File size: 74,471 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 |
{
"paper_id": "2019",
"header": {
"generated_with": "S2ORC 1.0.0",
"date_generated": "2023-01-19T07:30:05.395273Z"
},
"title": "Kunji : A Resource Management System for Higher Productivity in Computer Aided Translation Tools",
"authors": [
{
"first": "Priyank",
"middle": [],
"last": "Gupta",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "IIIT-Hyderabad",
"location": {}
},
"email": "priyank.gupta@research.iiit.ac.in"
},
{
"first": "Rashid",
"middle": [],
"last": "Ahmad",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "IIIT-Hyderabad",
"location": {}
},
"email": "rashid.ahmadpg08@research.iiit.ac.in"
},
{
"first": "Manish",
"middle": [],
"last": "Shrivastava",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "IIIT-Hyderabad",
"location": {}
},
"email": "m.shrivastava@iiit.ac.in"
},
{
"first": "Dipti",
"middle": [],
"last": "Misra",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "IIIT-Hyderabad",
"location": {}
},
"email": ""
},
{
"first": "Sharma",
"middle": [],
"last": "Ltrc",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "IIIT-Hyderabad",
"location": {}
},
"email": ""
}
],
"year": "",
"venue": null,
"identifiers": {},
"abstract": "Complex NLP applications, such as machine translation systems, utilize various kinds of resources namely lexical, multiword, domain dictionaries, maps and rules etc. Similarly, translators working on Computer Aided Translation workbenches, also require help from various kinds of resources-glossaries, terminologies, concordances and translation memory in the workbenches in order to increase their productivity. Additionally, translators have to look away from the workbenches for linguistic resources like Named Entities, Multiwords, lexical and lexeme dictionaries in order to get help, as the available resources like concordances, terminologies and glossaries are often not enough. In this paper we present Kunji, a resource management system for translation workbenches and MT modules. This system can be easily integrated in translation workbenches and can also be used as a management tool for resources for MT systems. The described resource management system has been integrated in a translation workbench Transzaar. We also study the impact of providing this resource management system along with linguistic resources on the productivity of translators for English-Hindi language pair. When the linguistic resources like lexeme, NER and MWE dictionaries were made available to translators in addition to their regular translation memories, concordances and terminologies, their productivity increased by 15.61%.",
"pdf_parse": {
"paper_id": "2019",
"_pdf_hash": "",
"abstract": [
{
"text": "Complex NLP applications, such as machine translation systems, utilize various kinds of resources namely lexical, multiword, domain dictionaries, maps and rules etc. Similarly, translators working on Computer Aided Translation workbenches, also require help from various kinds of resources-glossaries, terminologies, concordances and translation memory in the workbenches in order to increase their productivity. Additionally, translators have to look away from the workbenches for linguistic resources like Named Entities, Multiwords, lexical and lexeme dictionaries in order to get help, as the available resources like concordances, terminologies and glossaries are often not enough. In this paper we present Kunji, a resource management system for translation workbenches and MT modules. This system can be easily integrated in translation workbenches and can also be used as a management tool for resources for MT systems. The described resource management system has been integrated in a translation workbench Transzaar. We also study the impact of providing this resource management system along with linguistic resources on the productivity of translators for English-Hindi language pair. When the linguistic resources like lexeme, NER and MWE dictionaries were made available to translators in addition to their regular translation memories, concordances and terminologies, their productivity increased by 15.61%.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Abstract",
"sec_num": null
}
],
"body_text": [
{
"text": "NLP applications -machine translation systems and translation workbenches (which can have multiple MT systems integrated), are complex in nature as they are built with various complex heterogeneous NLP modules. These complex NLP applications are compute and knowledge intensive in nature and require various types of resources at different levels of processing.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "The translators using the translation workbenches have to look for various resources -glossaries, terminologies, concordances and translation memory in order to supplement their translation tasks. It is still hard to maintain consistency while maintaining high productivity. Sometimes, the aid the translators get from the resources is not enough and they have to look for linguistic resources -Named Entities, lexical, bilingual, Multiwords and lexeme dictionaries -offline or online for the correct and appropriate meaning of a given phrase or word leading to additional cognitive load on translators.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "In MT system, when MT modules execute on the given input, various types of resources namely lexical, bilingual, domain, multiword dictionaries, paradigms, maps files and rules are required at various stages of the processing. With the exploration of Sampark and Anusaaraka MT systems (ILMT, 2009; Chaudhury et al., 2010) , it has been found that the resources can be used in a more efficient way in the both systems. These resources are developed by by NLP researchers who face resource management issues like exporting, importing, storage etc. For a large scale NLP application development, inefficient storage and management of these resources can become a bottleneck for productivity.",
"cite_spans": [
{
"start": 284,
"end": 296,
"text": "(ILMT, 2009;",
"ref_id": "BIBREF8"
},
{
"start": 297,
"end": 320,
"text": "Chaudhury et al., 2010)",
"ref_id": "BIBREF3"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "For the translation tasks and MT applications' development for Indian languages, the above mentioned issues become more critical as the digital content available in Indian languages are way less compared to other languages. As Hindi is one of the top 5 spoken languages of the world but still its digital content available on the web is less than 0.05% while the content of the other Indian languages are still lower. Hence the need of a resource management system arises which can ad-dress the issues of both translators as well as MT modules' development.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "In this paper, we present a resource management system named Kunji, which addresses the issues related to resources of both translation workbenches and MT system development for Indian languages. It can be used in translation workbenches to facilitate the translators to use additional linguistic resources -domain, lexical, named entities, multiword and WSD dictionaries -along with the terminologies, concordances and glossaries. It facilitates the translators in two ways -i) reduction of their repetitive load by providing searching and filter mechanism and ii) reuse of their previously translated words, phrases and sentences by providing the provision of personal and domain dictionaries. It shifts the problem of recall to recognition which causes less cognitive load on translators by providing them provision to search the corresponding terms and recognize. It also facilitates the NLP researchers to manage the resources for MT modules in a robust manner while addressing their issues of management, export, import, formats and efficient search etc. It also lets them evaluate and verify their work from other senior expert. Such system facilitates the large scale MT applications development by making the resources available to multiple modules and their facilitation to process and reuse them efficiently in an MT system.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Additionally we performed an experiment to study the impact of providing this resource management system along with linguistic resources on the productivity of translators. The described resource management system, Kunji, has been integrated in the translation workbench Transzaar (Ahmad et al., 2018) . The resources with linguistic features like lexeme, NER and MWE dictionaries have been provided in Kunji to facilitate translators in addition to their regular translation memories, concordances and terminologies for English-Hindi language pair. We observed that the productivity of the translators increased by 15.61% when the resources with linguistic features were provided to the translators in addition to their regular resources (translation memories, glossaries, concordances and terminologies).",
"cite_spans": [
{
"start": 281,
"end": 301,
"text": "(Ahmad et al., 2018)",
"ref_id": "BIBREF0"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "In this paper, background and motivation for Kunji is described in Section-2 followed by related work in Section-3. The detailed architecture and functionality of Kunji is described in Section-4. Section-5 describes the experiment performed on Kunji followed by its results and discussion in Section-6. In Section-7, we conclude our work.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "In current scenario, various translation workbenches(CAT tools) are available. Some of them offer the resources like translation memories(TM), concordances, terminologies and glossaries to translators but there is a lack of the resource management systems which provide the provision by which a translator can use additional linguistic resources like NE, MWE and lexeme dictionaries. Additionally there is a lack of the provisions of reuse, management and verification of these resources which can lead a process to improvement and development of the MT systems.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Background and Motivation",
"sec_num": "2"
},
{
"text": "There are several tools for resource development but they are built with separate purposes and languages as they are for task specific. Like for annotators they have separate tools in Indian languages like Sanchay (Singh, 2008) but it is desktop based application. Some tools like GATE (Cunningham, 2002) addresses the some of the problems faced by NLP researchers while developing NLP resources but it is desktop based application.",
"cite_spans": [
{
"start": 214,
"end": 227,
"text": "(Singh, 2008)",
"ref_id": "BIBREF11"
},
{
"start": 286,
"end": 304,
"text": "(Cunningham, 2002)",
"ref_id": "BIBREF4"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Background and Motivation",
"sec_num": "2"
},
{
"text": "It motivated us to develop Kunji, a resource management tool, which facilitates the translators in translator workbenches as well as MT system development process. It tries to address the issues faced by both translators as well as MT module developers. It allows the translators to use and manage the linguistic resources along with their terminologies and glossaries.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Background and Motivation",
"sec_num": "2"
},
{
"text": "There are some translation workbenches with separate resource management systems such as: Smartcat 1 , Memsource 2 and Matecat (Federico et al., 2014) are web based CAT tools, provide central resources management for a project but they do not provide the provision for the linguistic resources. CASMACAT (Alabau et al., 2014 ) is a translation workbench which is web based and offers advanced functionality for computer-aided translation. It offers TMs but it also lacks the provision for the linguistic resources. PET (Aziz et al., 2012 ) is a tool to postedit to for evaluating the quality of translations. It evaluates the efforts required in translations in order to be fixed. SDL Trados, MemoQ 3 and Anubis (Jaworski, 2013) are CAT tools which are not web based and both lack the provision to use linguistic resources.",
"cite_spans": [
{
"start": 127,
"end": 150,
"text": "(Federico et al., 2014)",
"ref_id": "BIBREF6"
},
{
"start": 304,
"end": 324,
"text": "(Alabau et al., 2014",
"ref_id": "BIBREF1"
},
{
"start": 519,
"end": 537,
"text": "(Aziz et al., 2012",
"ref_id": "BIBREF2"
},
{
"start": 712,
"end": 728,
"text": "(Jaworski, 2013)",
"ref_id": "BIBREF9"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "3"
},
{
"text": "Brat (Stenetorp et al., 2012) is basically a web based annotation tool which focuses mainly on text annotations to enhance annotators productivity by closely integrating NLP technology into the annotation process. It doesnt address the issues in resources management with respect to translators and large scale NLP application development. Creating language resources for NLP in Indian languages (Sangal and Sharma, 2001 ) defines a novel idea in which the development of lexical resources is linked with an example NLP application like MT then it can act as a test bed for the developing resources and provide constant feedback. Sanchay (Singh, 2008) provides an extra layer of easily customizable language encoding support for less computerized languages along with an editor named Sanchay with different types of fonts and language encoding supports but it is desktop based application. GATE (Cunningham, 2002) enables users to develop and deploy the language engineering tools and resources in a robust manner. It is a desktop based tool which supports many NLP and information extraction tasks in multiple languages. Nancy et al (Ide and Romary, 2009) presents an abstract data model and its implementation for linguistic annotations. Annomarket (Dimitrov et al., 2014) described a cloud-based open platform for text mining, which aims to assist the development and deployment of robust, largescale text processing applications.",
"cite_spans": [
{
"start": 5,
"end": 29,
"text": "(Stenetorp et al., 2012)",
"ref_id": "BIBREF12"
},
{
"start": 396,
"end": 420,
"text": "(Sangal and Sharma, 2001",
"ref_id": "BIBREF10"
},
{
"start": 638,
"end": 651,
"text": "(Singh, 2008)",
"ref_id": "BIBREF11"
},
{
"start": 895,
"end": 913,
"text": "(Cunningham, 2002)",
"ref_id": "BIBREF4"
},
{
"start": 1134,
"end": 1156,
"text": "(Ide and Romary, 2009)",
"ref_id": "BIBREF7"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "3"
},
{
"text": "We see that in some translation workbenches, the resource management systems for TM, Terminologies and glossaries are available but they don't provide the provision for linguistic resources as well.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Related Work",
"sec_num": "3"
},
{
"text": "We propose Kunji-a Resources management System, which is a web based system based on microservices architecture. It can be easily integrated to a translation workbench and provides a mechanism to facilitate translators to use and manage the various types of resources (i.e. like terminologies, glossaries and domain dictionaries along with linguistic resources(lexeme dictionar-ies, NERs, MWEs etc.)). It does not only facilitate the translators but also facilitates the MT modules development by facilitating language researchers to manage the language engineering resources in a robust manner for various language processing tasks and evaluate them which boosts the process of large scale MT development.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Kunji : Resource Management System",
"sec_num": "4"
},
{
"text": "Microservices (Th\u00f6nes, 2015) based approach is an architectural concept of developing an application or software systems according to which a functionality or process can be developed, deployed and tested independently. An application is divided into set of such modular components or functionalities in which each functionality is an independent or disjoint from the others' in the application. In this approach, each functionality or component has a proper structured interface which is called API and is used to communicate with the corresponding module. It helps to overcome the drawbacks of monolithic approach. The complex NLP applications-especially MT systems or translation workbenches are knowledge and compute intensive in nature. So management and development of the resources for them requires deep knowledge of corresponding domain, nature of language while implementing algorithm for it requires knowledge of both streams NLP and computer science. For the development of the resource management tools for a translation workbench, the requirements of translators need to be understood. Many existing NLP applications for resource management follow monolithic design and hence static in nature. In NLP, for building scalable, distributed resource management systems , microservices based architecture can overcome the barriers of the monolithic architecture.",
"cite_spans": [
{
"start": 14,
"end": 28,
"text": "(Th\u00f6nes, 2015)",
"ref_id": "BIBREF13"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Microservices Based Architecture and NLP Applications",
"sec_num": "4.1"
},
{
"text": "Various NLP applications -CAT tools have been analyzed for their resource management, their procedures to utilize different kinds of resources. For the translation workbenches, various aspects like their reuse, verification and sharing of resources in a projects among various translators which are linked to a project. In MT modules, the procedures to access, management and reuse of the resources have been analyzed along with their functional and development requirements. We figured out the fol-lowing Architectural requirements for the system which are given below:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Architectural requirements",
"sec_num": "4.2"
},
{
"text": "Web based tool:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Microservices based architecture and",
"sec_num": "4.2.1"
},
{
"text": "The need of microservices based architecture arise to overcome the barriers imposed by monolithic architecture which also facilitates the creation of the dynamic web based tool for the same. In such paradigm, every functionality of the resource management system will be exposed as an independent service which can be utilized by the resource management in translation workbenches in order to facilitate the translators. It can also facilitate the NLP resource management for large scale MT development.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Microservices based architecture and",
"sec_num": "4.2.1"
},
{
"text": "Today is the era of the web, everything we see and work on is on web or is exposed as a web based service or tool. Due to mobile and internet evolution, many people are internet friendly even non-technical people. So in order to make people use and develop resources in more efficient way it is a major need to make it available in web as a form of web-tool which will also makes the application platform independent.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Microservices based architecture and",
"sec_num": "4.2.1"
},
{
"text": "The resources management system should be designed in such a way so that it should be able to integrate in the translation workbench in order to facilitate translators to increase their productivity. So that translators can manage and use their own private dictionaries, domain dictionaries, terminologies among with the project or task specific dictionaries.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Facilitation in translation workbench:",
"sec_num": "4.2.2"
},
{
"text": "It should support the translators to use, reuse and manage the linguistic resources like NE, MWE and lexeme dictionaries dictionaries as well.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Support for linguistic resources in translation workbenches:",
"sec_num": "4.2.3"
},
{
"text": "We explored the various aspects of NLP resources management and the ways in which they can be utilized in a translation workbenches to aid the translation tasks to translators as well as their facilitation to NLP researchers and annotators towards development of large Scale MT systems. Taking such requirements in consideration, we designed the architecture of the system such that it is a collection of independent microservices for each of the functionality which can be deployed independently. We resolve the problem of monolithic architecture by exposing each functionality as a microservice which can be executed and interacted via RESTful API.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Our Architecture",
"sec_num": "4.3"
},
{
"text": "The architecture has been shown in Fig. 1 following with it's explanation.",
"cite_spans": [],
"ref_spans": [
{
"start": 35,
"end": 41,
"text": "Fig. 1",
"ref_id": "FIGREF0"
}
],
"eq_spans": [],
"section": "Kunji : Architecture Explanation",
"sec_num": "4.3.1"
},
{
"text": "\u2022 Application logic : It is the block or portion containing the complete functionality of the system. Hence the system's functionalities are structured, designed and implemented in the functional way independently.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Kunji : Architecture Explanation",
"sec_num": "4.3.1"
},
{
"text": "\u2022 In-Memory Resource Loading : The different types of resources are required in different modules of MT as well as for providing help in the translation workbenches. Their loading from database at each execution process makes process slow as well as it imposes an unnecessary load to the system. Hence the services of the system are designed in such a way that resources load at the time of the starting of the service not at the time of invocation of the service.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Kunji : Architecture Explanation",
"sec_num": "4.3.1"
},
{
"text": "\u2022 REST APIs : REST(Representational state transfer) API is basically an approach or technology for communications which is used in developing web services. It provides many methods such as HTTP or HTTPs for implementation. We expose each of the functionality of the resource management system as a microservice with REST API interface to interact with. The APIs are created for create, add, import, export, delete, link, update and verify.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Kunji : Architecture Explanation",
"sec_num": "4.3.1"
},
{
"text": "\u2022 UI The simple UI has been created with the HTML5 and JavaScript",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Kunji : Architecture Explanation",
"sec_num": "4.3.1"
},
{
"text": "First the configuration of a resource is defined followed by the corresponding entries of the data inside the resource as follows:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Structure of Resources",
"sec_num": "4.3.2"
},
{
"text": "\u2022 Configuration of a resource : It contains the following fields as given below in the First the configuration of the resource is created as the configuration mentioned in Table- 1:",
"cite_spans": [],
"ref_spans": [
{
"start": 172,
"end": 178,
"text": "Table-",
"ref_id": null
}
],
"eq_spans": [],
"section": "Structure of Resources",
"sec_num": "4.3.2"
},
{
"text": "\u2022 A data record of a resource will look like as mentioned in Table-2 : It contains the record with its target value and category and other required fields.",
"cite_spans": [],
"ref_spans": [
{
"start": 61,
"end": 70,
"text": "Table-2 :",
"ref_id": "TABREF3"
}
],
"eq_spans": [],
"section": "Structure of Resources",
"sec_num": "4.3.2"
},
{
"text": "\u2022 Explanation: First the configuration of a particular resource type is created then its corresponding data entries can be created. So after the resources are created then we read the corresponding configuration of the resource and load it in memory in order to make process efficient. ",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Structure of Resources",
"sec_num": "4.3.2"
},
{
"text": "For the controller part, for complete in-memory based architecture and for creating the microservices we used Java servlets with Apache-Tomcat-9 web server. For the storage of the resources we used MongoDB-3.6.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Technologies used",
"sec_num": "4.3.3"
},
{
"text": "For UI portion of the management system HTML5 along with CSS and Ajax have been used.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Technologies used",
"sec_num": "4.3.3"
},
{
"text": "This resource management system has been integrated with two translation workbenches Transzaar (Ahmad et al., 2018) . The features of the resource management system are given as follows:",
"cite_spans": [
{
"start": 95,
"end": 115,
"text": "(Ahmad et al., 2018)",
"ref_id": "BIBREF0"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Features:",
"sec_num": "4.3.4"
},
{
"text": "\u2022 Addition of the resources in single and bulk mode.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Features:",
"sec_num": "4.3.4"
},
{
"text": "\u2022 Editing and updating of the resources and their corresponding meta data.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Features:",
"sec_num": "4.3.4"
},
{
"text": "\u2022 Provision to export and reuse the resources in various tools.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Features:",
"sec_num": "4.3.4"
},
{
"text": "\u2022 Search: N-gram dynamic search mechanism.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Features:",
"sec_num": "4.3.4"
},
{
"text": "\u2022 Easy import the resource in a translation workbench.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Features:",
"sec_num": "4.3.4"
},
{
"text": "\u2022 Provision to develop and manage resources with linguistic features.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Features:",
"sec_num": "4.3.4"
},
{
"text": "\u2022 Provision to verify them with a senior language expert of the corresponding language.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Features:",
"sec_num": "4.3.4"
},
{
"text": "By using Kunji, translators can get aid in their translation tasks while using translation workbench. We demonstrated it with facilitating the translation workbench transzaar with Kunji. In kunji, they can create or import and use various kinds of resources like terminologies, bilingual and glossary dictionaries. Consider a scenario where a translator working on the workbench has some domain terminology dictionaries and wants to get help in his/her translation tasks. Then he can simply take that dictionary text file with source and target tab separated in each line in it. These resources can be imported/uploaded in the Kunji with the user metadata i.e. username, domain, language pair, and project etc.. So when translator opens his translation tasks and do editing then he can get the appropriate meanings of the terms or words of source text from terminology dictionary uploaded in Kunji. Those words are highlighted in source pane and on clicking them we can get appropriate meanings of the terms. Kunji facilitates a user working on workbench to add, import, export, update, n-gram search and edit the resources. Possible types of the resources which can be uploaded are terminologies, glossaries, domain and bilingual dictionaries etc. Kunji also facilitates translator to import and use linguistic resources namely lexemes, named entities(NE), multiword dictionaries which can be helpful in selecting the correct form of the meaning of a word they should use.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Process Flow",
"sec_num": "4.3.5"
},
{
"text": "It helps the translators by allowing them to search and look-up into their corresponding user specific or system specific resources, terminologies, glossaries which they can link with their translation tasks of a project to increase their productivity. Repetitive tasks are also facilitated by search or look up of a term which repeats many times in the task from the dictionaries and hence their consistency for given translation task would not be affected.Translators can import and use the linguistic resources like lexeme, NE and MWE dictionaries which can be helpful to decrease their cognitive load like selecting the correct form of the meaning of a word they should use. An experiment has been performed and is explained in Section-5 which would show how the linguistics resources can be beneficial for translators when we use this resource management system with a translation workbench.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Benefits of Kunji",
"sec_num": "4.4"
},
{
"text": "Kunji also provides various aspects for managing and developing resources for MT application development. For NER and MWE, it provides provisions of category and subcategories like person/location/organization or Idiom/multiword phrases/Domain terms etc. Similarly, it also provides the ways to create different types of domain dictionaries according to the need. So the development process of MT can be hugely facilitated by using this NLP researchers-linguists, annotators and and NLP module developers can be benefited. It addresses issues like encodings, fonts, usage of the common resources and lexical dictionaries for multiple tools, import and export etc so using such tool can enhance their productivity.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Benefits of Kunji",
"sec_num": "4.4"
},
{
"text": "The described resource management system -Kunji has been integrated in translation workbenches Transzaar (Ahmad et al., 2018) . We study the impact of providing this resource management system in CAT tool-Transzaar along with linguistic resources on the productivity of translators for English-Hindi language pair. The various types of resources namely terminologies, glossaries, TM along with linguistic resources like lexeme, NE and MWE dictionaries were made available in the resources management tool in the workbench.",
"cite_spans": [
{
"start": 105,
"end": 125,
"text": "(Ahmad et al., 2018)",
"ref_id": "BIBREF0"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Experiment",
"sec_num": "5"
},
{
"text": "We setup the experiment of translation tasks from the news data of \"The Hindu\" news website 4 .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Experimental setup",
"sec_num": "5.1"
},
{
"text": "For the experiment, 20 news stories from The Hindu news paper From national domain have been taken. The data stories have been divided into 4 sets with each set containing 5 stories. The details of the data sets are given in Table 3 and Table 4 .",
"cite_spans": [],
"ref_spans": [
{
"start": 225,
"end": 245,
"text": "Table 3 and Table 4",
"ref_id": "TABREF5"
}
],
"eq_spans": [],
"section": "Data Set",
"sec_num": "5.1.1"
},
{
"text": "We integrated the Kunji, the resource management system to the Tanszaar and integrated the resources namely terminologies, glossaries, TM along with linguistic resources like lexeme, NE and MWE dictionaries were made available in the resources management tool in the workbench. We divided the experiment of translation tasks into 6 scenarios which are described in Table 5 . We have chosen the English-Hindi language pair for our experiment and created the translation tasks from the dataset for each of the mentioned scenarios in Table 5 and assigned them to the four translators. The translation tasks have been assigned in such a way that no translator would repeat the same set again in a given scenario.",
"cite_spans": [],
"ref_spans": [
{
"start": 365,
"end": 372,
"text": "Table 5",
"ref_id": "TABREF7"
},
{
"start": 531,
"end": 538,
"text": "Table 5",
"ref_id": "TABREF7"
}
],
"eq_spans": [],
"section": "Data Set",
"sec_num": "5.1.1"
},
{
"text": "The results of experiments are given in Table 6 and Table 7 . Each cell in the Table 6 presents the total time taken(in hours) by corresponding translator in the post-editing of the each set. The cells in the last row present the average time taken on all sets in the corresponding scenario. Table 7 presents per sentence time taken by each translator(in minutes) in the 6 different scenarios given in Table 5 .",
"cite_spans": [],
"ref_spans": [
{
"start": 40,
"end": 60,
"text": "Table 6 and Table 7",
"ref_id": "TABREF9"
},
{
"start": 80,
"end": 87,
"text": "Table 6",
"ref_id": "TABREF9"
},
{
"start": 293,
"end": 300,
"text": "Table 7",
"ref_id": null
},
{
"start": 403,
"end": 410,
"text": "Table 5",
"ref_id": "TABREF7"
}
],
"eq_spans": [],
"section": "Results and Discussion",
"sec_num": "6"
},
{
"text": "From Table 6 and Table 7 we see the impact of a translation workbench with and without the availability of different kinds of resources. We observed the translation data with professional translators in Table 6 and Table 7 . Scenario-3 is when we use a workbench without any resources. Compared to Scenario-3, in Scenario-4 we use the workbench with translation resources but exclude linguistic resources. We observe that the productivity is significantly improved by 29.93%. When we contrast this with Scenario-5 where only linguistic resources are used then the productivity is only improved by 17.12% over Scenario-3. When we use all the resources including linguistic resources then the productivity is improved by 45.54% as compared to Scenario-3. Hence, we see that the productivity is improved additionally by 15.61% over Scenario-4 when the resources with linguistic features are made available for the translators in the workbench along with TM, Terms and Glossaries. From the above mentioned results, it is evident that the translators' productivity is positively affected when they get help from the utilization of resources with linguistic features in CAT tool.",
"cite_spans": [],
"ref_spans": [
{
"start": 5,
"end": 24,
"text": "Table 6 and Table 7",
"ref_id": "TABREF9"
},
{
"start": 203,
"end": 222,
"text": "Table 6 and Table 7",
"ref_id": "TABREF9"
}
],
"eq_spans": [],
"section": "Results and Discussion",
"sec_num": "6"
},
{
"text": "The scenario-wise time consumed per sentence translation by each of the translators is presented in Table 7 . It is observed that the productivity of each translator increases when we go from Scenario-1 to Scenario-6. For each translator, it is observed that their productivity increases as they use the translation workbench with resources. Their productivity significantly increase by more than 66% in scenario-6 i.e. when the linguistic resources are made available for the translators in the workbench along with TM, Terms and Glossaries. In scenario-6, the corresponding productivity of the translators are increased by 66%, 71.6%, 71.3% and 68.2% respectively. We also see the significant improvement in productivity when we go from scenario-4(workbench with translation resources but exclude linguistic resources) to scenario-6((workbench with translation resources but including linguistic resources).",
"cite_spans": [],
"ref_spans": [
{
"start": 100,
"end": 107,
"text": "Table 7",
"ref_id": null
}
],
"eq_spans": [],
"section": "Results and Discussion",
"sec_num": "6"
},
{
"text": "There are some observations of the translators about the data set. The Set-1 is the toughest of all four sets. It is evident from table-7 that, translators took most time in translation of this set in each of the scenario. The Set-4 is the easiest of the sets as less time is consumed in its translation. Translators found the linguistic dictionaries (lex- 1.84(S1) 0.938(S4) 1.058(S3) 1.195(S2) Scenario-2 0.78(S2) 1.03(S1) 0.569(S4) 0.61(S3) Scenario-3 0.6(S3) 0.733(S2) 1.07(S1) 0.52(S4) Scenario-4 0.506(S2) 0.76(S1) 0.359(S4) 0.43(S3) Scenario-5 0.41(S4) 0.49(S3) 0.59(S2) 0.88(S1) Scenario-6 0.64(S1) 0.267(S4) 0.304(S3) 0.38(S2) Improvement (in %)",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Results and Discussion",
"sec_num": "6"
},
{
"text": "71.6% 71.3% 68.2% Table 7 : Scenario-wise Results of Experiment with Professional Translators eme, named entities and MWE ) very helpful in the translation. For example the meaning of the word \"dreaded\" contains two senses in different contexts in Hindi. The lexeme dictionary aided the translators in disambiguation of word senses.",
"cite_spans": [],
"ref_spans": [
{
"start": 18,
"end": 25,
"text": "Table 7",
"ref_id": null
}
],
"eq_spans": [],
"section": "66%",
"sec_num": null
},
{
"text": "We have proposed a resource management system which addresses the issues of the translators as well as MT system development. Using Kunji, a translator can utilize, reuse and manage his resources which can be shared across projects and can be used further. Also the NLP researchers get benefited using it in their resources management in MT system development so that the resources can be utilized and managed in an efficient way. It also provides the mechanism which allows a translator to get help from the linguistic resourceslexeme, NER and MWE dictionaries in addition to their regular translation memories, concordances and terminologies. We performed experiments which show the effect of the linguistic resources in the productivity of the translators. We see that their productivity increased by 15.61% when they use linguistic resources along with their regular TM, terminologies and glossaries as mentioned in Table 6 and Table 7 . We hope to extend this system by analyzing the impact of more complex NLP module like Morphological analyzer in the translation pipeline.",
"cite_spans": [],
"ref_spans": [
{
"start": 920,
"end": 939,
"text": "Table 6 and Table 7",
"ref_id": "TABREF9"
}
],
"eq_spans": [],
"section": "Conclusion",
"sec_num": "7"
},
{
"text": "https://www.smartcat.ai 2 https://www.memsource.com",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "https://www.memoq.com/en",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
},
{
"text": "https://www.thehindu.com",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
}
],
"back_matter": [],
"bib_entries": {
"BIBREF0": {
"ref_id": "b0",
"title": "Transzaar: Empowers human translators",
"authors": [
{
"first": "Rashid",
"middle": [],
"last": "Ahmad",
"suffix": ""
},
{
"first": "Priyank",
"middle": [],
"last": "Gupta",
"suffix": ""
},
{
"first": "Nagaraju",
"middle": [],
"last": "Vuppala",
"suffix": ""
},
{
"first": "Ashutosh",
"middle": [],
"last": "Sanket Kumar Pathak",
"suffix": ""
},
{
"first": "Gagan",
"middle": [],
"last": "Kumar",
"suffix": ""
},
{
"first": "Sravan",
"middle": [],
"last": "Soni",
"suffix": ""
},
{
"first": "Manish",
"middle": [],
"last": "Kumar",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Shrivastava",
"suffix": ""
},
{
"first": "K",
"middle": [],
"last": "Avinash",
"suffix": ""
},
{
"first": "Arbind",
"middle": [
"K"
],
"last": "Singh",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Gangwar",
"suffix": ""
}
],
"year": 2018,
"venue": "",
"volume": "",
"issue": "",
"pages": "1--8",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Rashid Ahmad, Priyank Gupta, Nagaraju Vuppala, Sanket Kumar Pathak, Ashutosh Kumar, Gagan Soni, Sravan Kumar, Manish Shrivastava, Avinash K Singh, Arbind K Gangwar, et al. 2018. Transzaar: Empowers human translators. pages 1-8.",
"links": null
},
"BIBREF1": {
"ref_id": "b1",
"title": "Casmacat: A computer-assisted translation workbench",
"authors": [
{
"first": "Vicent",
"middle": [],
"last": "Alabau",
"suffix": ""
},
{
"first": "Christian",
"middle": [],
"last": "Buck",
"suffix": ""
},
{
"first": "Michael",
"middle": [],
"last": "Carl",
"suffix": ""
},
{
"first": "Francisco",
"middle": [],
"last": "Casacuberta ; Luis Leiva",
"suffix": ""
}
],
"year": 2014,
"venue": "",
"volume": "",
"issue": "",
"pages": "25--28",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Vicent Alabau, Christian Buck, Michael Carl, Fran- cisco Casacuberta, Mercedes Garc\u00eda-Mart\u00ednez, Ul- rich Germann, Jes\u00fas Gonz\u00e1lez-Rubio, Robin Hill, Philipp Koehn, Luis Leiva, et al. 2014. Casmacat: A computer-assisted translation workbench. pages 25-28.",
"links": null
},
"BIBREF2": {
"ref_id": "b2",
"title": "Pet: a tool for post-editing and assessing machine translation",
"authors": [
{
"first": "Wilker",
"middle": [],
"last": "Aziz",
"suffix": ""
},
{
"first": "Sheila",
"middle": [],
"last": "Castilho",
"suffix": ""
},
{
"first": "Lucia",
"middle": [],
"last": "Specia",
"suffix": ""
}
],
"year": 2012,
"venue": "",
"volume": "",
"issue": "",
"pages": "3982--3987",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Wilker Aziz, Sheila Castilho, and Lucia Specia. 2012. Pet: a tool for post-editing and assessing machine translation. pages 3982-3987.",
"links": null
},
"BIBREF3": {
"ref_id": "b3",
"title": "Anusaaraka: An expert system based machine translation system",
"authors": [
{
"first": "Sriram",
"middle": [],
"last": "Chaudhury",
"suffix": ""
},
{
"first": "Ankitha",
"middle": [],
"last": "Rao",
"suffix": ""
},
{
"first": "Dipti M",
"middle": [],
"last": "Sharma",
"suffix": ""
}
],
"year": 2010,
"venue": "",
"volume": "",
"issue": "",
"pages": "1--6",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Sriram Chaudhury, Ankitha Rao, and Dipti M Sharma. 2010. Anusaaraka: An expert system based machine translation system. pages 1-6.",
"links": null
},
"BIBREF4": {
"ref_id": "b4",
"title": "Gate, a general architecture for text engineering",
"authors": [
{
"first": "Hamish",
"middle": [],
"last": "Cunningham",
"suffix": ""
}
],
"year": 2002,
"venue": "Computers and the Humanities",
"volume": "36",
"issue": "2",
"pages": "223--254",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Hamish Cunningham. 2002. Gate, a general architec- ture for text engineering. Computers and the Hu- manities, 36(2):223-254.",
"links": null
},
"BIBREF5": {
"ref_id": "b5",
"title": "Annomarket-multilingual text analytics at scale on the cloud",
"authors": [
{
"first": "Marin",
"middle": [],
"last": "Dimitrov",
"suffix": ""
},
{
"first": "Hamish",
"middle": [],
"last": "Cunningham",
"suffix": ""
},
{
"first": "Ian",
"middle": [],
"last": "Roberts",
"suffix": ""
},
{
"first": "Petar",
"middle": [],
"last": "Kostov",
"suffix": ""
},
{
"first": "Alex",
"middle": [],
"last": "Simov",
"suffix": ""
}
],
"year": 2014,
"venue": "",
"volume": "",
"issue": "",
"pages": "315--319",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Marin Dimitrov, Hamish Cunningham, Ian Roberts, Petar Kostov, Alex Simov, Philippe Rigaux, and He- len Lippell. 2014. Annomarket-multilingual text analytics at scale on the cloud. pages 315-319.",
"links": null
},
"BIBREF6": {
"ref_id": "b6",
"title": "The matecat tool",
"authors": [
{
"first": "Marcello",
"middle": [],
"last": "Federico",
"suffix": ""
},
{
"first": "Nicola",
"middle": [],
"last": "Bertoldi",
"suffix": ""
},
{
"first": "Mauro",
"middle": [],
"last": "Cettolo",
"suffix": ""
},
{
"first": "Matteo",
"middle": [],
"last": "Negri",
"suffix": ""
},
{
"first": "Marco",
"middle": [],
"last": "Turchi",
"suffix": ""
},
{
"first": "Marco",
"middle": [],
"last": "Trombetti",
"suffix": ""
},
{
"first": "Alessandro",
"middle": [],
"last": "Cattelan",
"suffix": ""
},
{
"first": "Antonio",
"middle": [],
"last": "Farina",
"suffix": ""
},
{
"first": "Domenico",
"middle": [],
"last": "Lupinetti",
"suffix": ""
},
{
"first": "Andrea",
"middle": [],
"last": "Martines",
"suffix": ""
}
],
"year": 2014,
"venue": "",
"volume": "",
"issue": "",
"pages": "129--132",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Marcello Federico, Nicola Bertoldi, Mauro Cettolo, Matteo Negri, Marco Turchi, Marco Trombetti, Alessandro Cattelan, Antonio Farina, Domenico Lupinetti, Andrea Martines, et al. 2014. The mate- cat tool. pages 129-132.",
"links": null
},
"BIBREF7": {
"ref_id": "b7",
"title": "Standards for language resources",
"authors": [
{
"first": "Nancy",
"middle": [],
"last": "Ide",
"suffix": ""
},
{
"first": "Laurent",
"middle": [],
"last": "Romary",
"suffix": ""
}
],
"year": 2009,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {
"arXiv": [
"arXiv:0909.2719"
]
},
"num": null,
"urls": [],
"raw_text": "Nancy Ide and Laurent Romary. 2009. Stan- dards for language resources. arXiv preprint arXiv:0909.2719.",
"links": null
},
"BIBREF8": {
"ref_id": "b8",
"title": "Sampark: Machine translation system among indian languages",
"authors": [
{
"first": "",
"middle": [],
"last": "Group Ilmt",
"suffix": ""
}
],
"year": 2009,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "GROUP ILMT. 2009. Sampark: Machine transla- tion system among indian languages. http:// sampark.iiit.ac.in.",
"links": null
},
"BIBREF9": {
"ref_id": "b9",
"title": "Anubis-speeding up computeraided translation",
"authors": [
{
"first": "Rafa\u0142",
"middle": [],
"last": "Jaworski",
"suffix": ""
}
],
"year": 2013,
"venue": "",
"volume": "",
"issue": "",
"pages": "263--280",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Rafa\u0142 Jaworski. 2013. Anubis-speeding up computer- aided translation. pages 263-280.",
"links": null
},
"BIBREF10": {
"ref_id": "b10",
"title": "Creating language resources for nlp in indian languages 1. background",
"authors": [
{
"first": "Rajeev",
"middle": [],
"last": "Sangal",
"suffix": ""
},
{
"first": "Dipti Misra",
"middle": [],
"last": "Sharma",
"suffix": ""
}
],
"year": 2001,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Rajeev Sangal and Dipti Misra Sharma. 2001. Creat- ing language resources for nlp in indian languages 1. background.",
"links": null
},
"BIBREF11": {
"ref_id": "b11",
"title": "A mechanism to provide language-encoding support and an nlp friendly editor",
"authors": [
{
"first": "Anil Kumar",
"middle": [],
"last": "Singh",
"suffix": ""
}
],
"year": 2008,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Anil Kumar Singh. 2008. A mechanism to provide language-encoding support and an nlp friendly ed- itor.",
"links": null
},
"BIBREF12": {
"ref_id": "b12",
"title": "Brat: a web-based tool for nlp-assisted text annotation",
"authors": [
{
"first": "Pontus",
"middle": [],
"last": "Stenetorp",
"suffix": ""
},
{
"first": "Sampo",
"middle": [],
"last": "Pyysalo",
"suffix": ""
},
{
"first": "Goran",
"middle": [],
"last": "Topi\u0107",
"suffix": ""
},
{
"first": "Tomoko",
"middle": [],
"last": "Ohta",
"suffix": ""
},
{
"first": "Sophia",
"middle": [],
"last": "Ananiadou",
"suffix": ""
},
{
"first": "Jun'ichi",
"middle": [],
"last": "Tsujii",
"suffix": ""
}
],
"year": 2012,
"venue": "",
"volume": "",
"issue": "",
"pages": "102--107",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Pontus Stenetorp, Sampo Pyysalo, Goran Topi\u0107, Tomoko Ohta, Sophia Ananiadou, and Jun'ichi Tsu- jii. 2012. Brat: a web-based tool for nlp-assisted text annotation. pages 102-107.",
"links": null
},
"BIBREF13": {
"ref_id": "b13",
"title": "Microservices. IEEE software",
"authors": [
{
"first": "Johannes",
"middle": [],
"last": "Th\u00f6nes",
"suffix": ""
}
],
"year": 2015,
"venue": "",
"volume": "32",
"issue": "",
"pages": "116--116",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Johannes Th\u00f6nes. 2015. Microservices. IEEE soft- ware, 32(1):116-116.",
"links": null
}
},
"ref_entries": {
"FIGREF0": {
"text": "Kunji : Architecture",
"num": null,
"uris": null,
"type_str": "figure"
},
"TABREF1": {
"text": "",
"num": null,
"html": null,
"content": "<table/>",
"type_str": "table"
},
"TABREF3": {
"text": "Resource Record Details.",
"num": null,
"html": null,
"content": "<table/>",
"type_str": "table"
},
"TABREF5": {
"text": "Data sets description-1",
"num": null,
"html": null,
"content": "<table><tr><td>Total words in dataset</td><td>5693</td></tr><tr><td>Total sentences in dataset</td><td>309</td></tr><tr><td>Total paras in dataset</td><td>219</td></tr><tr><td>Average no. of sentences in each data set</td><td>15.45</td></tr><tr><td>Average sentence length</td><td>18.34 tokens</td></tr><tr><td>Average paras in complete data set</td><td>11</td></tr><tr><td>Average para length</td><td>26 tokens</td></tr></table>",
"type_str": "table"
},
"TABREF6": {
"text": "Data sets description-2",
"num": null,
"html": null,
"content": "<table><tr><td>Scenarios Description</td></tr><tr><td>Scenario-1 Manual Translation</td></tr><tr><td>Scenario-2 Post-editing in GT on Text editor</td></tr><tr><td>Scenario-3 GT with Transzaar without Kunji</td></tr><tr><td>Scenario-4 GT with Transzaar and Kunji</td></tr><tr><td>(onlyTM, Terms, Glossaries)</td></tr><tr><td>Scenario-5 GT with Transzaar and Kunji</td></tr><tr><td>(only linguistic resources)</td></tr><tr><td>Scenario-6 GT with Transzaar and Kunji</td></tr><tr><td>(TM, Terms, Glossaries and linguistic resources)</td></tr></table>",
"type_str": "table"
},
"TABREF7": {
"text": "Description of Various scenarios : Here GT refers to Google translation output",
"num": null,
"html": null,
"content": "<table/>",
"type_str": "table"
},
"TABREF9": {
"text": "Results of Experiment",
"num": null,
"html": null,
"content": "<table><tr><td>Scenarios</td><td>Total Time</td><td>Total Time</td><td>Total Time</td><td>Total Time</td></tr><tr><td/><td>taken</td><td>taken</td><td>taken</td><td>taken</td></tr><tr><td/><td>by U1</td><td>by U2</td><td>by U3</td><td>by U4</td></tr><tr><td/><td>per sent.</td><td>per sent.</td><td>per sent.</td><td>per sent.</td></tr><tr><td/><td>(in min.)</td><td>(in min.)</td><td>(in min.)</td><td>(in min.)</td></tr><tr><td>Scenario-1</td><td/><td/><td/><td/></tr></table>",
"type_str": "table"
}
}
}
} |