File size: 98,045 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 |
{
"paper_id": "2020",
"header": {
"generated_with": "S2ORC 1.0.0",
"date_generated": "2023-01-19T07:12:26.124029Z"
},
"title": "Towards Creating Interoperable Resources for Conceptual Annotation of Multilingual Domain Corpora",
"authors": [
{
"first": "Svetlana",
"middle": [],
"last": "Sheremetyeva",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "South Ural State University",
"location": {
"addrLine": "76, Lenin pr. 454080",
"settlement": "Chelyabinsk",
"country": "Russia"
}
},
"email": "sheremetevaso@susu.ru"
}
],
"year": "",
"venue": null,
"identifiers": {},
"abstract": "In this paper we focus on creation of interoperable annotation resources that make up a significant proportion of an ongoing project on the development of conceptually annotated multilingual corpora for the domain of terrorist attacks in three languages (English, French and Russian) that can be used for comparative linguistic research, intelligent content and trend analysis, summarization, machine translation, etc. Conceptual annotation is understood as a type of task-oriented domain-specific semantic annotation. The annotation process in our project relies on ontological analysis. The paper details on the issues of the development of both static and dynamic resources such as a universal conceptual annotation scheme, multilingual domain ontology and multipurpose annotation platform with flexible settings, which can be used for the automation of the conceptual resource acquisition and of the annotation process, as well as for the documentation of the annotated corpora specificities. The resources constructed in the course of the research are also to be used for developing concept disambiguation metrics by means of qualitative and quantitative analysis of the golden portion of the conceptually annotated multilingual corpora and of the annotation platform linguistic knowledge.",
"pdf_parse": {
"paper_id": "2020",
"_pdf_hash": "",
"abstract": [
{
"text": "In this paper we focus on creation of interoperable annotation resources that make up a significant proportion of an ongoing project on the development of conceptually annotated multilingual corpora for the domain of terrorist attacks in three languages (English, French and Russian) that can be used for comparative linguistic research, intelligent content and trend analysis, summarization, machine translation, etc. Conceptual annotation is understood as a type of task-oriented domain-specific semantic annotation. The annotation process in our project relies on ontological analysis. The paper details on the issues of the development of both static and dynamic resources such as a universal conceptual annotation scheme, multilingual domain ontology and multipurpose annotation platform with flexible settings, which can be used for the automation of the conceptual resource acquisition and of the annotation process, as well as for the documentation of the annotated corpora specificities. The resources constructed in the course of the research are also to be used for developing concept disambiguation metrics by means of qualitative and quantitative analysis of the golden portion of the conceptually annotated multilingual corpora and of the annotation platform linguistic knowledge.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Abstract",
"sec_num": null
}
],
"body_text": [
{
"text": "The importance of linguistic annotations and, especially, semantic annotations over raw textual data is widely acknowledged as critical in developing language technologies, such as intelligent content and trend analysis, classification, machine learning, summarization, machine translation, etc. (Mair, 2005; Pustejovsky, 2012) . However, and this is also widely recognized, annotated corpora are quite sparse and their availability is often problematic due to no or restricted access, differences in volume and principles of construction, nonstandardized and/or unsuitable annotations for specific language technology tasks. There are good reasons for this, -annotating a comprehensive corpus with semantic representations is a hard, costly and time-consuming task. In spite of quite a number of attempts to facilitate the problem by developing reusable annotations, including semantic annotation formats, such as, for example, XML, SGML, etc., and the introduction of increasingly convivial and hardware-independent application software, it is difficult to find a system that matches exactly end-user requirements. For quality semantic annotation, the portable annotation software packages, as the main dynamic annotation resource should contain a significant amount of linguistic knowledge, acquisition of which so far is highly problematic. If, however, genericity is considered as applied to a family of applications, i.e., applications sharing tasks and domains, one can probably suggest particular approaches to solve the problem, even cross linguistically. In this paper we attempt just that.",
"cite_spans": [
{
"start": 296,
"end": 308,
"text": "(Mair, 2005;",
"ref_id": "BIBREF13"
},
{
"start": 309,
"end": 327,
"text": "Pustejovsky, 2012)",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1."
},
{
"text": "Our ultimate goal is to develop a methodology for developing annotation resources and resources themselves for the conceptual annotation of multilingual domain corpora, which are interoperable across languages and targeted to the automation of the annotation process primarily, but not exclusively, for such tasks as intelligent content analysis, machine learning, and classification. In our project, conceptual annotation is understood as a type of domain-specific task-oriented semantic annotation as opposed to the annotation with high level semantic properties, such as animacy, being human, person, etc.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1."
},
{
"text": "We demonstrate our approach on the domain of e-news on terrorist attacks in three languages, English, French and Russian. Our motivation to focus on the domain on terrorist attacks is that counterterrorist activity requires, among others, operative analysis of unstructured einformation and the availability of means to speed up the creation of annotated corpora in this particular domain is of high importance. We here focus on the development of both static and dynamic annotation resources such as a universal conceptual annotation scheme, multilingual domain ontology and annotation platform with flexible settings. The platform is multipurpose; it can be used for the automation of the conceptual resource acquisition and of the annotation process itself, as well as for the documentation of the annotated corpora specificities. The resources constructed in the course of the research are also to be used for developing concept disambiguation metrics by means of qualitative and quantitative analysis of the golden portion of the conceptually annotated multilingual corpora and of the annotation platform linguistic knowledge.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1."
},
{
"text": "The rest of the paper is organized as follows. Section 2 gives an overview of the related work. Section 3 defines the research tasks and introduces our data set. In Section 4 we suggest a methodology of building interoperable domain-specific conceptual annotation resources and describe the pool of static and dynamic resources built in the course of the current phase of the research. Section 5 describes the \"first-machine-then-human\" workflow of the conceptual annotation procedure. We conclude with the research overview and future work.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1."
},
{
"text": "Today the area of language annotation research witnesses the tendency towards semantization and, in particular, domain semantization (in our research, domain conceptualization), as the most realistic way to solve language technology tasks. The current trend is to use domain ontologies as conceptual annotation instruments, which, in turn, boosts the research in the field of ontology development. Ontologies are most often created for the annotation of unilingual (most often, English) domain corpora oriented to particular tasks. For example, to name just a few, the ontology described in (Roberts A. et. al., 2009) is created for the analysis of English medical records. (Tenenboim L et al., 2008) present the domain ontology for personalized filtration of English eNews. (Mannes and Golbeck, 2005; Najgebauer et al., 2008; Inyaem et al., 2009) devote their efforts to building ontologies for forecasting terror attacks and extraction of terrorist events from eNews. There is much less research on the ontology-based annotation in other languages, among which (as most closely related to our research) are (Dobrov et al., 2015) who suggest ways to semantically annotate a Russian domain corpus, and (Djemaa et al., 2016) focusing on a French corpus, correspondingly, As ontology development is a very tough and timeconsuming task, there are attempts to save effort in constructing ontological resources by making them multilingual. Multilingualism in ontologies is generally understood in two major senses: 1) as the adaptation (or understandability) of the ontology labels for the usersnative speakers of different national languages and 2) as the capability of one ontology to be applied to processing texts in different languages regardless of the language used for wording concept labels. These understandings of ontology multilingualism directly depend on the interpretation of ontology either as a language-dependent or language-independent resource.",
"cite_spans": [
{
"start": 591,
"end": 617,
"text": "(Roberts A. et. al., 2009)",
"ref_id": "BIBREF20"
},
{
"start": 674,
"end": 700,
"text": "(Tenenboim L et al., 2008)",
"ref_id": "BIBREF25"
},
{
"start": 775,
"end": 801,
"text": "(Mannes and Golbeck, 2005;",
"ref_id": "BIBREF12"
},
{
"start": 802,
"end": 826,
"text": "Najgebauer et al., 2008;",
"ref_id": "BIBREF17"
},
{
"start": 827,
"end": 847,
"text": "Inyaem et al., 2009)",
"ref_id": "BIBREF9"
},
{
"start": 1109,
"end": 1130,
"text": "(Dobrov et al., 2015)",
"ref_id": null
},
{
"start": 1202,
"end": 1223,
"text": "(Djemaa et al., 2016)",
"ref_id": "BIBREF3"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2."
},
{
"text": "Language-dependent ontologies are thesaurus-like structures whose elements are defined by the properties of a specific language. A well-known example of such resource, often called ontological, is the famous WordNet thesaurus (Miller et al., 1990) . The research on providing ontological multilingualism here goes in the direction of localization of the labels of ontology concepts, rather than modification of the ontological conceptualization. The localization procedure can go in different ways. For example, (Montiel-Ponsoda et al., 2008) propose the association of word senses in different languages to ontology concepts through a special linguistic model, while (Espinoza et al., 2008) suggest translating ontology labels into the user's language. One more localization technique is to manually annotate ontological concepts with labels in different languages (Chaves and Trojahn, 2010) . (Alatrish et al., 2014) direct their efforts to the development of universal tools that could be used for semi-automatic procedure of building separate ontologies tuned to different languages. (Embley et al., 2019) suggest methodologies on how to relate unilingual ontologies by mapping both the data and the meta-data of these ontologies. The use of language-dependent ontologies for interoperable semantic (conceptual) annotation of multilingual corpora does not seem quite doable.",
"cite_spans": [
{
"start": 226,
"end": 247,
"text": "(Miller et al., 1990)",
"ref_id": "BIBREF15"
},
{
"start": 512,
"end": 542,
"text": "(Montiel-Ponsoda et al., 2008)",
"ref_id": "BIBREF14"
},
{
"start": 668,
"end": 691,
"text": "(Espinoza et al., 2008)",
"ref_id": "BIBREF7"
},
{
"start": 866,
"end": 892,
"text": "(Chaves and Trojahn, 2010)",
"ref_id": "BIBREF2"
},
{
"start": 895,
"end": 918,
"text": "(Alatrish et al., 2014)",
"ref_id": "BIBREF0"
},
{
"start": 1088,
"end": 1109,
"text": "(Embley et al., 2019)",
"ref_id": "BIBREF6"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2."
},
{
"text": "Language-independent ontologies, like e.g., Mikrokosmos (Nirenburg and Raskin, 2004) , SUMO (Niles et al., 2003) and BFO (Arp et al., 2015) , allow multilingualism in the second sense (the applicability to processing texts in multiple languages) per definition, provided that each lexical unit (one-or multi-component) in the vocabulary of a particular language is mapped (according to special rules) into such ontology concept. This is the basic feature that makes language-independent ontologies applicable to semantic (conceptual, including) annotations that can be interoperable across languages. Given the expense of manual work, unavoidable in semantic (conceptual) annotation a lot of effort in using language-independent ontologies as annotation instruments is currently devoted to the creation of different tools to increase annotators' productivity. As a rule, so far, such annotation tools are user interfaces for mapping lexical units into ontological concepts and/or postediting the results of the automated annotation (Zagorul'ko et al., 2012; Stenetorp et al., 2012) .",
"cite_spans": [
{
"start": 56,
"end": 84,
"text": "(Nirenburg and Raskin, 2004)",
"ref_id": "BIBREF19"
},
{
"start": 92,
"end": 112,
"text": "(Niles et al., 2003)",
"ref_id": "BIBREF18"
},
{
"start": 121,
"end": 139,
"text": "(Arp et al., 2015)",
"ref_id": null
},
{
"start": 1032,
"end": 1057,
"text": "(Zagorul'ko et al., 2012;",
"ref_id": "BIBREF29"
},
{
"start": 1058,
"end": 1081,
"text": "Stenetorp et al., 2012)",
"ref_id": "BIBREF24"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "2."
},
{
"text": "Creation of interoperable resources for annotation makes should be closely associated with the annotation procedure that in our research is defined by the intersection of the following criteria: (i) data-driven methodology directed from analysis to representation, (ii) domain orientation, (iv) interoperability across languages, (v) automation of the annotation process, (vi) reusability of resources. We argue that interoperability of content annotation across languages calls for a clear division between language-dependent lexical knowledge and language-independent conceptual knowledge that can be best represented in ontology. We consider ontological analysis as a main instrument for interoperable conceptual annotation with a tagset defined by the ontological concepts. We are fully aware that ontological analysis has a serious limitation that lies in its practical realization. The shortcomings of ontological analysis are well-known and include the difficulty of clearly specifying the boundaries of the analysis and the influence of objective human judgments. There is no universal recipe for ideal ontological analysis therefore, as a rule, in every practical project, specific approaches are developed to deal with the problems above. Our solutions are domain-constraint and data-driven. Then, to reduce manual work, a decision was made to experiment as much as possible with the \"firstmachine, then human\" set-up of annotation work and to postpone the actual annotation process till later stages of the research and to first focus on the creation of the resources for annotation, which, following the classification given in (Witt et al., 2009) are divided into static and dynamic. In our research static resources include a conceptual annotation knowledge that consists of multilingual comparable domain corpora on terrorist acts in three languages (English, French and Russian), a universal conceptual annotation scheme, a multilingual domain ontology, domain-related unilingual lexicons and lexical-ontological mappings. The dynamic resources are tools to automate the creation of both static resources, and the annotation procedure.",
"cite_spans": [
{
"start": 1640,
"end": 1659,
"text": "(Witt et al., 2009)",
"ref_id": "BIBREF27"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Task Definition",
"sec_num": "3.1"
},
{
"text": "The road map for this research is as follows. First, the data set for the study was acquired and conceptualized resulting into lists of conceptually classified lexical items, and then the upper-level ontology and representation formalism were decided on followed by the development of a seed multilingual ontology for the terrorist attack domain. The seed ontology was further refined and populated with the text template technique. In parallel with the research on the content (knowledge) side of the project, a toolkit to automate the work on all its stages was being developed.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Task Definition",
"sec_num": "3.1"
},
{
"text": "First of all, the advantage was taken of the previously built domain resources created for our earlier CAT project that include a 400 000 word Russian terrorist domain corpus of 2016-2017 e-news acquired in the Internet and a Russian-English lexicon of multicomponent lexical units built over the corpus. The lexicon includes initial corpusbased Russian vocabulary translated into English by professional translators. These data were used, first of all, to acquire knowledge for the built-in house Internet crawlers to automatically collect new portions of Russian, as well as English and French domain corpora. The crawler knowledge was decided to consist of key phrases rather than single words as the use of key phrases has the immediate effect in improving precision in keyword related tasks (Lefever et al., 2009) . Then, a general opinion that content resides in noun phrases (Witschel, 2005) , made us vote in favor of keywords/phrases as grammatically well-formed noun phrases. The key noun phrases were automatically extracted from the \"old\" 400 000 word Russian corpus by means of the tool described in (Sheremetyeva, 2012) that we trained for the Russian terrorist domain. The top 30% of the extracted Russian key noun phrases and their translations into English and French were used as the knowledge for the crawlers, by means of which the second part of the raw data, -multilingual terrorist act corpora of 100,000 words published on the Internet in 2018-2019 in the three languages were automatically acquired. For feasibility reasons, we excluded news on terrorist military activities and focused on the news on terrorist attack committed by individuals or terrorist groups in different countries.",
"cite_spans": [
{
"start": 796,
"end": 818,
"text": "(Lefever et al., 2009)",
"ref_id": null
},
{
"start": 882,
"end": 898,
"text": "(Witschel, 2005)",
"ref_id": "BIBREF26"
},
{
"start": 1113,
"end": 1133,
"text": "(Sheremetyeva, 2012)",
"ref_id": "BIBREF21"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Data Set",
"sec_num": "3.2"
},
{
"text": "In this section we describe the process of acquisition of static resources for conceptual annotation at the preannotation stage. The results of the acquisition were used as the knowledge base for the NLP annotation platform (see section 4.3) and were further augmented in the course of the whole research period.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Static resources",
"sec_num": "4.1"
},
{
"text": "Data set analysis The first step in building resources for interoperable conceptual annotation consisted in classifying the multilingual corpora lexis into domain-relevant conceptual classes (or categories). It included decisions on i) the units of conceptual classification, which we took to be both single words and multi-word phrases of different POS classes, and ii) the list of categories/concepts. The set up for this work included an initial intuitively prescribed universal list of conceptual classes with definitions, unilingual (English, French and Russian) corpora-based frequency lists of multicomponent noun phrases as most closely content-related textual units (note, not only key phrases), raw corpora for context check, if needed, and conceptualization guidelines that were the same throughout the languages. The lists of noun phrases for conceptualization were constructed in two takes. First, the set of noun phrases up to four components long 1 were 1 Constraint to four component extraction units is explained by the limitations of the extractor. automatically extracted from the English, French, and Russian corpora with the lexical extractor (Sheremetyeva, cf.) after it was trained for the terrorist domain in all the three languages. Then, every unilingual corpus was searched for longer noun phrases with the regular \"find\" functionality using the seed set of automatically extracted 4-component phrases. The domain-relevant units where then manually classified into conceptual classes (starting with the prescribed set) and following the guidelines. Special attention was paid to the selection of concept labels that were worded in English and made as descriptive as possible. Throughout the whole research period, weekly discussions were held by the project participants to provide for inter-conceptualization consistency and brush up. This stage resulted in the specification of the seed set of domain concepts. Other types of phrases were then extracted and classified in the same way followed by further brush up and extension of the cross-language conceptual class set. In general, the concept set was elaborated to specify a 3-level tree-like structure of concept organization with 97 fine-grained conceptual categories, assigned to 20 top-level domain categories. Table 1 shows a fragment of the top level domain concept list with definitions; Table 2 lists the second level grained concepts for the top domain concepts COUNTER-TERRORISM and CONSEQUENCES, and Table 3 presents Table 3 : Fragments of the most frequent unilingual lexical units put into the \"AGENT-TERRORIST\" class.",
"cite_spans": [],
"ref_spans": [
{
"start": 2297,
"end": 2304,
"text": "Table 1",
"ref_id": "TABREF0"
},
{
"start": 2377,
"end": 2384,
"text": "Table 2",
"ref_id": "TABREF1"
},
{
"start": 2493,
"end": 2509,
"text": "Table 3 presents",
"ref_id": null
},
{
"start": 2510,
"end": 2517,
"text": "Table 3",
"ref_id": null
}
],
"eq_spans": [],
"section": "4.1.1",
"sec_num": null
},
{
"text": "Like any work on semantics based on human judgment, the concept specification process, in spite of all the domain constraints and guidelines, was not free from different levels of detalization, overlaps in interpretation and even contradictions. In such cases, reasonably strict decisions were taken by the project leader.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "4.1.1",
"sec_num": null
},
{
"text": "In our project, we follow three basic methodological assumptions on ontology definition. The first is that ontology is a reusable language-independent resource; the second is that \"domain-specific knowledge is not isolated from general world knowledge\" (Moreno & P\u00e9rez, 2011, p . 233) and we, therefore, link our ontological resource to the upper-level Mikrokosmos ontology (Nirenburg & Raskin, cf.) to reuse the knowledge that is already there. We also follow the initial Mikrokosmos division of the reality into OBJECTS, EVENTS, and PROPERTIES, and use its formalism. We keep concept labels worded in English, the scopes of which, like in Mikrokosmos, are only specified by their definitions. Our third assumption is that interoperable domain ontological knowledge can be extracted from multilingual comparable domain corpora using mixed (top-down/bottom-up) acquisition techniques (Francesconi et al., 2010) .",
"cite_spans": [
{
"start": 253,
"end": 277,
"text": "(Moreno & P\u00e9rez, 2011, p",
"ref_id": null
},
{
"start": 374,
"end": 399,
"text": "(Nirenburg & Raskin, cf.)",
"ref_id": null
},
{
"start": 884,
"end": 910,
"text": "(Francesconi et al., 2010)",
"ref_id": "BIBREF8"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Ontology",
"sec_num": "4.1.2"
},
{
"text": "The set of domain concepts defined at the lexical analysis stage formed the seed e-news terrorist ontology, whose pool of concepts was further augmented and refined by using the text-template technique. For example, such RELATION concepts as IS-A and INSTANCE-OF can be acquired (though not exclusively) using the following English/French/Russian parallel text templates: \"A / is / are / and other/ such as/ B\" (English) \"A est /somme/ comme / et autres / B\" (French) \"A / \u044d\u0442\u043e / \u0438 \u0434\u0440\u0443\u0433\u0438\u0435 / \u0442\u0430\u043a\u0438\u0435 \u043a\u0430\u043a / B\" (Russian), where B is a lexeme that can signal of a more general concept; A is a lexeme of a more specific class.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Ontology",
"sec_num": "4.1.2"
},
{
"text": "The top domain concept MEANS OF ATTACK can be further split by means of such corsslingual templates as \"attack /with/ using/involving/ C\" (English) \"attaque /avec/au moyen de/ C\" (French) \"\u0430\u0442\u0430\u043a\u0430 /\u0441 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u043e\u0432\u0430\u043d\u0438\u0435\u043c/ \u0441 \u043f\u0440\u0438\u043c\u0435\u043d\u0435\u043d\u0438\u0435\u043c/ C\" (Russian), where C stands for lexemes of a weapon type concept.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Ontology",
"sec_num": "4.1.2"
},
{
"text": "The resulted ontology currently consists of 112 OBJECT and EVENT concepts and 27 PROPERTY concepts, see details in (Sheremetyeva & Zinovyeva, 2018) .",
"cite_spans": [
{
"start": 115,
"end": 147,
"text": "(Sheremetyeva & Zinovyeva, 2018)",
"ref_id": "BIBREF22"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Ontology",
"sec_num": "4.1.2"
},
{
"text": "Lexical-Ontological Mapping Our main methodology for the interoperable conceptual annotation is ontological analysis. In practice, ontological analysis consists in mapping corpora lexical units into ontological concepts that, in our case, calls for creating unilingual lexicons, in which every domain-related unit is explicitly linked to an ontological concept. The boundaries of such mappings were specified by the domain data analysis and where allowed to be one-tomany, many-to-one or many-to-many. This had to follow human judgement, though strictly regulated by the mapping guidelines. For example, the French named entity \"Charlie Hebdo\" is mapped into the concepts OBJECT OF ATTACK (its office was targeted by terrorists in 2015) and SOURCE (it is a weekly newspaper that published info on terrorist attacks). Among lexical items mapped to several concepts there are, for example, the English word \"police officer\" and its French and Russian equivalents \"policier\" and \"\u043f\u043e\u043b\u0438\u0446\u0435\u0439\u0441\u043a\u0438\u0439\", correspondingly. Namely, following their use in the corresponding unilingual corpus, these lexical items are mapped into the 4 concepts of the multilingual ontology (the order of examples below are English, French and Russian): COUNTER-TERRORISM: After the explosions, the authorities deployed police officers. / Apr\u00e8s les explosions, les autorit\u00e9s ont d\u00e9ploy\u00e9 des policiers. / \u041f\u043e\u0441\u043b\u0435 \u0432\u0437\u0440\u044b\u0432\u043e\u0432 \u0432\u043b\u0430\u0441\u0442\u0438 \u0432\u044b\u0441\u0442\u0430\u0432\u0438\u043b\u0438 \u043f\u043e\u043b\u0438\u0446\u0435\u0439\u0441\u043a\u0438\u0445.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "4.1.3",
"sec_num": null
},
{
"text": "CONSEQUENCES: A police officer was killed. / Un policier est tu\u00e9. / \u041f\u043e\u043b\u0438\u0446\u0435\u0439\u0441\u043a\u0438\u0439 \u0431\u044b\u043b \u0443\u0431\u0438\u0442. SOURCE: According to a police officer, the man shouted \"Allahu akbar\". / Selon des policiers, l'homme aurait cri\u00e9 \u00ab Allah akbar \u00bb. / \u041f\u043e \u0441\u043b\u043e\u0432\u0430\u043c \u043f\u043e\u043b\u0438\u0446\u0435\u0439\u0441\u043a\u0438\u0445 \u043c\u0443\u0436\u0447\u0438\u043d\u0430 \u043a\u0440\u0438\u0447\u0430\u043b \u00ab\u0410\u043b\u043b\u0430\u0445 \u0430\u043a\u0431\u0430\u0440\u00bb",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "4.1.3",
"sec_num": null
},
{
"text": "Russia's ambassador is assassinated in Ankara by a police officer. / L'ambassadeur de Russie est assassin\u00e9 \u00e0 Ankara par un policier. / \u0420\u043e\u0441\u0441\u0438\u0439\u0441\u043a\u0438\u0439 \u043f\u043e\u0441\u043e\u043b \u0443\u0431\u0438\u0442 \u0432 \u0410\u043d\u043a\u0430\u0440\u0435 \u043f\u043e\u043b\u0438\u0446\u0435\u0439\u0441\u043a\u0438\u043c.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "AGENT-TERRORIST:",
"sec_num": null
},
{
"text": "We also introduced a convention that is not very obvious and generally accepted. It concerns the ontology mapping of multicomponent lexical units, in which individual components bear domain-related conceptual meanings that translate different aspects of content and do not contradict one another. For example, in the English phrase \"airport shooting suspect\", the word \"shooting\" conveys the information on the type of attack, the word \"airport\" points to the location where the attack took place, while the word \"suspect\" has two conceptual meanings \"assumption\" and \"performer of the terrorist attack\". All these content components are sincretically united in the phrase. Therefore, the convention is to map this multicomponent lexeme into 4 concepts, -AGENT-TERRORIST, ASSUMPTION, TYPE OF ATTACK and LOCATION. Similarly, the phrase \"Algerian terrorist\" is mapped into the AGENT-TERRORIST and NATION concepts. Multiple ontological-lexical mappings will obviously lead to assigning multiple concept tags to textual units in the annotation procedure. However, as seen from the examples above, in our approach to annotation, it might or might not signal of lexical unit conceptual ambiguity. The situation forecast the need to make decisions on when multiple conceptual tags have to be disambiguated and when it should not be done to preserve as much domain-related content as possible. This issue is a matter of further investigation.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "AGENT-TERRORIST:",
"sec_num": null
},
{
"text": "The domain conceptualization described in this section resulted in the acquisition of the pre-annotation static knowledge including the multilingual terrorist act domain ontology and ontology-mapped corpora-based unilingual lexicons of English, French and Russian. This knowledge was used to create the first version of the multilingual annotation platform described in Section 4.2.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "AGENT-TERRORIST:",
"sec_num": null
},
{
"text": "A tool, which we call annotation platform, is the main part of all dynamic resources we used in our work. We approached its design with several considerations in mind. First of all, the annotation platform should automate the process of conceptual annotation and mark-up every unilingual corpus with the universal set of concept tags defined by the multilingual domain ontology. It is also desirable for the platform to contain knowledge that could help conceptual disambiguation. Further, it should be possible to configure the platform settings to different languages and language-dependent types of linguistic information. The annotation platform should allow for the knowledge administration and, therefore, be provided with the acquisition interface.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Annotation Platform",
"sec_num": "4.2"
},
{
"text": "To save the development effort we reused, though sufficiently updated two software modules from our earlier (different type) project (Sheremetyeva, 2013) that meet most of the expectations on the annotation platform. The first module is the program shell of the multilingual TransDict e-lexicon and the second is the tagger to which TransDict is pipelined. TransDict is built over a powerful set of linguistic features that have a tree-like structure. It is realized as a number of cross-referenced monolingual lexicons. Every monolingual lexicon consists of a set of entries with semantic, syntactic and morphological zones of flexible settings. The TransDict entry is meant for one meaning (semantic class) of a lexeme in a given language. The morphological zone can contain the morphological information, such as part-of-speech, number, gender, etc., and word paradigms of a lexical unit up to 10 components long explicitly listed in the entry. The latter makes recognition of text wordforms straightforward. Depending on the configuration of linguistic information, every wordform in the lexicon entry is automatically assigned a supertag that codes semantic and morphological information, such as part-of speech and typed morphological features that are language-dependent. TransDict, what is important for our project, has an advanced knowledge administration user interface, builtin search module with flexible search masks and a lot of other effort-saving functionalities, like automatic generation of entry structures and entry-fillers. The TransDict shell allows increasing the number of languages as necessary and can be configured to any type of knowledge. The adaptation of TransDict for the conceptual annotation task (see Figure 1 ) was as follows. We configured the program to three languages, -English, French and Russian. Semantic classes were set to the ontology concepts and some other classes like \"Other\" \"Numerals\", \"Definiteness\", etc., for mapping the lexemes of not specifically domain-related meaning. For feasibility reasons, so far, only upper-level ontology concepts were coded in TransDict. The morphological zones of the entries within each conceptual class were filled up with the explicitly listed morphological paradigms of the lexemes mapped to the ontology at the pre-annotation static resource acquisition stage (see Section 4.1.). If a lexical unit was mapped to several conceptual classes, several entries for this unit were created, each linked to a particular concept. Figure 1 shows a fragment of the TransDict main acquisition interface with the word list filtered by the mask \"English\" & \"mapped into the TERRORIST-AGENT (tag A) concept\" & \"also to any other concept\". The duplication of the lexical units shown in the left column of the interface displays multiple mappings. For example, the two-component lexeme \"alleged terrorist\" is listed twice as it is mapped into the TERRORIST-AGENT concept (tag A) and into the ASSUMPTION concept (tag I). The screenshot displays the \"alleged terrorist\" lexeme entry mapped into the ASSUMPTION concept. The morphological zone is filled with the lexeme wordforms that are automatically assigned supertags Ni and Nip, where N stands for \"noun\", \"I\" for the concept ASSUMPTION and \"p\" for plural. Supertags are positional, a concept code is the second in order; this coding format is inherited from the parent TransDict application. To allow the acquirers working independently at their own pace, TransDict is programmed in two variants, as MASTER with a full set of functionalities and as the so-called SLAVE -an empty program shell configured exactly as the corresponding version of MASTER but of a limited capability, namely, the user cannot change the dictionary settings (sets of languages, conceptual classes, entry structures and tags). SlAVEs filled by the acquirers with new portions of lexical conceptual knowledge are merged into MASTER on a regular basis. TransDict entries can be created for a single lexeme or for whole lists in batch mode. Figure 2 shows the window for ontological mapping when a lexeme is to be added to the TransDict knowledge. The window popsup following a click on the \"Add\" button in the interface. The selection of a conceptual class calls for another popup window for part-of-speech specification, after which an entry with typed morphological and syntactic zones appears that could be filled, if and as necessary. Fillers of the TransDict morphological zone fields supply knowledge to the tagger for conceptual annotation.",
"cite_spans": [
{
"start": 133,
"end": 153,
"text": "(Sheremetyeva, 2013)",
"ref_id": "BIBREF23"
}
],
"ref_spans": [
{
"start": 1737,
"end": 1745,
"text": "Figure 1",
"ref_id": "FIGREF1"
},
{
"start": 2511,
"end": 2519,
"text": "Figure 1",
"ref_id": "FIGREF1"
},
{
"start": 4039,
"end": 4047,
"text": "Figure 2",
"ref_id": "FIGREF2"
}
],
"eq_spans": [],
"section": "Annotation Platform",
"sec_num": "4.2"
},
{
"text": "As said above, original TransDict shell was substantially updated for the annotation knowledge management and now includes quite a number of new effort-saving acquisition and analysis functionalities, substantially augmented search/filtering possibilities, export/ import functions, etc. The new TransDict search module with a lot of possible search masks is shown in Figure 3 . The main update here is filtering according to the concept class parameters (combined or not with other mask parameters). One can filter lexemes of one conceptual class, lexemes of one class that are also mapped to any other concepts, and lexemes assigned to a fixed set of conceptual classes. This function shows knowledge lacuna to be filled. Filtration on the concept parameters can be done in two modes: based on lexeme main forms only or based on the whole paradigm of lexeme wordforms listed in the TransDict morphological zone. This obviously gives different results, comparing which one can find morphological hints for concept disambiguation in an automatically annotated text. In general, all types of filtering including the concept class masks give a lot of information on the domain annotation statistics that can be used e.g., for forecasting the conceptual ambiguity rate in a particular language and for developing automatic disambiguation metrics.",
"cite_spans": [],
"ref_spans": [
{
"start": 368,
"end": 376,
"text": "Figure 3",
"ref_id": "FIGREF3"
}
],
"eq_spans": [],
"section": "Annotation Platform",
"sec_num": "4.2"
},
{
"text": "The second module of the annotation platform is the tagger pipelined to TransDict. The tagger has a control interface and compilers which, if necessary, can be used for the acquisition of disambiguation rules and syntactic analysis rules. The tagger can be set to coarse-grain or fine-grain corpus mark-up. The coarse-grain mark-up outputs annotation with concept tags only, which can be enough for certain text-mining and content/knowledge extraction tasks. The fine-grain mark-up assigns a full range of linguistic features coded in the TransDict supertags that can be useful for disambiguation purposes. A screenshot of the control interface of the annotation platform tagger with the results of coarse-grain automatic conceptual tagging is shown in Figure 4 (see the concepts tags in Figure 2 ). Some lexemes shown in the tagger interface screenshot have multiple tags that signals of possible conceptual ambiguity. This version of the tagger does yet support concept disambiguation and, in general, the problem of automated conceptual disambiguation is out of the scope of this paper. We can only say at this stage that both statistical and, if necessary, linguistic information will be used for this purpose. This, among others, motivated the main change in the current tagging module as compared to the parent application, -two level fine-grained and coarse-grained annotation. The annotation platform is currently implemented as a PC application and includes three pipelined modules, TransDict MASTER, TransDict SLAVE and Tagger that can also be used as stand-alone tools.",
"cite_spans": [],
"ref_spans": [
{
"start": 753,
"end": 761,
"text": "Figure 4",
"ref_id": "FIGREF4"
},
{
"start": 788,
"end": 796,
"text": "Figure 2",
"ref_id": "FIGREF2"
}
],
"eq_spans": [],
"section": "Annotation Platform",
"sec_num": "4.2"
},
{
"text": "In our approach, the process of conceptual annotation as the implementation of ontological analysis is the process of mapping text strings (in our case grammatical phrases of different types) into the domain multilingual ontology. The annotation procedure is identical for each unilingual corpus. It is \"first-machine, then human\" and is incremental in nature. We tested the approach, given the volume of multilingual effort and expectations about reasonable annotator tasks, on relatively small portions of unilingual corpora of 20,000 words each. However, the process and the results of such annotation, which in the long run was potedited into golden, gave us a lot of experience and leads on how to treat conceptual annotation problems.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Annotation Procedure",
"sec_num": "5."
},
{
"text": "During the beginning annotation phases covered in this paper, the types of conceptual categories included in the annotation were constrained to 21 top-level domain concepts and the concept \"OTHER\", to which domainneutral lexemes are mapped. The annotators, who had already been trained in conceptualizing during the lexical analysis stage, were given a code-book with the sets of concepts associated with definitions and tags. The annotation process itself was done in several takes in an iterative manner. First, a weakly portion of the raw text meant to be gold-annotated was automatically tagged by our annotation platform described in Section 4.2 and then passed for postediting to the annotators. Conceptual ambiguity, if any, was resolved manually. In case a domain-relevant lexeme was left untagged or tagged incorrectly, it was supplied with correct linguistic information into the acquirer's personal TransDict SLAVE program to be further merged in TransDict MASTER (see Section 4.2) and the platform knowledge was thus updated, after which the annotation platform was used to automatically annotate the next portion of the corpus leading to a new knowledge update, etc. The knowledge was updated on a regular basis and the accuracy of the automatic annotation increased with very iteration. The accuracy was so far evaluated based on the annotators' reports on the amount of time spent on postediting and on the number of new lexical items to be merged into TransDict after every annotation iteration. Evidently, one cannot hope for a 100% correct automatic annotation without some risk of reducing annotation quality and, hence, human judgements cannot be avoided. However, our experiment shows that automation as used in the current research significantly augments and supports the annotation process.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Annotation Procedure",
"sec_num": "5."
},
{
"text": "The annotation procedure resulted in three golden conceptually annotated comparable English, French and Russian corpora of the e-news on terrorist acts and a substantial augmentation of the annotation platform knowledge. The TransDict lexicon currently consists of three unilingual lexicons of the English, French and Russian languages, that amount to around 43000 crossreferenced lexical entries acquired both at the preannotation stage, and in the course of annotation.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Annotation Procedure",
"sec_num": "5."
},
{
"text": "In this paper, we suggested a methodology of creating static and dynamic resources for interoperable conceptual annotation of domain corpora and presented actual annotation resources built along the suggested methodology for the multilingual (English, French and Russian) domain corpora of e-news on terrorist attacks. The resources include a universal conceptual annotation scheme, multilingual domain ontology, annotation platform with flexible settings and comparable golden conceptually annotated corpora in the three languages. This research is one of the major parts of an annotation project, which is significantly different from those that concentrate on morphological, syntactic or general types of semantic annotation. The emphasis of the presented work is on: i) a domain-specific level of annotation; ii) the assignment of well-defined interoperable conceptual representations based on multilingual domain ontology; and iii)\"first-machine-then-human\" approach to the annotation process.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusion",
"sec_num": "6."
},
{
"text": "Qualitative and quantitative investigation of the annotation resources we have constructed open quite a number of research opportunities for, e.g., theoretical aspects of social and comparative linguistics, as well as for research and development in Natural Language Processing technologies including multilingual Information Extraction, Generation, Question Answering, etc., and Machine Translation. The conceptual annotation knowledge can directly be used for developing machine learning techniques. In particular, the resource analysis findings can be used for developing concept disambiguation metrics, which, on top of increasing the volume of the annotation resources and annotated corpora, we see as our future work.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusion",
"sec_num": "6."
}
],
"back_matter": [],
"bib_entries": {
"BIBREF0": {
"ref_id": "b0",
"title": "Building Ontologies for Different Natural Languages",
"authors": [
{
"first": "E",
"middle": [
"A"
],
"last": "Alatrish",
"suffix": ""
},
{
"first": "D",
"middle": [],
"last": "To\u0161i\u0107",
"suffix": ""
},
{
"first": "N",
"middle": [],
"last": "Milenkov",
"suffix": ""
}
],
"year": 2014,
"venue": "Building Computer Science and Information Systems",
"volume": "11",
"issue": "2",
"pages": "623--644",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Alatrish E.A., To\u0161i\u0107 D., Milenkov N. (2014). Building Ontologies for Different Natural Languages. Building Computer Science and Information Systems. -Vol. 11(2). pp. 623-644.",
"links": null
},
"BIBREF1": {
"ref_id": "b1",
"title": "Building Ontologies with Basic Formal Ontology",
"authors": [
{
"first": "R",
"middle": [],
"last": "Arp",
"suffix": ""
},
{
"first": "B",
"middle": [],
"last": "Smith",
"suffix": ""
},
{
"first": "A",
"middle": [
"D"
],
"last": "Spear",
"suffix": ""
}
],
"year": 2010,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Arp, R., Smith, B., Spear, A.D. (2010). Building Ontologies with Basic Formal Ontology. MIT Press, Cambridge.",
"links": null
},
"BIBREF2": {
"ref_id": "b2",
"title": "Towards a Multilingual Ontology for Ontology-driven Content Mining in Social Web Sites",
"authors": [
{
"first": "",
"middle": [],
"last": "Chaves",
"suffix": ""
},
{
"first": "C",
"middle": [],
"last": "Trojahn",
"suffix": ""
}
],
"year": 2010,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Chaves, M and Trojahn C. (2010). Towards a Multilingual Ontology for Ontology-driven Content Mining in Social Web Sites -URL: https://goo.gl/sZKmS2(09.11.2019).",
"links": null
},
"BIBREF3": {
"ref_id": "b3",
"title": "Corpus annotation within the French FrameNet: a domain-by-domain methodology",
"authors": [
{
"first": "M",
"middle": [],
"last": "Djemaa",
"suffix": ""
},
{
"first": "M",
"middle": [],
"last": "Candito",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Muller Ph",
"suffix": ""
},
{
"first": "L",
"middle": [],
"last": "Vieu",
"suffix": ""
}
],
"year": 2016,
"venue": "Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC 2016)",
"volume": "",
"issue": "",
"pages": "3794--3801",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Djemaa M., Candito M., Muller Ph., Vieu L. (2016). Corpus annotation within the French FrameNet: a domain-by-domain methodology. Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC 2016), May 2016, Portoro\u017e, Slovenia pp. 3794-3801.",
"links": null
},
"BIBREF5": {
"ref_id": "b5",
"title": "Semanticheskij analiz novostnyh soobshchenij po teme \u00abElektronnye uslugi\u00bb: opyt primeneniya metodov ontologicheskoj semantiki Trudy XVIII ob\"edinennoj konferencii \u00abInternet i sovremennoe obshchestvo\u00bb (IMS-2015)",
"authors": [],
"year": null,
"venue": "",
"volume": "",
"issue": "",
"pages": "120--125",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Semanticheskij analiz novostnyh soobshchenij po teme \u00abElektronnye uslugi\u00bb: opyt primeneniya metodov ontologicheskoj semantiki Trudy XVIII ob\"edinennoj konferencii \u00abInternet i sovremennoe obshchestvo\u00bb (IMS-2015). pp. 120-125. (in Russian).",
"links": null
},
"BIBREF6": {
"ref_id": "b6",
"title": "Multilingual Ontologies for Cross-Language Information Extraction and Semantic Search",
"authors": [
{
"first": "D",
"middle": [
"W"
],
"last": "Embley",
"suffix": ""
},
{
"first": "S",
"middle": [
"W"
],
"last": "Liddle",
"suffix": ""
},
{
"first": "D",
"middle": [
"W"
],
"last": "Lonsdale",
"suffix": ""
},
{
"first": "Y",
"middle": [],
"last": "Tijerino",
"suffix": ""
}
],
"year": 2019,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Embley D. W., Liddle S. W., Lonsdale D. W., Tijerino Y. (2019). Multilingual Ontologies for Cross-Language Information Extraction and Semantic Search. -URL: https://pdfs.semanticscholar.org/6884/41a96b6da61295 c7df39b70db2f28531370a.pdf ((09.11.2019)",
"links": null
},
"BIBREF7": {
"ref_id": "b7",
"title": "Enriching an Ontology with Multilingual Information. The Semantic Web: Research and Applications. ESWC Lecture Notes in Computer Science",
"authors": [
{
"first": "M",
"middle": [],
"last": "Espinoza",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "G\u00f3mez-P\u00e9rez",
"suffix": ""
},
{
"first": "E",
"middle": [],
"last": "Mena",
"suffix": ""
}
],
"year": 2008,
"venue": "",
"volume": "5021",
"issue": "",
"pages": "333--347",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Espinoza, M., G\u00f3mez-P\u00e9rez A., Mena E. (2008). Enriching an Ontology with Multilingual Information. The Semantic Web: Research and Applications. ESWC Lecture Notes in Computer Science. -Springer, Berlin, Heidelberg. -Vol. 5021. pp. 333-347.",
"links": null
},
"BIBREF8": {
"ref_id": "b8",
"title": "Integrating a Bottom-Up and Top-Down Methodology for Building Semantic Resources for the Multilingual Legal Domain Semantic Processing of Legal Texts",
"authors": [
{
"first": "E",
"middle": [],
"last": "Francesconi",
"suffix": ""
},
{
"first": "S",
"middle": [],
"last": "Montemagni",
"suffix": ""
},
{
"first": "W",
"middle": [],
"last": "Peters",
"suffix": ""
},
{
"first": "D",
"middle": [],
"last": "Tiscornia",
"suffix": ""
}
],
"year": 2010,
"venue": "LNAE",
"volume": "6036",
"issue": "",
"pages": "95--121",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Francesconi E., Montemagni S., Peters W., Tiscornia D. (2010). Integrating a Bottom-Up and Top-Down Methodology for Building Semantic Resources for the Multilingual Legal Domain Semantic Processing of Legal Texts . LNAE. -Vol. 6036, pp. 95-121.",
"links": null
},
"BIBREF9": {
"ref_id": "b9",
"title": "Ontology-Based Terrorism Event Extraction Proceedings of the 1st International Conference on Information Science and Engineering",
"authors": [
{
"first": "U",
"middle": [],
"last": "Inyaem",
"suffix": ""
},
{
"first": "Haruechaiyasak",
"middle": [],
"last": "Ch",
"suffix": ""
},
{
"first": "Tran",
"middle": [
"D"
],
"last": "Meesad Ph",
"suffix": ""
}
],
"year": 2009,
"venue": "",
"volume": "",
"issue": "",
"pages": "912--915",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Inyaem U, Haruechaiyasak Ch., Meesad Ph,,Tran D. (2009). Ontology-Based Terrorism Event Extraction Proceedings of the 1st International Conference on Information Science and Engineering.. -P. 912-915.",
"links": null
},
"BIBREF11": {
"ref_id": "b11",
"title": "Languageindependent bilingual terminology extraction from a multilingual parallel corpus",
"authors": [
{
"first": "V",
"middle": [],
"last": "Hoste",
"suffix": ""
}
],
"year": 2009,
"venue": "Proceedings of the 12th Conference of the European Chapter of the ACL",
"volume": "",
"issue": "",
"pages": "496--504",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Hoste V. (2009). Language- independent bilingual terminology extraction from a multilingual parallel corpus'. In Proceedings of the 12th Conference of the European Chapter of the ACL, Athens, Greece. pp. 496-504.",
"links": null
},
"BIBREF12": {
"ref_id": "b12",
"title": "Building a Terrorism Ontology",
"authors": [
{
"first": "A",
"middle": [],
"last": "Mannes",
"suffix": ""
},
{
"first": "J",
"middle": [],
"last": "Golbeck",
"suffix": ""
}
],
"year": 2005,
"venue": "Proceedings of the ISWC Workshop on Ontology Patterns for the Semantic Web",
"volume": "36",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Mannes, A., Golbeck J. (2005). Building a Terrorism Ontology/ Proceedings of the ISWC Workshop on Ontology Patterns for the Semantic Web 36. URL: https://pdfs.semanticscholar.org/9bcb/90e48677e39da7 b84939e8c8da2b2a63cde7.pdf (25.09.2019).",
"links": null
},
"BIBREF13": {
"ref_id": "b13",
"title": "The corpus-based study of language change in progress: The extra value of tagged corpora",
"authors": [
{
"first": "C",
"middle": [],
"last": "Mair",
"suffix": ""
}
],
"year": 2005,
"venue": "The AAACL/ICAME Conference",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Mair, C. (2005). The corpus-based study of language change in progress: The extra value of tagged corpora. The AAACL/ICAME Conference, Ann Arbor, 2005.",
"links": null
},
"BIBREF14": {
"ref_id": "b14",
"title": "Modelling Multilinguality in Ontologies",
"authors": [
{
"first": "E",
"middle": [],
"last": "Montiel-Ponsoda",
"suffix": ""
},
{
"first": "G",
"middle": [],
"last": "Aguado De Cea",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "G\u00f3mez-P\u00e9rez",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "Peters",
"suffix": ""
}
],
"year": 2008,
"venue": "Proceedings of COLING 2008, Companion volume -Posters and Demonstrations",
"volume": "",
"issue": "",
"pages": "67--70",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Montiel-Ponsoda E., Aguado de Cea G., G\u00f3mez-P\u00e9rez A., Peters A. (2008). Modelling Multilinguality in Ontologies. Proceedings of COLING 2008, Companion volume -Posters and Demonstrations. pp. 67-70.",
"links": null
},
"BIBREF15": {
"ref_id": "b15",
"title": "Introduction to WordNet: An Online Lexical Database",
"authors": [
{
"first": "G",
"middle": [
"A"
],
"last": "Miller",
"suffix": ""
},
{
"first": "R",
"middle": [],
"last": "Beckwith",
"suffix": ""
},
{
"first": "C",
"middle": [],
"last": "Fellbaum",
"suffix": ""
},
{
"first": "D",
"middle": [],
"last": "Gross",
"suffix": ""
},
{
"first": "K",
"middle": [
"J"
],
"last": "Miller",
"suffix": ""
}
],
"year": 1990,
"venue": "International Journal of Lexicography",
"volume": "3",
"issue": "4",
"pages": "235--244",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Miller, G.A., Beckwith, R., Fellbaum, C., Gross, D., Miller, K.J. (1990). Introduction to WordNet: An On- line Lexical Database. International Journal of Lexicography 3 (4), pp. 235-244.",
"links": null
},
"BIBREF16": {
"ref_id": "b16",
"title": "From Text to Ontology Extraction and Representation of Conceptual Information",
"authors": [
{
"first": "A",
"middle": [],
"last": "Moreno",
"suffix": ""
},
{
"first": "P\u00e9rez",
"middle": [],
"last": "Ch",
"suffix": ""
}
],
"year": 2011,
"venue": "Actes de quatri\u00e8mes rencontres \u00abTerminologie et Intelligence Artifi-cielle",
"volume": "",
"issue": "",
"pages": "233--242",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Moreno A., P\u00e9rez Ch. (2011). From Text to Ontology Extraction and Representation of Conceptual Information. Actes de quatri\u00e8mes rencontres \u00abTerminologie et Intelligence Artifi-cielle\u00bb, pp.233- 242.",
"links": null
},
"BIBREF17": {
"ref_id": "b17",
"title": "Prediction of Terrorist Threat on the basis of Semantic Association acquisition and Complex Network Evolution",
"authors": [
{
"first": "A",
"middle": [],
"last": "Najgebauer",
"suffix": ""
},
{
"first": "R",
"middle": [],
"last": "Antkiewicz",
"suffix": ""
},
{
"first": "M",
"middle": [],
"last": "Chmielewski",
"suffix": ""
},
{
"first": "R",
"middle": [],
"last": "Kasprzyk",
"suffix": ""
}
],
"year": 2008,
"venue": "The Journal of Telecommunications and Information Technology",
"volume": "2",
"issue": "",
"pages": "14--20",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Najgebauer A., Antkiewicz R., Chmielewski M., Kasprzyk R., (2008). Prediction of Terrorist Threat on the basis of Semantic Association acquisition and Complex Network Evolution. The Journal of Telecommunications and Information Technology. Vol. 2. pp. 14-20.",
"links": null
},
"BIBREF18": {
"ref_id": "b18",
"title": "Linking Lexicons and Ontologies: Mapping WordNet to the Suggested Upper Merged Ontology",
"authors": [
{
"first": "I",
"middle": [],
"last": "Niles",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "Pease",
"suffix": ""
}
],
"year": 2003,
"venue": "Proceedings of the 2003 International Conference on Information and Knowledge Engineering",
"volume": "",
"issue": "",
"pages": "412--416",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Niles I. & Pease A. (2003). Linking Lexicons and Ontologies: Mapping WordNet to the Suggested Upper Merged Ontology. Proceedings of the 2003 International Conference on Information and Knowledge Engineering (IKE 03), pp. 412-416.",
"links": null
},
"BIBREF19": {
"ref_id": "b19",
"title": "Natural Language Annotation for Machine Learning",
"authors": [
{
"first": "S",
"middle": [],
"last": "Nirenburg",
"suffix": ""
},
{
"first": "V",
"middle": [],
"last": "Raskin",
"suffix": ""
}
],
"year": 2004,
"venue": "Cambridge Pustejovsky J",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Nirenburg S. & Raskin V. (2004). Ontological Semantics. MIT Press, Cambridge Pustejovsky J. (2012). Natural Language Annotation for Machine Learning. O'Reilly Media; 1 edition. 342 P.",
"links": null
},
"BIBREF20": {
"ref_id": "b20",
"title": "Building a semantically annotated corpus of clinical texts",
"authors": [
{
"first": "A",
"middle": [],
"last": "Roberts",
"suffix": ""
},
{
"first": "R",
"middle": [],
"last": "Gaizauskas",
"suffix": ""
},
{
"first": "M",
"middle": [],
"last": "Hepple",
"suffix": ""
},
{
"first": "G",
"middle": [],
"last": "Demetriou",
"suffix": ""
},
{
"first": "Y",
"middle": [],
"last": "Guo",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "Roberts",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "Setzer",
"suffix": ""
}
],
"year": 2009,
"venue": "Journal of Biomedical Informatics",
"volume": "42",
"issue": "5",
"pages": "950--966",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Roberts A., Gaizauskas R., Hepple M., Demetriou G., Guo Y., Roberts A., Setzer A. (2009). Building a semantically annotated corpus of clinical texts. Journal of Biomedical Informatics. -Vol. 42 (5), pp. 950-966.",
"links": null
},
"BIBREF21": {
"ref_id": "b21",
"title": "Automatic Extraction of Linguistic Resources in Multiple Languages",
"authors": [
{
"first": "S",
"middle": [],
"last": "Sheremetyeva",
"suffix": ""
}
],
"year": 2012,
"venue": "Proceedings of NLPCS 2012, 9th International Workshop on Natural Language Processing and Cognitive Science in conjunction with ICEIS 2012",
"volume": "",
"issue": "",
"pages": "44--52",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Sheremetyeva S. (2012). Automatic Extraction of Linguistic Resources in Multiple Languages. Proceedings of NLPCS 2012, 9th International Workshop on Natural Language Processing and Cognitive Science in conjunction with ICEIS 2012, Wroclaw, Poland, pp. 44-52.",
"links": null
},
"BIBREF22": {
"ref_id": "b22",
"title": "On Modelling Domain Ontology Knowledge for Processing Multilingual Texts of Terroristic Content",
"authors": [
{
"first": "S",
"middle": [],
"last": "Sheremetyeva",
"suffix": ""
},
{
"first": "A",
"middle": [],
"last": "Zinovyeva",
"suffix": ""
}
],
"year": 2018,
"venue": "Communications in Computer and Information Science",
"volume": "859",
"issue": "",
"pages": "368--379",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Sheremetyeva S. & Zinovyeva A. (2018). On Modelling Domain Ontology Knowledge for Processing Multilingual Texts of Terroristic Content. Communications in Computer and Information Science, 859. Springer, Cham, pp. 368-379.",
"links": null
},
"BIBREF23": {
"ref_id": "b23",
"title": "On Integrating Hybrid and Rule-Based Components For Patent MT with Several Levels of Output. Proceedings of \"The Fifth Workshop on Patent Translation in conjunction of the fourteenth Machine Translation Summit",
"authors": [
{
"first": "S",
"middle": [],
"last": "Sheremetyeva",
"suffix": ""
}
],
"year": 2006,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Sheremetyeva S. (2013). On Integrating Hybrid and Rule- Based Components For Patent MT with Several Levels of Output. Proceedings of \"The Fifth Workshop on Patent Translation in conjunction of the fourteenth Machine Translation Summit 2013\", Nice, France, September 2-6.",
"links": null
},
"BIBREF24": {
"ref_id": "b24",
"title": "BRAT: a Web-based Tool for NLP-Assisted Text Annotation",
"authors": [
{
"first": "P",
"middle": [],
"last": "Stenetorp",
"suffix": ""
},
{
"first": "S",
"middle": [],
"last": "Pyysalo",
"suffix": ""
},
{
"first": "G",
"middle": [],
"last": "Topic",
"suffix": ""
},
{
"first": "T",
"middle": [],
"last": "Ohta",
"suffix": ""
},
{
"first": "S",
"middle": [],
"last": "Ananiadou",
"suffix": ""
},
{
"first": "'",
"middle": [],
"last": "Jun",
"suffix": ""
},
{
"first": "J",
"middle": [],
"last": "",
"suffix": ""
}
],
"year": 2012,
"venue": "Proceedings of the 13th Conference of the European Chapter of the Association for Computational Linguistics",
"volume": "",
"issue": "",
"pages": "102--107",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Stenetorp P., Pyysalo S., Topic G., Ohta T., Ananiadou S., Jun'ichiTsujii J. (2012). BRAT: a Web-based Tool for NLP-Assisted Text Annotation. Proceedings of the 13th Conference of the European Chapter of the Association for Computational Linguistics. Avignon, France, April 23 -27. 2012. pp. 102-107.",
"links": null
},
"BIBREF25": {
"ref_id": "b25",
"title": "Ontology-Based Classification of News in an Electronic Newspaper. International Book Series",
"authors": [
{
"first": "L",
"middle": [],
"last": "Tenenboim",
"suffix": ""
},
{
"first": "B",
"middle": [],
"last": "Shapira",
"suffix": ""
},
{
"first": "P",
"middle": [],
"last": "Shoval",
"suffix": ""
}
],
"year": 2008,
"venue": "Information Science and Computing",
"volume": "",
"issue": "",
"pages": "89--97",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Tenenboim L., Shapira B, Shoval P. (2008). Ontology- Based Classification of News in an Electronic Newspaper. International Book Series \"Information Science and Computing\", pp. 89-97.",
"links": null
},
"BIBREF26": {
"ref_id": "b26",
"title": "Terminology extraction and automatic indexing -comparison and qualitative evaluation of methods",
"authors": [
{
"first": "H",
"middle": [
"F"
],
"last": "Witschel",
"suffix": ""
}
],
"year": 2005,
"venue": "Terminology and Knowledge Engineering",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Witschel H. F. (2005). Terminology extraction and automatic indexing -comparison and qualitative evaluation of methods. Terminology and Knowledge Engineering (TKE)",
"links": null
},
"BIBREF27": {
"ref_id": "b27",
"title": "Multilingual language resources and interoperability",
"authors": [
{
"first": "A",
"middle": [],
"last": "Witt",
"suffix": ""
},
{
"first": "U",
"middle": [],
"last": "Heid",
"suffix": ""
},
{
"first": "F",
"middle": [],
"last": "Sasaki",
"suffix": ""
},
{
"first": "Gilles",
"middle": [],
"last": "S\u00e9rasset",
"suffix": ""
}
],
"year": 2009,
"venue": "Lang Resources & Evaluation",
"volume": "43",
"issue": "",
"pages": "1--14",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Witt, A., Heid, U., Sasaki, F., Gilles S\u00e9rasset (2009). Multilingual language resources and interoperability. Lang Resources & Evaluation 43, 1-14 (2009).",
"links": null
},
"BIBREF29": {
"ref_id": "b29",
"title": "Sistema semanticheskoj razmetki korpusa tekstov v ogranichennoj predmetnoj oblasti",
"authors": [
{
"first": "M",
"middle": [
"Yu"
],
"last": "Zagorul'ko",
"suffix": ""
},
{
"first": "I",
"middle": [
"S"
],
"last": "Kononenko",
"suffix": ""
},
{
"first": "E",
"middle": [
"A"
],
"last": "Sidorova",
"suffix": ""
}
],
"year": 2012,
"venue": "Proceeding of the international conference Komp'yuternaya lingvistika i intellektual'nye tekhnologii",
"volume": "",
"issue": "",
"pages": "674--683",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Zagorul'ko, M. YU., Kononenko I. S., Sidorova E. A. (2012). Sistema semanticheskoj razmetki korpusa tekstov v ogranichennoj predmetnoj oblasti. Proceeding of the international conference Komp'yuternaya lingvistika i intellektual'nye tekhnologii, pp. 674-683. (in Russian).",
"links": null
}
},
"ref_entries": {
"FIGREF0": {
"type_str": "figure",
"num": null,
"uris": null,
"text": "fragments of unilingual lexica lists assigned to the conceptual class \"AGENT -TERRORIST\". AGENT -TERRORIST: Executor of a terrorist act ASSUMPTION: Assumption on who could commit a terr. act CAUSE: What caused a terrorist attack CLAIM RESPOSSIBILITY: terr. act responsibility claims CONSEQUENCES: Aftermath of the terrorist attack COUNTER-TERRORISM: People and measures against terr. GOAL OF ATTACK: Demands of terrorists LOCATION: Place where a terrorist act was committed MEANS OF ATTACK: Items used for a terrorist act NATION: person citizenship or country related to terrorism OBJECT OF ATTACK: Who or what was hit in terr. act TIME: Date and time when the terrorist attack happened TYPE OF ATTACK: shooting, explosion, stubbing, arson SOURCE : Sources of attack reports: newspapers, TV, etc."
},
"FIGREF1": {
"type_str": "figure",
"num": null,
"uris": null,
"text": "A fragment of the main TransDict interface."
},
"FIGREF2": {
"type_str": "figure",
"num": null,
"uris": null,
"text": "TransDict pop-up window for lexicalontological mapping and assigning concept tags."
},
"FIGREF3": {
"type_str": "figure",
"num": null,
"uris": null,
"text": "TransDict pop-up window for lexicalontological analysis."
},
"FIGREF4": {
"type_str": "figure",
"num": null,
"uris": null,
"text": "The tagger interface showing the results automatic coarse-grained conceptual annotation."
},
"TABREF0": {
"num": null,
"type_str": "table",
"content": "<table><tr><td>COUNTER-TERRORISM</td></tr><tr><td>COUNTER-TERRORISM AGENT : People fighting terr.</td></tr><tr><td>COUNTER-TERRORISM MEASURES: counter-terr. action</td></tr><tr><td>CONSEQUENCES</td></tr><tr><td>PUBLIC LOSS: killed, wounded, hostage, no damage</td></tr><tr><td>DESTRUCTION: objects damaged or destructed</td></tr><tr><td>TERRORISTS' LOSS: suicided, killed, wounded, detained</td></tr><tr><td>TERRORISTS' GAIN: terrorists' demands answered</td></tr><tr><td>PUBLIC REACTION: manifestation of support</td></tr><tr><td>RECONSTRUCTION: restoration of destroyed objects</td></tr></table>",
"text": "A fragment of the domain conceptual class list.",
"html": null
},
"TABREF1": {
"num": null,
"type_str": "table",
"content": "<table><tr><td>Language</td><td>Most frequent domain lexica of the class</td></tr><tr><td>English</td><td>terrorist, militant, fighter, gunman, suicide</td></tr><tr><td/><td>bomber, jihadi, female suicide bomber,</td></tr><tr><td/><td>female terrorist, lone-wolf terrorist, ISIS</td></tr><tr><td/><td>terrorist</td></tr><tr><td>French</td><td>terroriste, kamikaze, combattant, femme</td></tr><tr><td/><td>kamikaze, djihadiste. loup solitaire, terroriste de</td></tr><tr><td/><td>l'EI, combattant terroriste, femme terroriste</td></tr><tr><td>Russian</td><td>\u0442\u0435\u0440\u0440\u043e\u0440\u0438\u0441\u0442, \u0431\u043e\u0435\u0432\u0438\u043a, \u0441\u043c\u0435\u0440\u0442\u043d\u0438\u043a, \u0442\u0435\u0440\u0440\u043e\u0440\u0438\u0441\u0442-</td></tr><tr><td/><td>\u0441\u043c\u0435\u0440\u0442\u043d\u0438\u043a, \u0442\u0435\u0440\u0440\u043e\u0440\u0438\u0441\u0442\u043a\u0430-\u0441\u043c\u0435\u0440\u0442\u043d\u0438\u0446\u0430, \u0438\u0433\u0438\u043b\u043e\u0432\u0435\u0446,</td></tr><tr><td/><td>\u0442\u0435\u0440\u0440\u043e\u0440\u0438\u0441\u0442\u043a\u0430, \u0434\u0436\u0438\u0445\u0430\u0434\u0438\u0441\u0442, \u0442\u0435\u0440\u0440\u043e\u0440\u0438\u0441\u0442-</td></tr><tr><td/><td>\u043e\u0434\u0438\u043d\u043e\u0447\u043a\u0430</td></tr></table>",
"text": "Second level concepts for the top concepts COUNTER-TERRORISM and CONSEQUENCES.",
"html": null
}
}
}
} |