by using CBoW and Skip Grams to train the model. However, it does have a few drawbacks. Firstly, for terms that aren't encountered by the model in training corpus, a word embedding is not constructed, hence, concepts remaining unclustered. Secondly, Word2Vec doesn't have any shared representations at sub-word levels. It represents each word as an independent vector, though there could be morphologically similar terms. It also detects concepts that are too close to the original term, like plurals and synonyms which are unneccessarily added to the ontology as separate concepts. Lastly, it necessitates manual intervention after every iteration, unlike the previous algorithms, which in turn means it suffers from scalability issues.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Word2Vec-based algorithm",
"sec_num": "4.4"
},
{
"text": "The Guided Agglomerative Clustering algorithms used Wordnet and Hearst Patterns on corpora to build its hypernym oracle. While Wordnet is able to provide hypernyms for common nouns, it cannot handle proper nouns and phrases, which are often the primary focus while enriching domain specific ontologies. Using Hearst Patterns is inefficient too and results in a lot of noise, due to pattern being matching being purely syntactic with no attention paid to context. Though C-PANKOW is able to improve on precision, efficiency and also partially address the issue of context relevance (using a mixture of frequency-based mapping and document similarity scores), it uses naive syntactic pattern matching which results in selecting irrelevant terms and dropping relevant ones. The DYNAMO-MAS algorithm, despite solving disambiguation and having better precision, has serious limitations like data sparsity and unscalability. FCA, which uses pseudo-syntactic dependencies, was found to have better precision than both Clustering and C-PANKOW. But construction of a concept hierarchy is time inefficient, which is where it loses out to C-PANKOW. The Word2Vec algorithm was able to improve the problems of efficiency, precision and data sparsity by using word embeddings and skip-grams, and was found to outperform previously mentioned algorithms. However, this algorithm also suffers from some shortcomings like the inability to handle previously unencountered words, selecting of too similar terms, scalability issues due to manual intervention etc.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Discussion",
"sec_num": "5"
},
{
"text": "We used the 'Information Security' ontology (Ekelhart et al., 2006 ) based on ISO 27001 for comparing the algorithms. Figure 4 shows comparison of the metrics across these algorithms. All these algorithms have an area of improvement when the current concept and its pronouns are being extracted from text. In the previous approaches, the attributes and relations were mapped to the pronouns and not the concept itself. To address this gap, the algorithm needs to retain in memory the concept being extracted, instead of using naive pattern-matching approaches. Also, the analysis of the shortlisted research papers in each category state the declining research on Clustering and Set-Theoretic algorithms, and an increasing in research of learning algorithms.",
"cite_spans": [
{
"start": 44,
"end": 66,
"text": "(Ekelhart et al., 2006",
"ref_id": "BIBREF18"
}
],
"ref_spans": [
{
"start": 118,
"end": 126,
"text": "Figure 4",
"ref_id": "FIGREF2"
}
],
"eq_spans": [],
"section": "Discussion",
"sec_num": "5"
},
{
"text": "The algorithms proposed above use either pattern matching techniques, naive SVO (subject-verb-object) triplet extraction techniques or semantic similarity techniques for extracting concepts. All of these techniques were at the concept level, and though extension of algorithms like C-PANKOW used Doc2Vec to gauge similarity of documents, none of these algorithms involved understanding of the text corpus to filter out irrelevant data. Hence, we suggest a need to incorporate Deep Learning to enrich ontologies.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Possible solution: Long Short Term Memory Networks",
"sec_num": "5.1"
},
{
"text": "We propose a Deep Learning solution using Long Short Term Memory Networks (LSTMs) 5 to address the identified gaps. We explain our reason for proposing an LSTM with the help of an example.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Possible solution: Long Short Term Memory Networks",
"sec_num": "5.1"
},
{
"text": "\"Cross-Frame Scripting (XFS) is a browser based attack that combines malicious JavaScript with an iframe while loading a legitimate site. This attack is one of the most common attacks against IE. This is due to it leaking keyboard events across HTML framesets.\".",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Possible solution: Long Short Term Memory Networks",
"sec_num": "5.1"
},
{
"text": "On passing these sentences to a concept-relationship extraction system, such as the ones described previously, a \"one-of\" relationship would be formed between \"This attack\" and \"one of the most common attacks against IE\" and a \"due-to\" relationship would be formed between \"this\" and \"leaking keyboard events across HTML framesets\". However, in the second sentence, \"This attack\" refers to \"XFS attack\" (from the first sentence) and is the concept identified and can be abstracted to \"browser based attack\". But in the third sentence, the current concept has changed and \"this\" refers to \"attacks against IE\". Hence, normal concept extraction techniques would not work for these examples, since the current concept may change every sentence. LSTMs can be trained to learn optimal forget matrices that continually update the cell state, thereby, enabling the model to maintain the state of a concept (by addding new concepts and removing old ones) for longer durations. Thus, LSTMs can enable greater semantic understanding as well as detection of long ranging patterns, which theoretically should improve precision.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Possible solution: Long Short Term Memory Networks",
"sec_num": "5.1"
},
{
"text": "We started this survey paper by describing the need for enrichment of ontologies. We proceeded to survey the existing domain literature in the field of ontology learning from text and got a subset of 166 research papers and 20 survey papers. From shortlisted 101 papers, we narrowed down to the 23 most relevant research papers. These 23 papers were classified into four categories based on the approach used for ontology enrichment, namely Clustering, Set-Theoretic, Web Corpus-based and Learning-based Algorithms.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusion",
"sec_num": "6"
},
{
"text": "We selected a seminal paper from each category, based on criteria like the date of publication, the number of citations, relevance to our end goal etc. and then described the approach of the algorithms. Next, we compared algorithms performance (context relevance, precision and efficiency) on the enrichment of Information Security ontology. We found that with each trend, some of the gaps were overcome but there still remained the problem of retaining concepts to improve relevance in a scalable manner. We proposed LSTMs as a possible solution for concept retention, since they use a memory state to partially remember/forget concepts over long periods of time as require. In future, we plan on implementing the proposed LSTM model to improve precision and efficiency of the state-of-the-art. We also plan to validate further with complex ontologies, and extend our concept enrichment model to the addition of instances for building knowledge base.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusion",
"sec_num": "6"
},
{
"text": "https://expertsystem.com/what-is-the-semantic-web/ 3 https://db-x.org/blog/2016/04/15/semantic-web-2/",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "https://tinyurl.com/OntoSurvey",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "https://colah.github.io/posts/2015-08-Understanding-LSTMs/",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
}
],
"back_matter": [],
"bib_entries": {
"BIBREF0": {
"ref_id": "b0",
"title": "Effective Ontology Learning: Concepts' Hierarchy Building using Plain Text Wikipedia",
"authors": [
{
"first": "Adil",
"middle": [],
"last": "Khalida Bensidi Ahmed",
"suffix": ""
},
{
"first": "Mimoun",
"middle": [],
"last": "Toumouh",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Malki",
"suffix": ""
}
],
"year": 2012,
"venue": "ICWIT",
"volume": "",
"issue": "",
"pages": "170--178",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Khalida Bensidi Ahmed, Adil Toumouh, and Mimoun Malki. 2012. Effective Ontology Learning: Concepts' Hierarchy Building using Plain Text Wikipedia. In ICWIT, pages 170-178.",
"links": null
},
"BIBREF1": {
"ref_id": "b1",
"title": "Automated Ontology Generation Framework Powered by Linked Biomedical Ontologies for Disease-Drug Domain",
"authors": [
{
"first": "Mazen",
"middle": [],
"last": "Alobaidi",
"suffix": ""
},
{
"first": "Khalid Mahmood",
"middle": [],
"last": "Malik",
"suffix": ""
},
{
"first": "Maqbool",
"middle": [],
"last": "Hussain",
"suffix": ""
}
],
"year": 2018,
"venue": "Computer Methods and Programs in Biomedicine",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Mazen Alobaidi, Khalid Mahmood Malik, and Maqbool Hussain. 2018. Automated Ontology Generation Framework Powered by Linked Biomedical Ontologies for Disease-Drug Domain. Computer Methods and Programs in Biomedicine.",
"links": null
},
"BIBREF2": {
"ref_id": "b2",
"title": "Horn-Representation of a Concept Lattice",
"authors": [
{
"first": "Kamel",
"middle": [],
"last": "Ben-Khalifa",
"suffix": ""
},
{
"first": "Susanne",
"middle": [],
"last": "Motameny",
"suffix": ""
}
],
"year": 2007,
"venue": "",
"volume": "38",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Kamel Ben-Khalifa and Susanne Motameny. 2007. Horn-Representation of a Concept Lattice. volume 38.",
"links": null
},
"BIBREF3": {
"ref_id": "b3",
"title": "An ontologist Feedback driven Ontology Evolution with an Adaptive Multi-agent System",
"authors": [
{
"first": "Souad",
"middle": [],
"last": "Benomrane",
"suffix": ""
},
{
"first": "Zied",
"middle": [],
"last": "Sellami",
"suffix": ""
},
{
"first": "Mounir",
"middle": [],
"last": "Ben Ayed",
"suffix": ""
}
],
"year": 2016,
"venue": "Advanced Engineering Informatics",
"volume": "30",
"issue": "3",
"pages": "337--353",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Souad Benomrane, Zied Sellami, and Mounir Ben Ayed. 2016. An ontologist Feedback driven Ontology Evolution with an Adaptive Multi-agent System. Advanced Engineering Informatics, 30(3):337-353.",
"links": null
},
"BIBREF4": {
"ref_id": "b4",
"title": "Towards Adding Linked Data to Ontology Learning Layers",
"authors": [
{
"first": "Meisam",
"middle": [],
"last": "Booshehri",
"suffix": ""
},
{
"first": "Peter",
"middle": [],
"last": "Luksch",
"suffix": ""
}
],
"year": 2014,
"venue": "Proceedings of the 16th International Conference on Information Integration and Webbased Applications & Services",
"volume": "",
"issue": "",
"pages": "401--409",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Meisam Booshehri and Peter Luksch. 2014. Towards Adding Linked Data to Ontology Learning Layers. In Proceedings of the 16th International Conference on Information Integration and Web- based Applications & Services, pages 401-409. ACM.",
"links": null
},
"BIBREF5": {
"ref_id": "b5",
"title": "An Ontology Enrichment Approach by using DbPedia",
"authors": [
{
"first": "Meisam",
"middle": [],
"last": "Booshehri",
"suffix": ""
},
{
"first": "Peter",
"middle": [],
"last": "Luksch",
"suffix": ""
}
],
"year": 2015,
"venue": "Proceedings of the International Conference on Web Intelligence, Mining and Semantics",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Meisam Booshehri and Peter Luksch. 2015. An Ontology Enrichment Approach by using DbPedia. In Proceedings of the International Conference on Web Intelligence, Mining and Semantics, page 5. ACM.",
"links": null
},
"BIBREF6": {
"ref_id": "b6",
"title": "Ontology Learning from Text: An overview",
"authors": [
{
"first": "Paul",
"middle": [],
"last": "Buitelaar",
"suffix": ""
},
{
"first": "Philipp",
"middle": [],
"last": "Cimiano",
"suffix": ""
},
{
"first": "Bernardo",
"middle": [],
"last": "Magnini",
"suffix": ""
}
],
"year": 2005,
"venue": "Ontology Learning from Text: Methods, Evaluation and Applications",
"volume": "123",
"issue": "",
"pages": "3--12",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Paul Buitelaar, Philipp Cimiano, and Bernardo Magnini. 2005. Ontology Learning from Text: An overview. Ontology Learning from Text: Methods, Evaluation and Applications, 123:3-12.",
"links": null
},
"BIBREF7": {
"ref_id": "b7",
"title": "Ontology Learning with Deep Learning: a Case Study on Patient Safety Using PubMed",
"authors": [
{
"first": "Mercedes",
"middle": [],
"last": "Arg\u00fcello Casteleiro",
"suffix": ""
},
{
"first": "Maria Jesus Fernandez",
"middle": [],
"last": "Prieto",
"suffix": ""
},
{
"first": "George",
"middle": [],
"last": "Demetriou",
"suffix": ""
},
{
"first": "Nava",
"middle": [],
"last": "Maroto",
"suffix": ""
},
{
"first": "J",
"middle": [],
"last": "Warren",
"suffix": ""
},
{
"first": "Diego",
"middle": [],
"last": "Read",
"suffix": ""
},
{
"first": "Jose Julio Des",
"middle": [],
"last": "Maseda-Fernandez",
"suffix": ""
},
{
"first": "Goran",
"middle": [],
"last": "Diz",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Nenadic",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "John",
"suffix": ""
},
{
"first": "Robert",
"middle": [],
"last": "Keane",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Stevens",
"suffix": ""
}
],
"year": 2016,
"venue": "SWAT4LS",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Mercedes Arg\u00fcello Casteleiro, Maria Jesus Fernandez Prieto, George Demetriou, Nava Maroto, Warren J Read, Diego Maseda-Fernandez, Jose Julio Des Diz, Goran Nenadic, John A Keane, and Robert Stevens. 2016. Ontology Learning with Deep Learning: a Case Study on Patient Safety Using PubMed. In SWAT4LS.",
"links": null
},
"BIBREF8": {
"ref_id": "b8",
"title": "A Knowledge-based Clustering Algorithm Driven by Gene Ontology",
"authors": [
{
"first": "Jill",
"middle": [],
"last": "Cheng",
"suffix": ""
},
{
"first": "Melissa",
"middle": [],
"last": "Cline",
"suffix": ""
},
{
"first": "John",
"middle": [],
"last": "Martin",
"suffix": ""
},
{
"first": "David",
"middle": [],
"last": "Finkelstein",
"suffix": ""
},
{
"first": "Tarif",
"middle": [],
"last": "Awad",
"suffix": ""
},
{
"first": "David",
"middle": [],
"last": "Kulp",
"suffix": ""
},
{
"first": "Michael",
"middle": [
"A"
],
"last": "Siani-Rose",
"suffix": ""
}
],
"year": 2004,
"venue": "Journal of Biopharmaceutical statistics",
"volume": "14",
"issue": "3",
"pages": "687--700",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Jill Cheng, Melissa Cline, John Martin, David Finkelstein, Tarif Awad, David Kulp, and Michael A Siani-Rose. 2004. A Knowledge-based Clustering Algorithm Driven by Gene Ontology. Journal of Biopharmaceutical statistics, 14(3):687-700.",
"links": null
},
"BIBREF9": {
"ref_id": "b9",
"title": "Towards the self-annotating web",
"authors": [
{
"first": "Philipp",
"middle": [],
"last": "Cimiano",
"suffix": ""
},
{
"first": "Siegfried",
"middle": [],
"last": "Handschuh",
"suffix": ""
},
{
"first": "Steffen",
"middle": [],
"last": "Staab",
"suffix": ""
}
],
"year": 2004,
"venue": "Proceedings of the 13th international conference on World Wide Web",
"volume": "",
"issue": "",
"pages": "462--471",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Philipp Cimiano, Siegfried Handschuh, and Steffen Staab. 2004. Towards the self-annotating web. In Proceedings of the 13th international conference on World Wide Web, pages 462-471. ACM.",
"links": null
},
"BIBREF10": {
"ref_id": "b10",
"title": "Learning Concept Hierarchies from Text Corpora using Formal Concept Analysis",
"authors": [
{
"first": "Philipp",
"middle": [],
"last": "Cimiano",
"suffix": ""
},
{
"first": "Andreas",
"middle": [],
"last": "Hotho",
"suffix": ""
},
{
"first": "Steffen",
"middle": [],
"last": "Staab",
"suffix": ""
}
],
"year": 2005,
"venue": "Journal of Artificial Intelligence Research",
"volume": "24",
"issue": "",
"pages": "305--339",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Philipp Cimiano, Andreas Hotho, and Steffen Staab. 2005a. Learning Concept Hierarchies from Text Corpora using Formal Concept Analysis. Journal of Artificial Intelligence Research, 24:305-339.",
"links": null
},
"BIBREF11": {
"ref_id": "b11",
"title": "Gimme'the Context: Context-driven Automatic Semantic Annotation with C-PANKOW",
"authors": [
{
"first": "Philipp",
"middle": [],
"last": "Cimiano",
"suffix": ""
},
{
"first": "G\u00fcnter",
"middle": [],
"last": "Ladwig",
"suffix": ""
},
{
"first": "Steffen",
"middle": [],
"last": "Staab",
"suffix": ""
}
],
"year": 2005,
"venue": "Proceedings of the 14th international conference on World Wide Web",
"volume": "",
"issue": "",
"pages": "332--341",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Philipp Cimiano, G\u00fcnter Ladwig, and Steffen Staab. 2005b. Gimme'the Context: Context-driven Automatic Semantic Annotation with C-PANKOW. In Proceedings of the 14th international conference on World Wide Web, pages 332-341. ACM.",
"links": null
},
"BIBREF12": {
"ref_id": "b12",
"title": "Learning Concept Hierarchies from Text with a Guided Agglomerative clustering Algorithm",
"authors": [
{
"first": "Philipp",
"middle": [],
"last": "Cimiano",
"suffix": ""
},
{
"first": "Steffen",
"middle": [],
"last": "Staab",
"suffix": ""
}
],
"year": 2005,
"venue": "Proceedings of the Workshop on Learning and Extending Lexical Ontologies with Machine Learning Methods",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Philipp Cimiano and Steffen Staab. 2005. Learning Concept Hierarchies from Text with a Guided Agglomerative clustering Algorithm. In Proceedings of the Workshop on Learning and Extending Lexical Ontologies with Machine Learning Methods.",
"links": null
},
"BIBREF14": {
"ref_id": "b14",
"title": "Learning Ontology from Text: A Storytelling Exploratory Case Study",
"authors": [],
"year": null,
"venue": "International Conference on Knowledge Management in Organizations",
"volume": "",
"issue": "",
"pages": "477--491",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Learning Ontology from Text: A Storytelling Exploratory Case Study. In International Conference on Knowledge Management in Organizations, pages 477-491. Springer.",
"links": null
},
"BIBREF15": {
"ref_id": "b15",
"title": "Mining Concepts from Wikipedia for Ontology Construction",
"authors": [
{
"first": "Gaoying",
"middle": [],
"last": "Cui",
"suffix": ""
},
{
"first": "Qin",
"middle": [],
"last": "Lu",
"suffix": ""
},
{
"first": "Wenjie",
"middle": [],
"last": "Li",
"suffix": ""
},
{
"first": "Yirong",
"middle": [],
"last": "Chen",
"suffix": ""
}
],
"year": 2009,
"venue": "Proceedings of the IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology",
"volume": "03",
"issue": "",
"pages": "287--290",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Gaoying Cui, Qin Lu, Wenjie Li, and Yirong Chen. 2009. Mining Concepts from Wikipedia for Ontology Construction. In Proceedings of the IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology, volume 03, pages 287-290. IEEE.",
"links": null
},
"BIBREF16": {
"ref_id": "b16",
"title": "Towards an Automatic Fuzzy Ontology Generation",
"authors": [
{
"first": "Carmen",
"middle": [],
"last": "De Maio",
"suffix": ""
},
{
"first": "Giuseppe",
"middle": [],
"last": "Fenza",
"suffix": ""
},
{
"first": "Vincenzo",
"middle": [],
"last": "Loia",
"suffix": ""
},
{
"first": "Sabrina",
"middle": [],
"last": "Senatore",
"suffix": ""
}
],
"year": 2009,
"venue": "IEEE International Conference on Fuzzy Systems",
"volume": "",
"issue": "",
"pages": "1044--1049",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Carmen De Maio, Giuseppe Fenza, Vincenzo Loia, and Sabrina Senatore. 2009. Towards an Automatic Fuzzy Ontology Generation. In IEEE International Conference on Fuzzy Systems, pages 1044-1049. IEEE.",
"links": null
},
"BIBREF17": {
"ref_id": "b17",
"title": "An Experiment using Markov Logic Networks to Extract Ontology Concepts From Text. lLearning",
"authors": [
{
"first": "Lucas",
"middle": [],
"last": "Drumond",
"suffix": ""
},
{
"first": "Rosario",
"middle": [],
"last": "Girardi",
"suffix": ""
}
],
"year": 2010,
"venue": "",
"volume": "1",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Lucas Drumond and Rosario Girardi. 2010. An Experiment using Markov Logic Networks to Extract Ontology Concepts From Text. lLearning, 1:2.",
"links": null
},
"BIBREF18": {
"ref_id": "b18",
"title": "Security Ontology: Simulating Threats to Corporate assets",
"authors": [
{
"first": "Andreas",
"middle": [],
"last": "Ekelhart",
"suffix": ""
},
{
"first": "Stefan",
"middle": [],
"last": "Fenz",
"suffix": ""
},
{
"first": "Markus",
"middle": [
"D"
],
"last": "Klemen",
"suffix": ""
},
{
"first": "Edgar",
"middle": [
"R"
],
"last": "Weippl",
"suffix": ""
}
],
"year": 2006,
"venue": "International Conference on Information Systems Security",
"volume": "",
"issue": "",
"pages": "249--259",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Andreas Ekelhart, Stefan Fenz, Markus D Klemen, and Edgar R Weippl. 2006. Security Ontology: Simulating Threats to Corporate assets. In International Conference on Information Systems Security, pages 249-259. Springer.",
"links": null
},
"BIBREF19": {
"ref_id": "b19",
"title": "Ontology Enrichment with Texts from the WWW. Semantic Web Mining",
"authors": [
{
"first": "Andreas",
"middle": [],
"last": "Faatz",
"suffix": ""
},
{
"first": "Ralf",
"middle": [],
"last": "Steinmetz",
"suffix": ""
}
],
"year": 2002,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Andreas Faatz and Ralf Steinmetz. 2002. Ontology Enrichment with Texts from the WWW. Semantic Web Mining, 20.",
"links": null
},
"BIBREF20": {
"ref_id": "b20",
"title": "Formal Concept Analysis. Wissenschaftliche Zeitschrift-Technischen Universitat Dresden",
"authors": [
{
"first": "Bernhard",
"middle": [],
"last": "Ganter",
"suffix": ""
},
{
"first": "Rudolf",
"middle": [],
"last": "Wille",
"suffix": ""
}
],
"year": 1996,
"venue": "",
"volume": "45",
"issue": "",
"pages": "8--13",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Bernhard Ganter and Rudolf Wille. 1996. Formal Concept Analysis. Wissenschaftliche Zeitschrift- Technischen Universitat Dresden, 45:8-13.",
"links": null
},
"BIBREF21": {
"ref_id": "b21",
"title": "Contextual Attribute Logic",
"authors": [
{
"first": "Bernhard",
"middle": [],
"last": "Ganter",
"suffix": ""
},
{
"first": "Rudolf",
"middle": [],
"last": "Wille",
"suffix": ""
}
],
"year": 1999,
"venue": "International Conference on Conceptual Structures",
"volume": "",
"issue": "",
"pages": "377--388",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Bernhard Ganter and Rudolf Wille. 1999. Contextual Attribute Logic. In International Conference on Conceptual Structures, pages 377-388. Springer.",
"links": null
},
"BIBREF22": {
"ref_id": "b22",
"title": "A Semi-automatic Method to Ontology Design by Using FCA",
"authors": [
{
"first": "Hele-Mai",
"middle": [],
"last": "Haav",
"suffix": ""
}
],
"year": 2004,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Hele-Mai Haav. 2004. A Semi-automatic Method to Ontology Design by Using FCA. In CLA. Citeseer.",
"links": null
},
"BIBREF23": {
"ref_id": "b23",
"title": "Ontology Learning from Text using Relational Concept Analysis",
"authors": [
{
"first": "Amedeo",
"middle": [],
"last": "Mohamed Rouane Hacene",
"suffix": ""
},
{
"first": "Petko",
"middle": [],
"last": "Napoli",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Valtchev",
"suffix": ""
}
],
"year": 2008,
"venue": "International MCETECH Conference on e-Technologies",
"volume": "",
"issue": "",
"pages": "154--163",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Mohamed Rouane Hacene, Amedeo Napoli, Petko Valtchev, Yannick Toussaint, and Rokia Bendaoud. 2008. Ontology Learning from Text using Relational Concept Analysis. In International MCETECH Conference on e-Technologies, pages 154-163. IEEE.",
"links": null
},
"BIBREF24": {
"ref_id": "b24",
"title": "Automatic Acquisition of Hyponyms from Large Text Corpora",
"authors": [
{
"first": "A",
"middle": [],
"last": "Marti",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Hearst",
"suffix": ""
}
],
"year": 1992,
"venue": "Proceedings of the 14th conference on Computational linguistics",
"volume": "2",
"issue": "",
"pages": "539--545",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Marti A Hearst. 1992. Automatic Acquisition of Hyponyms from Large Text Corpora. In Proceedings of the 14th conference on Computational linguistics, volume 2, pages 539- 545. Association for Computational Linguistics.",
"links": null
},
"BIBREF25": {
"ref_id": "b25",
"title": "Procedures for Performing Systematic Reviews",
"authors": [
{
"first": "Barbara",
"middle": [],
"last": "Kitchenham",
"suffix": ""
}
],
"year": 2004,
"venue": "",
"volume": "33",
"issue": "",
"pages": "1--26",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Barbara Kitchenham. 2004. Procedures for Performing Systematic Reviews. Keele, UK, Keele University, 33(2004):1-26.",
"links": null
},
"BIBREF26": {
"ref_id": "b26",
"title": "An Empirical Evaluation of Doc2Vec with Practical Insights into Document Embedding Generation",
"authors": [
{
"first": "Han",
"middle": [],
"last": "Jey",
"suffix": ""
},
{
"first": "Timothy",
"middle": [],
"last": "Lau",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Baldwin",
"suffix": ""
}
],
"year": 2016,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {
"arXiv": [
"arXiv:1607.05368"
]
},
"num": null,
"urls": [],
"raw_text": "Jey Han Lau and Timothy Baldwin. 2016. An Empirical Evaluation of Doc2Vec with Practical Insights into Document Embedding Generation. arXiv preprint arXiv:1607.05368.",
"links": null
},
"BIBREF27": {
"ref_id": "b27",
"title": "Automated Ontology Construction for Unstructured Text Documents",
"authors": [
{
"first": "Chang-Shing",
"middle": [],
"last": "Lee",
"suffix": ""
},
{
"first": "Yuan-Fang",
"middle": [],
"last": "Kao",
"suffix": ""
},
{
"first": "Yau-Hwang",
"middle": [],
"last": "Kuo",
"suffix": ""
},
{
"first": "Mei-Hui",
"middle": [],
"last": "Wang",
"suffix": ""
}
],
"year": 2007,
"venue": "Data & Knowledge Engineering",
"volume": "60",
"issue": "3",
"pages": "547--566",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Chang-Shing Lee, Yuan-Fang Kao, Yau-Hwang Kuo, and Mei-Hui Wang. 2007. Automated Ontology Construction for Unstructured Text Documents. Data & Knowledge Engineering, 60(3):547-566.",
"links": null
},
"BIBREF28": {
"ref_id": "b28",
"title": "Clustering Ontology-based Metadata in the Semantic Web",
"authors": [
{
"first": "Alexander",
"middle": [],
"last": "Maedche",
"suffix": ""
},
{
"first": "Valentin",
"middle": [],
"last": "Zacharias",
"suffix": ""
}
],
"year": 2002,
"venue": "European Conference on Principles of Data Mining and Knowledge Discovery",
"volume": "",
"issue": "",
"pages": "348--360",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Alexander Maedche and Valentin Zacharias. 2002. Clustering Ontology-based Metadata in the Semantic Web. In European Conference on Principles of Data Mining and Knowledge Discovery, pages 348-360. Springer.",
"links": null
},
"BIBREF29": {
"ref_id": "b29",
"title": "A Situated Ontology for Practical NLP",
"authors": [
{
"first": "Kavi",
"middle": [],
"last": "Mahesh",
"suffix": ""
},
{
"first": "Sergei",
"middle": [],
"last": "Nirenburg",
"suffix": ""
}
],
"year": 1995,
"venue": "Proceedings of the IJCAI-95 Workshop on Basic Ontological Issues in Knowledge Sharing",
"volume": "19",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Kavi Mahesh, Sergei Nirenburg, et al. 1995. A Situated Ontology for Practical NLP. In Proceedings of the IJCAI-95 Workshop on Basic Ontological Issues in Knowledge Sharing, volume 19, page 21. Citeseer.",
"links": null
},
"BIBREF30": {
"ref_id": "b30",
"title": "Mining Meaning from Wikipedia",
"authors": [
{
"first": "Olena",
"middle": [],
"last": "Medelyan",
"suffix": ""
},
{
"first": "David",
"middle": [],
"last": "Milne",
"suffix": ""
},
{
"first": "Catherine",
"middle": [],
"last": "Legg",
"suffix": ""
},
{
"first": "Ian",
"middle": [
"H"
],
"last": "Witten",
"suffix": ""
}
],
"year": 2009,
"venue": "International Journal of Human-Computer Studies",
"volume": "67",
"issue": "9",
"pages": "716--754",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Olena Medelyan, David Milne, Catherine Legg, and Ian H Witten. 2009. Mining Meaning from Wikipedia. International Journal of Human- Computer Studies, 67(9):716-754.",
"links": null
},
"BIBREF31": {
"ref_id": "b31",
"title": "Harvesting Domain Specific Ontologies from Text",
"authors": [
{
"first": "Hamid",
"middle": [],
"last": "Mousavi",
"suffix": ""
},
{
"first": "Deirdre",
"middle": [],
"last": "Kerr",
"suffix": ""
},
{
"first": "Markus",
"middle": [],
"last": "Iseli",
"suffix": ""
},
{
"first": "Carlo",
"middle": [],
"last": "Zaniolo",
"suffix": ""
}
],
"year": 2014,
"venue": "IEEE International Conference on Semantic Computing",
"volume": "",
"issue": "",
"pages": "211--218",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Hamid Mousavi, Deirdre Kerr, Markus Iseli, and Carlo Zaniolo. 2014. Harvesting Domain Specific Ontologies from Text. In IEEE International Conference on Semantic Computing, pages 211- 218. IEEE.",
"links": null
},
"BIBREF32": {
"ref_id": "b32",
"title": "Ontology Evaluation through Text Classification",
"authors": [
{
"first": "Yael",
"middle": [],
"last": "Netzer",
"suffix": ""
},
{
"first": "David",
"middle": [],
"last": "Gabay",
"suffix": ""
},
{
"first": "Meni",
"middle": [],
"last": "Adler",
"suffix": ""
},
{
"first": "Yoav",
"middle": [],
"last": "Goldberg",
"suffix": ""
},
{
"first": "Michael",
"middle": [],
"last": "Elhadad",
"suffix": ""
}
],
"year": 2009,
"venue": "Advances in Web and Network Technologies, and Information Management",
"volume": "",
"issue": "",
"pages": "210--221",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Yael Netzer, David Gabay, Meni Adler, Yoav Goldberg, and Michael Elhadad. 2009. Ontology Evaluation through Text Classification. In Advances in Web and Network Technologies, and Information Management, pages 210-221. Springer.",
"links": null
},
"BIBREF33": {
"ref_id": "b33",
"title": "Ontology Evolution: Not the same as Schema Evolution",
"authors": [
{
"first": "F",
"middle": [],
"last": "Natalya",
"suffix": ""
},
{
"first": "Michel",
"middle": [],
"last": "Noy",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Klein",
"suffix": ""
}
],
"year": 2004,
"venue": "Knowledge and Information Systems",
"volume": "6",
"issue": "4",
"pages": "428--440",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Natalya F Noy and Michel Klein. 2004. Ontology Evolution: Not the same as Schema Evolution. Knowledge and Information Systems, 6(4):428-440.",
"links": null
},
"BIBREF34": {
"ref_id": "b34",
"title": "The Distributional Hypothesis",
"authors": [
{
"first": "Magnus",
"middle": [],
"last": "Sahlgren",
"suffix": ""
}
],
"year": 2008,
"venue": "Italian Journal of Disability Studies",
"volume": "20",
"issue": "",
"pages": "33--53",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Magnus Sahlgren. 2008. The Distributional Hypothesis. Italian Journal of Disability Studies, 20:33-53.",
"links": null
},
"BIBREF35": {
"ref_id": "b35",
"title": "A Methodology to Learn Ontological Attributes from the Web",
"authors": [
{
"first": "David",
"middle": [],
"last": "S\u00e1nchez",
"suffix": ""
}
],
"year": 2010,
"venue": "Data & Knowledge Engineering",
"volume": "69",
"issue": "6",
"pages": "573--597",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "David S\u00e1nchez. 2010. A Methodology to Learn Ontological Attributes from the Web. Data & Knowledge Engineering, 69(6):573-597.",
"links": null
},
"BIBREF36": {
"ref_id": "b36",
"title": "DYNAMO-MAS: a Multi-agent System for Ontology Evolution from Text",
"authors": [
{
"first": "Zied",
"middle": [],
"last": "Sellami",
"suffix": ""
},
{
"first": "Val\u00e9rie",
"middle": [],
"last": "Camps",
"suffix": ""
},
{
"first": "Nathalie",
"middle": [],
"last": "Aussenac-Gilles",
"suffix": ""
}
],
"year": 2013,
"venue": "Journal on Data Semantics",
"volume": "2",
"issue": "2-3",
"pages": "145--161",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Zied Sellami, Val\u00e9rie Camps, and Nathalie Aussenac- Gilles. 2013. DYNAMO-MAS: a Multi-agent System for Ontology Evolution from Text. Journal on Data Semantics, 2(2-3):145-161.",
"links": null
},
"BIBREF37": {
"ref_id": "b37",
"title": "Ontology Learning from Open Linked Data and Web Snippets",
"authors": [
{
"first": "Ilaria",
"middle": [],
"last": "Tiddi",
"suffix": ""
},
{
"first": "Nesrine",
"middle": [],
"last": "Ben Mustapha",
"suffix": ""
},
{
"first": "Yves",
"middle": [],
"last": "Vanrompay",
"suffix": ""
},
{
"first": "Marie-Aude",
"middle": [],
"last": "Aufaure",
"suffix": ""
}
],
"year": 2012,
"venue": "Confederated International Conferences\" On the Move to Meaningful Internet Systems",
"volume": "",
"issue": "",
"pages": "434--443",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Ilaria Tiddi, Nesrine Ben Mustapha, Yves Vanrompay, and Marie-Aude Aufaure. 2012. Ontology Learning from Open Linked Data and Web Snippets. In Confederated International Conferences\" On the Move to Meaningful Internet Systems\", pages 434- 443. Springer.",
"links": null
},
"BIBREF38": {
"ref_id": "b38",
"title": "Using word2vec to Build a Simple Ontology Learning System",
"authors": [
{
"first": "Gerhard",
"middle": [],
"last": "Wohlgenannt",
"suffix": ""
},
{
"first": "Filip",
"middle": [],
"last": "Minic",
"suffix": ""
}
],
"year": 2016,
"venue": "International Semantic Web Conference (Posters & Demos)",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Gerhard Wohlgenannt and Filip Minic. 2016. Using word2vec to Build a Simple Ontology Learning System. In International Semantic Web Conference (Posters & Demos).",
"links": null
}
},
"ref_entries": {
"FIGREF0": {
"text": "Trend Chart on Ontology Research",
"num": null,
"type_str": "figure",
"uris": null
},
"FIGREF1": {
"text": "Ontology Learning Categories",
"num": null,
"type_str": "figure",
"uris": null
},
"FIGREF2": {
"text": "Ontology Algorithms Comparison",
"num": null,
"type_str": "figure",
"uris": null
}
}
}
} |