File size: 88,654 Bytes
6fa4bc9 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 | {
"paper_id": "P06-1046",
"header": {
"generated_with": "S2ORC 1.0.0",
"date_generated": "2023-01-19T09:25:10.121927Z"
},
"title": "Scaling Distributional Similarity to Large Corpora",
"authors": [
{
"first": "James",
"middle": [],
"last": "Gorman",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "University of Sydney NSW 2006",
"location": {
"country": "Australia"
}
},
"email": "jgorman2@it.usyd.edu.au"
},
{
"first": "James",
"middle": [
"R"
],
"last": "Curran",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "University of Sydney NSW 2006",
"location": {
"country": "Australia"
}
},
"email": ""
}
],
"year": "",
"venue": null,
"identifiers": {},
"abstract": "Accurately representing synonymy using distributional similarity requires large volumes of data to reliably represent infrequent words. However, the na\u00efve nearestneighbour approach to comparing context vectors extracted from large corpora scales poorly (O(n 2) in the vocabulary size). In this paper, we compare several existing approaches to approximating the nearestneighbour search for distributional similarity. We investigate the trade-off between efficiency and accuracy, and find that SASH (Houle and Sakuma, 2005) provides the best balance.",
"pdf_parse": {
"paper_id": "P06-1046",
"_pdf_hash": "",
"abstract": [
{
"text": "Accurately representing synonymy using distributional similarity requires large volumes of data to reliably represent infrequent words. However, the na\u00efve nearestneighbour approach to comparing context vectors extracted from large corpora scales poorly (O(n 2) in the vocabulary size). In this paper, we compare several existing approaches to approximating the nearestneighbour search for distributional similarity. We investigate the trade-off between efficiency and accuracy, and find that SASH (Houle and Sakuma, 2005) provides the best balance.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Abstract",
"sec_num": null
}
],
"body_text": [
{
"text": "It is a general property of Machine Learning that increasing the volume of training data increases the accuracy of results. This is no more evident than in Natural Language Processing (NLP), where massive quantities of text are required to model rare language events. Despite the rapid increase in computational power available for NLP systems, the volume of raw data available still outweighs our ability to process it. Unsupervised learning, which does not require the expensive and timeconsuming human annotation of data, offers an opportunity to use this wealth of data. Curran and Moens (2002) show that synonymy extraction for lexical semantic resources using distributional similarity produces continuing gains in accuracy as the volume of input data increases.",
"cite_spans": [
{
"start": 575,
"end": 598,
"text": "Curran and Moens (2002)",
"ref_id": "BIBREF4"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Extracting synonymy relations using distributional similarity is based on the distributional hypothesis that similar words appear in similar contexts. Terms are described by collating informa-tion about their occurrence in a corpus into vectors. These context vectors are then compared for similarity. Existing approaches differ primarily in their definition of \"context\", e.g. the surrounding words or the entire document, and their choice of distance metric for calculating similarity between the context vectors representing each term.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Manual creation of lexical semantic resources is open to the problems of bias, inconsistency and limited coverage. It is difficult to account for the needs of the many domains in which NLP techniques are now being applied and for the rapid change in language use. The assisted or automatic creation and maintenance of these resources would be of great advantage.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Finding synonyms using distributional similarity requires a nearest-neighbour search over the context vectors of each term. This is computationally intensive, scaling to O(n 2 m) for the number of terms n and the size of their context vectors m. Increasing the volume of input data will increase the size of both n and m, decreasing the efficiency of a na\u00efve nearest-neighbour approach.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Many approaches to reduce this complexity have been suggested. In this paper we evaluate state-of-the-art techniques proposed to solve this problem. We find that the Spatial Approximation Sample Hierarchy (Houle and Sakuma, 2005) provides the best accuracy/efficiency trade-off.",
"cite_spans": [
{
"start": 205,
"end": 229,
"text": "(Houle and Sakuma, 2005)",
"ref_id": "BIBREF10"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Introduction",
"sec_num": "1"
},
{
"text": "Measuring distributional similarity first requires the extraction of context information for each of the vocabulary terms from raw text. These terms are then compared for similarity using a nearestneighbour search or clustering based on distance calculations between the statistical descriptions of their contexts.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Distributional Similarity",
"sec_num": "2"
},
{
"text": "A context relation is defined as a tuple (w, r, w ) where w is a term, which occurs in some grammatical relation r with another word w in some sentence. We refer to the tuple (r, w ) as an attribute of w. For example, (dog, direct-obj, walk) indicates that dog was the direct object of walk in a sentence.",
"cite_spans": [
{
"start": 218,
"end": 241,
"text": "(dog, direct-obj, walk)",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Extraction",
"sec_num": "2.1"
},
{
"text": "In our experiments context extraction begins with a Maximum Entropy POS tagger and chunker. The SEXTANT relation extractor (Grefenstette, 1994) produces context relations that are then lemmatised. The relations for each term are collected together and counted, producing a vector of attributes and their frequencies in the corpus.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Extraction",
"sec_num": "2.1"
},
{
"text": "Both nearest-neighbour and cluster analysis methods require a distance measure to calculate the similarity between context vectors. Curran (2004) decomposes this into measure and weight functions. The measure calculates the similarity between two weighted context vectors and the weight calculates the informativeness of each context relation from the raw frequencies.",
"cite_spans": [
{
"start": 132,
"end": 145,
"text": "Curran (2004)",
"ref_id": "BIBREF5"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Measures and Weights",
"sec_num": "2.2"
},
{
"text": "For these experiments we use the Jaccard (1) measure and the TTest (2) weight functions, found by Curran (2004) to have the best performance.",
"cite_spans": [
{
"start": 98,
"end": 111,
"text": "Curran (2004)",
"ref_id": "BIBREF5"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Measures and Weights",
"sec_num": "2.2"
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "(r,w ) min(w(w m , r, w ), w(w n , r, w )) (r,w ) max(w(w m , r, w ), w(w n , r, w )) (1) p(w, r, w ) \u2212 p( * , r, w )p(w, * , * ) p( * , r, w )p(w, * , * )",
"eq_num": "(2)"
}
],
"section": "Measures and Weights",
"sec_num": "2.2"
},
{
"text": "The simplest algorithm for finding synonyms is a k-nearest-neighbour (k-NN) search, which involves pair-wise vector comparison of the target term with every term in the vocabulary. Given an n term vocabulary and up to m attributes for each term, the asymptotic time complexity of nearestneighbour search is O(n 2 m). This is very expensive, with even a moderate vocabulary making the use of huge datasets infeasible. Our largest experiments used a vocabulary of over 184,000 words.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Nearest-neighbour Search",
"sec_num": "2.3"
},
{
"text": "Using a cut-off to remove low frequency terms can significantly reduce the value of n. Unfortunately, reducing m by eliminating low frequency contexts has a significant impact on the quality of the results. There are many techniques to reduce dimensionality while avoiding this problem. The simplest methods use feature selection techniques, such as information gain, to remove the attributes that are less informative. Other techniques smooth the data while reducing dimensionality. Latent Semantic Analysis (LSA, Landauer and Dumais, 1997) is a smoothing and dimensionality reduction technique based on the intuition that the true dimensionality of data is latent in the surface dimensionality. Landauer and Dumais admit that, from a pragmatic perspective, the same effect as LSA can be generated by using large volumes of data with very long attribute vectors. Experiments with LSA typically use attribute vectors of a dimensionality of around 1000. Our experiments have a dimensionality of 500,000 to 1,500,000. Decompositions on data this size are computationally difficult. Dimensionality reduction is often used before using LSA to improve its scalability.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Dimensionality Reduction",
"sec_num": "3"
},
{
"text": "Another technique is to use an initial heuristic comparison to reduce the number of full O(m) vector comparisons that are performed. If the heuristic comparison is sufficiently fast and a sufficient number of full comparisons are avoided, the cost of an additional check will be easily absorbed by the savings made. Curran and Moens (2002) introduces a vector of canonical attributes (of bounded length k m), selected from the full vector, to represent the term. These attributes are the most strongly weighted verb attributes, chosen because they constrain the semantics of the term more and partake in fewer idiomatic collocations. If a pair of terms share at least one canonical attribute then a full similarity comparison is performed, otherwise the terms are not compared. They show an 89% reduction in search time, with only a 3.9% loss in accuracy.",
"cite_spans": [
{
"start": 316,
"end": 339,
"text": "Curran and Moens (2002)",
"ref_id": "BIBREF4"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Heuristics",
"sec_num": "3.1"
},
{
"text": "There is a significant improvement in the computational complexity. If a maximum of p positive results are returned, our complexity becomes O(n 2 k + npm). When p n, the system will be faster as many fewer full comparisons will be made, but at the cost of accuracy as more possibly near results will be discarded out of hand.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Heuristics",
"sec_num": "3.1"
},
{
"text": "Conventional dimensionality reduction techniques can be computationally expensive: a more scal-able solution is required to handle the volumes of data we propose to use. Randomised techniques provide a possible solution to this.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Randomised Techniques",
"sec_num": "4"
},
{
"text": "We present two techniques that have been used recently for distributional similarity: Random Indexing (Kanerva et al., 2000) and Locality Sensitive Hashing (LSH, Broder, 1997) .",
"cite_spans": [
{
"start": 102,
"end": 124,
"text": "(Kanerva et al., 2000)",
"ref_id": "BIBREF13"
},
{
"start": 162,
"end": 175,
"text": "Broder, 1997)",
"ref_id": "BIBREF0"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Randomised Techniques",
"sec_num": "4"
},
{
"text": "Random Indexing (RI) is a hashing technique based on Sparse Distributed Memory (Kanerva, 1993) . Karlgren and Sahlgren (2001) showed RI produces results similar to LSA using the Test of English as a Foreign Language (TOEFL) evaluation. Sahlgren and Karlgren (2005) showed the technique to be successful in generating bilingual lexicons from parallel corpora.",
"cite_spans": [
{
"start": 79,
"end": 94,
"text": "(Kanerva, 1993)",
"ref_id": "BIBREF14"
},
{
"start": 97,
"end": 125,
"text": "Karlgren and Sahlgren (2001)",
"ref_id": "BIBREF15"
},
{
"start": 236,
"end": 264,
"text": "Sahlgren and Karlgren (2005)",
"ref_id": "BIBREF19"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Random Indexing",
"sec_num": "4.1"
},
{
"text": "In RI, we first allocate a d length index vector to each unique attribute. The vectors consist of a large number of 0s and small number ( ) number of randomly distributed \u00b11s. Context vectors, identifying terms, are generated by summing the index vectors of the attributes for each non-unique context in which a term appears. The context vector for a term t appearing in contexts",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Random Indexing",
"sec_num": "4.1"
},
{
"text": "c 1 = [1, 0, 0, \u22121] and c 2 = [0, 1, 0, \u22121] would be [1, 1, 0, \u22122].",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Random Indexing",
"sec_num": "4.1"
},
{
"text": "The distance between these context vectors is then measured using the cosine measure:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Random Indexing",
"sec_num": "4.1"
},
{
"text": "cos(\u03b8(u, v)) = u \u2022 v | u| | v| (3)",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Random Indexing",
"sec_num": "4.1"
},
{
"text": "This technique allows for incremental sampling, where the index vector for an attribute is only generated when the attribute is encountered. Construction complexity is O(nmd) and search complexity is O(n 2 d).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Random Indexing",
"sec_num": "4.1"
},
{
"text": "LSH is a probabilistic technique that allows the approximation of a similarity function. Broder (1997) proposed an approximation of the Jaccard similarity function using min-wise independent functions. Charikar (2002) proposed an approximation of the cosine measure using random hyperplanes Ravichandran et al. (2005) used this cosine variant and showed it to produce over 70% accuracy in extracting synonyms when compared against Pantel and Lin (2002) .",
"cite_spans": [
{
"start": 89,
"end": 102,
"text": "Broder (1997)",
"ref_id": "BIBREF0"
},
{
"start": 202,
"end": 217,
"text": "Charikar (2002)",
"ref_id": "BIBREF2"
},
{
"start": 291,
"end": 317,
"text": "Ravichandran et al. (2005)",
"ref_id": "BIBREF18"
},
{
"start": 431,
"end": 452,
"text": "Pantel and Lin (2002)",
"ref_id": "BIBREF17"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "Given we have n terms in an m dimensional space, we create d m unit random vectors also of m dimensions, labelled { r 1 , r 2 , ..., r d }. Each vector is created by sampling a Gaussian function m times, with a mean of 0 and a variance of 1.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "For each term w we construct its bit signature using the function",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "h r ( w) = 1 : r. w \u2265 0 0 : r. w < 0",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "where r is a spherically symmetric random vector of length d. The signature,w, is the d length bit vector:w",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "= {h r 1 ( w), h r 2 ( w), . . . , h r d ( w)}",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "The cost to build all n signatures is O(nm d).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "For terms u and v, Goemans and Williamson (1995) approximate the angular similarity by",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "p(h r ( u) = h r ( v)) = 1 \u2212 \u03b8( u, u) \u03c0 (4)",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "where \u03b8( u, u) is the angle between u and u. The angular similarity gives the cosine by",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "cos(\u03b8( u, u)) = cos((1 \u2212 p(h r ( u) = h r ( v)))\u03c0)",
"eq_num": "(5)"
}
],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "The probability can be derived from the Hamming distance:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "p(h r (u) = h r (v)) = 1 \u2212 H(\u016b,v) d",
"eq_num": "(6)"
}
],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "By combining equations 5 and 6 we get the following approximation of the cosine distance:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "cos(\u03b8( u, u)) = cos H(\u016b,v) d \u03c0",
"eq_num": "(7)"
}
],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "That is, the cosine of two context vectors is approximated by the cosine of the Hamming distance between their two signatures normalised by the size of the signatures. Search is performed using Equation 7 and scales to O(n 2 d).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Locality Sensitive Hashing",
"sec_num": "4.2"
},
{
"text": "The methods presented above fail to address the n 2 component of the search complexity. Many data structures have been proposed that can be used to address this problem in similarity searching. We present three data structures: the vantage point tree (VPT, Yianilos, 1993) , which indexes points in a metric space, Point Location in Equal Balls (PLEB , Indyk and Motwani, 1998) , a probabilistic structure that uses the bit signatures generated by LSH, and the Spatial Approximation Sample Hierarchy (SASH, Houle and Sakuma, 2005) , which approximates a k-NN search. Another option inspired by IR is attribute indexing (INDEX). In this technique, in addition to each term having a reference to its attributes, each attribute has a reference to the terms referencing it. Each term is then only compared with the terms with which it shares attributes. We will give a theoretically comparison against other techniques.",
"cite_spans": [
{
"start": 251,
"end": 272,
"text": "(VPT, Yianilos, 1993)",
"ref_id": null
},
{
"start": 353,
"end": 377,
"text": "Indyk and Motwani, 1998)",
"ref_id": "BIBREF12"
},
{
"start": 507,
"end": 530,
"text": "Houle and Sakuma, 2005)",
"ref_id": "BIBREF10"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Data Structures",
"sec_num": "5"
},
{
"text": "Metric space data structures provide a solution to near-neighbour searches in very high dimensions. These rely solely on the existence of a comparison function that satisfies the conditions of metricality: non-negativity, equality, symmetry and the triangle inequality.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Vantage Point Tree",
"sec_num": "5.1"
},
{
"text": "VPT is typical of these structures and has been used successfully in many applications. The VPT is a binary tree designed for range searches. These are searches limited to some distance from the target term but can be modified for k-NN search.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Vantage Point Tree",
"sec_num": "5.1"
},
{
"text": "VPT is constructed recursively. Beginning with a set of U terms, we take any term to be our vantage point p. This becomes our root. We now find the median distance m p of all other terms to p: m p = median{dist(p, u)|u \u2208 U }. Those terms u such that dist(p, u) \u2264 m p are inserted into the left sub-tree, and the remainder into the right subtree. Each sub-tree is then constructed as a new VPT, choosing a new vantage point from within its terms, until all terms are exhausted.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Vantage Point Tree",
"sec_num": "5.1"
},
{
"text": "Searching a VPT is also recursive. Given a term q and radius r, we begin by measuring the distance to the root term p. If dist(q, p) \u2264 r we enter p into our list of near terms. If dist(q, p) \u2212 r \u2264 m p we enter the left sub-tree and if dist(q, p) + r > mp we enter the right sub-tree. Both sub-trees may be entered. The process is repeated for each entered subtree, taking the vantage point of the sub-tree to be the new root term.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Vantage Point Tree",
"sec_num": "5.1"
},
{
"text": "To perform a k-NN search we use a backtracking decreasing radius search (Burkhard and Keller, 1973) . The search begins with r = \u221e, and terms are added to a list of the closest k terms. When the k th closest term is found, the radius is set to the distance between this term and the target. Each time a new, closer element is added to the list, the radius is updated to the distance from the target to the new k th closest term.",
"cite_spans": [
{
"start": 72,
"end": 99,
"text": "(Burkhard and Keller, 1973)",
"ref_id": "BIBREF1"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Vantage Point Tree",
"sec_num": "5.1"
},
{
"text": "Construction complexity is O(n log n). Search complexity is claimed to be O(log n) for small radius searches. This does not hold for our decreasing radius search, whose worst case complexity is O(n).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Vantage Point Tree",
"sec_num": "5.1"
},
{
"text": "PLEB is a randomised structure that uses the bit signatures generated by LSH. It was used by Ravichandran et al. (2005) to improve the efficiency of distributional similarity calculations.",
"cite_spans": [
{
"start": 93,
"end": 119,
"text": "Ravichandran et al. (2005)",
"ref_id": "BIBREF18"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Point Location in Equal Balls",
"sec_num": "5.2"
},
{
"text": "Having generated our d length bit signatures for each of our n terms, we take these signatures and randomly permute the bits. Each vector has the same permutation applied. This is equivalent to a column reordering in a matrix where the rows are the terms and the columns the bits. After applying the permutation, the list of terms is sorted lexicographically based on the bit signatures. The list is scanned sequentially, and each term is compared to its B nearest neighbours in the list. The choice of B will effect the accuracy/efficiency trade-off, and need not be related to the choice of k. This is performed q times, using a different random permutation function each time. After each iteration, the current closest k terms are stored.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Point Location in Equal Balls",
"sec_num": "5.2"
},
{
"text": "For a fixed d, the complexity for the permutation step is O(qn), the sorting O(qn log n) and the search O(qBn).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Point Location in Equal Balls",
"sec_num": "5.2"
},
{
"text": "SASH approximates a k-NN search by precomputing some near neighbours for each node (terms in our case). This produces multiple paths between terms, allowing SASH to shape itself to the data set (Houle, 2003) . The following description is adapted from Houle and Sakuma (2005) .",
"cite_spans": [
{
"start": 194,
"end": 207,
"text": "(Houle, 2003)",
"ref_id": "BIBREF11"
},
{
"start": 252,
"end": 275,
"text": "Houle and Sakuma (2005)",
"ref_id": "BIBREF10"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "The SASH is a directed, edge-weighted graph with the following properties (see Figure 1 ):",
"cite_spans": [],
"ref_spans": [
{
"start": 79,
"end": 87,
"text": "Figure 1",
"ref_id": null
}
],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "\u2022 Each term corresponds to a unique node.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "\u2022 The nodes are arranged into a hierarchy of levels, with the bottom level containing n 2 nodes and the top containing a single root node. Each level, except the top, will contain half as many nodes as the level below.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "\u2022 Edges between nodes are linked to consecutive levels. Each node will have at most p parent nodes in the level above, and c child nodes in the level below. \u2022 Every node must have at least one parent so that all nodes are reachable from the root.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "Construction begins with the nodes being randomly distributed between the levels. SASH is then constructed iteratively by each node finding its closest p parents in the level above. The parent will keep the closest c of these children, forming edges in the graph, and reject the rest. Any nodes without parents after being rejected are then assigned as children of the nearest node in the previous level with fewer than c children.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "Searching is performed by finding the k nearest nodes at each level, which are added to a set of near nodes. To limit the search, only those nodes whose parents were found to be nearest at the previous level are searched. The k closest nodes from the set of near nodes are then returned. The search complexity is O(ck log n).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "In Figure 1 , the filled nodes demonstrate a search for the near-neighbours of some node q, using k = 2. Our search begins with the root node A. As we are using k = 2, we must find the two nearest children of A using our similarity measure. In this case, C and D are closer than B. We now find the closest two children of C and D. E is not checked as it is only a child of B. All other nodes are checked, including F and G, which are shared as children by B and C. From this level we chose G and H. The final levels are considered similarly.",
"cite_spans": [],
"ref_spans": [
{
"start": 3,
"end": 11,
"text": "Figure 1",
"ref_id": null
}
],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "At this point we now have the list of near nodes A, C, D, G, H, I, J, K and L. From this we chose the two nodes nearest q, H and I marked in black, which are then returned.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "k can be varied at each level to force a larger number of elements to be tested at the base of the SASH using, for instance, the equation:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "k i = max{ k 1\u2212 h\u2212i log n , 1 2 pc }",
"eq_num": "(8)"
}
],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "This changes our search complexity to:",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "k 1+ 1 log n k 1 log n \u22121 + pc 2 2 log n",
"eq_num": "(9)"
}
],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "We use this geometric function in our experiments. Gorman and Curran (2005a; 2005b) found the performance of SASH for distributional similarity could be improved by replacing the initial random ordering with a frequency based ordering. In accordance with Zipf's law, the majority of terms have low frequencies. Comparisons made with these low frequency terms are unreliable (Curran and Moens, 2002) . Creating SASH with high frequency terms near the root produces more reliable initial paths, but comparisons against these terms are more expensive.",
"cite_spans": [
{
"start": 51,
"end": 76,
"text": "Gorman and Curran (2005a;",
"ref_id": "BIBREF7"
},
{
"start": 77,
"end": 83,
"text": "2005b)",
"ref_id": "BIBREF8"
},
{
"start": 374,
"end": 398,
"text": "(Curran and Moens, 2002)",
"ref_id": "BIBREF4"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "The best accuracy/efficiency trade-off was found when using more reliable initial paths rather than the most reliable. This is done by folding the data around some mean number of relations. For each term, if its number of relations m i is greater than some chosen number of relations M, it is given a new ranking based on the score M 2 m i . Otherwise its ranking based on its number of relations. This has the effect of pushing very high and very low frequency terms away from the root.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Spatial Approximation Sample Hierarchy",
"sec_num": "5.3"
},
{
"text": "The simplest method for evaluation is the direct comparison of extracted synonyms with a manually created gold standard (Grefenstette, 1994) . To reduce the problem of limited coverage, our evaluation combines three electronic thesauri: the Macquarie, Roget's and Moby thesauri.",
"cite_spans": [
{
"start": 120,
"end": 140,
"text": "(Grefenstette, 1994)",
"ref_id": "BIBREF9"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Evaluation Measures",
"sec_num": "6"
},
{
"text": "We follow Curran (2004) and use two performance measures: direct matches (DIRECT) and inverse rank (INVR). DIRECT is the percentage of returned synonyms found in the gold standard. INVR is the sum of the inverse rank of each matching synonym, e.g. matches at ranks 3, 5 and 28 CORPUS CUT-OFF TERMS AVERAGE RELATIONS PER TERM BNC 0 246,067 43 5 88,926 116 100 14,862 617 LARGE 0 541,722 97 5 184,494 281 100 35,618 1,400 give an inverse rank score of 1 3 + 1 5 + 1 28 . With at most 100 matching synonyms, the maximum INVR is 5.187. This more fine grained as it incorporates the both the number of matches and their ranking. The same 300 single word nouns were used for evaluation as used by Curran (2004) for his large scale evaluation. These were chosen randomly from WordNet such that they covered a range over the following properties: frequency, number of senses, specificity and concreteness. For each of these terms, the closest 100 terms and their similarity scores were extracted.",
"cite_spans": [
{
"start": 10,
"end": 23,
"text": "Curran (2004)",
"ref_id": "BIBREF5"
},
{
"start": 712,
"end": 725,
"text": "Curran (2004)",
"ref_id": "BIBREF5"
}
],
"ref_spans": [
{
"start": 284,
"end": 427,
"text": "CUT-OFF TERMS AVERAGE RELATIONS PER TERM BNC 0 246,067 43 5 88,926 116 100 14,862 617 LARGE 0 541,722 97 5 184,494 281 100",
"ref_id": "TABREF0"
}
],
"eq_spans": [],
"section": "Evaluation Measures",
"sec_num": "6"
},
{
"text": "We use two corpora in our experiments: the smaller is the non-speech portion of the British National Corpus (BNC), 90 million words covering a wide range of domains and formats; the larger consists of the BNC, the Reuters Corpus Volume 1 and most of the English news holdings of the LDC in 2003, representing over 2 billion words of text (LARGE, Curran, 2004) . The semantic similarity system implemented by Curran (2004) provides our baseline. This performs a brute-force k-NN search (NAIVE). We present results for the canonical attribute heuristic (HEURISTIC), RI, LSH, PLEB, VPT and SASH.",
"cite_spans": [
{
"start": 346,
"end": 359,
"text": "Curran, 2004)",
"ref_id": "BIBREF5"
},
{
"start": 408,
"end": 421,
"text": "Curran (2004)",
"ref_id": "BIBREF5"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Experiments",
"sec_num": "7"
},
{
"text": "We take the optimal canonical attribute vector length of 30 for HEURISTIC from Curran (2004) . For SASH we take optimal values of p = 4 and c = 16 and use the folded ordering taking M = 1000 from Gorman and Curran (2005b) .",
"cite_spans": [
{
"start": 79,
"end": 92,
"text": "Curran (2004)",
"ref_id": "BIBREF5"
},
{
"start": 196,
"end": 221,
"text": "Gorman and Curran (2005b)",
"ref_id": "BIBREF8"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Experiments",
"sec_num": "7"
},
{
"text": "For RI, LSH and PLEB we found optimal values experimentally using the BNC. For LSH we chose d = 3, 000 (LSH 3,000 ) and 10, 000 (LSH 10,000 ), showing the effect of changing the dimensionality. The frequency statistics were weighted using mutual information, as in Ravichandran et al. (2005) :",
"cite_spans": [
{
"start": 265,
"end": 291,
"text": "Ravichandran et al. (2005)",
"ref_id": "BIBREF18"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Experiments",
"sec_num": "7"
},
{
"text": "EQUATION",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [
{
"start": 0,
"end": 8,
"text": "EQUATION",
"ref_id": "EQREF",
"raw_str": "log( p(w, r, w ) p(w, * , * )p( * , r, w ) )",
"eq_num": "(10)"
}
],
"section": "Experiments",
"sec_num": "7"
},
{
"text": "PLEB used the values q = 500 and B = 100. The initial experiments on RI produced quite poor results. The intuition was that this was caused by the lack of smoothing in the algorithm. Experiments were performed using the weights given in Curran (2004) . Of these, mutual information (10), evaluated with an extra log 2 (f (w, r, w ) + 1) factor and limited to positive values, produced the best results (RI MI ). The values d = 1000 and = 5 were found to produce the best results.",
"cite_spans": [
{
"start": 237,
"end": 250,
"text": "Curran (2004)",
"ref_id": "BIBREF5"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Experiments",
"sec_num": "7"
},
{
"text": "All experiments were performed on 3.2GHz Xeon P4 machines with 4GB of RAM.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Experiments",
"sec_num": "7"
},
{
"text": "As the accuracy of comparisons between terms increases with frequency (Curran, 2004) , applying a frequency cut-off will both reduce the size of the vocabulary (n) and increase the average accuracy of comparisons. Table 1 shows the reduction in vocabulary and increase in average context relations per term as cut-off increases. For LARGE, the initial 541,722 word vocabulary is reduced by 66% when a cut-off of 5 is applied and by 86% when the cut-off is increased to 100. The average number of relations increases from 97 to 1400.",
"cite_spans": [
{
"start": 70,
"end": 84,
"text": "(Curran, 2004)",
"ref_id": "BIBREF5"
}
],
"ref_spans": [
{
"start": 214,
"end": 221,
"text": "Table 1",
"ref_id": "TABREF0"
}
],
"eq_spans": [],
"section": "Results",
"sec_num": "8"
},
{
"text": "The work by Curran (2004) largely uses a frequency cut-off of 5. When this cut-off was used with the randomised techniques RI and LSH, it produced quite poor results. When the cut-off was increased to 100, as used by Ravichandran et al. (2005) , the results improved significantly. Table 2 shows the INVR scores for our various techniques using the BNC with cut-offs of 5 and 100. Table 3 shows the results of a full thesaurus extraction using the BNC and LARGE corpora using a cut-off of 100. The average DIRECT score and INVR are from the 300 test words. The total execution time is extrapolated from the average search time of these test words and includes the setup time. For LARGE, extraction using NAIVE takes 444 hours: over 18 days. If the 184,494 word vocabulary were used, it would take over 7000 hours, or nearly 300 days. This gives some indication of The only technique to become less accurate when the corpus size is increased is RI; it is likely that RI is sensitive to high frequency, low information contexts that are more prevalent in LARGE. Weighting reduces this effect, improving accuracy.",
"cite_spans": [
{
"start": 12,
"end": 25,
"text": "Curran (2004)",
"ref_id": "BIBREF5"
},
{
"start": 217,
"end": 243,
"text": "Ravichandran et al. (2005)",
"ref_id": "BIBREF18"
}
],
"ref_spans": [
{
"start": 282,
"end": 289,
"text": "Table 2",
"ref_id": "TABREF2"
},
{
"start": 381,
"end": 388,
"text": "Table 3",
"ref_id": "TABREF4"
}
],
"eq_spans": [],
"section": "Results",
"sec_num": "8"
},
{
"text": "The importance of the choice of d can be seen in the results for LSH. While much slower, LSH 10,000 is also much more accurate than LSH 3,000 , while still being much faster than NAIVE. Introducing the PLEB data structure does not improve the efficiency while incurring a small cost on accuracy. We are not using large enough datasets to show the improved time complexity using PLEB.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Results",
"sec_num": "8"
},
{
"text": "VPT is only slightly faster slightly faster than NAIVE. This is not surprising in light of the original design of the data structure: decreasing radius search does not guarantee search efficiency.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Results",
"sec_num": "8"
},
{
"text": "A significant influence in the speed of the randomised techniques, RI and LSH, is the fixed dimensionality. The randomised techniques use a fixed length vector which is not influenced by the size of m. The drawback of this is that the size of the vector needs to be tuned to the dataset.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Results",
"sec_num": "8"
},
{
"text": "It would seem at first glance that HEURIS-TIC and SASH provide very similar results, with HEURISTIC slightly slower, but more accurate. This misses the difference in time complexity between the methods: HEURISTIC is n 2 and SASH n log n. The improvement in execution time over NAIVE decreases as corpus size increases and this would be expected to continue. Further tuning of SASH parameters may improve its accuracy.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Results",
"sec_num": "8"
},
{
"text": "RI MI produces similar result using LARGE to SASH using BNC. This does not include the cost of extracting context relations from the raw text, so the true comparison is much worse. SASH allows the free use of weight and measure functions, but RI is constrained by having to transform any context space into a RI space. This is important when (Weeds and Weir, 2005) . RI also suffers n 2 complexity, where as SASH is n log n. Taking these into account, and that the improvements are barely significant, SASH is a better choice.",
"cite_spans": [
{
"start": 342,
"end": 364,
"text": "(Weeds and Weir, 2005)",
"ref_id": "BIBREF20"
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Results",
"sec_num": "8"
},
{
"text": "The results for LSH are disappointing. It performs consistently worse than the other methods except VPT. This could be improved by using larger bit vectors, but there is a limit to the size of these as they represent a significant memory overhead, particularly as the vocabulary increases. Table 4 presents the theoretical analysis of attribute indexing. The average number of comparisons made for various cut-offs of LARGE are shown. NAIVE and INDEX are the actual values for those techniques. The values for SASH are worst case, where the maximum number of terms are compared at each level. The actual number of comparisons made will be much less. The efficiency of INDEX is sensitive to the density of attributes and increasing the cut-off increases the density. This is seen in the dramatic drop in performance as the cut-off increases. This problem of density will increase as volume of raw input data increases, further reducing its effectiveness. SASH is only dependent on the number of terms, not the density.",
"cite_spans": [],
"ref_spans": [
{
"start": 290,
"end": 297,
"text": "Table 4",
"ref_id": "TABREF6"
}
],
"eq_spans": [],
"section": "Results",
"sec_num": "8"
},
{
"text": "Where the need for computationally efficiency out-weighs the need for accuracy, RI MI provides better results. SASH is the most balanced of the techniques tested and provides the most scalable, high quality results.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Results",
"sec_num": "8"
},
{
"text": "We have evaluated several state-of-the-art techniques for improving the efficiency of distributional similarity measurements. We found that, in terms of raw efficiency, Random Indexing (RI) was significantly faster than any other technique, but at the cost of accuracy. Even after our modifications to the RI algorithm to significantly improve its accuracy, SASH still provides a better accuracy/efficiency trade-off. This is more evident when considering the time to extract context information from the raw text. SASH, unlike RI, also allows us to choose both the weight and the measure used. LSH and PLEB could not match either the efficiency of RI or the accuracy of SASH.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusion",
"sec_num": "9"
},
{
"text": "We intend to use this knowledge to process even larger corpora to produce more accurate results.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusion",
"sec_num": "9"
},
{
"text": "Having set out to improve the efficiency of distributional similarity searches while limiting any loss in accuracy, we are producing full nearestneighbour searches 18 times faster, with only a 2% loss in accuracy.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Conclusion",
"sec_num": "9"
}
],
"back_matter": [
{
"text": "We would like to thank our reviewers for their helpful feedback and corrections. This work has been supported by the Australian Research Council under Discovery Project DP0453131.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Acknowledgements",
"sec_num": null
}
],
"bib_entries": {
"BIBREF0": {
"ref_id": "b0",
"title": "On the resemblance and containment of documents",
"authors": [
{
"first": "Andrei",
"middle": [],
"last": "Broder",
"suffix": ""
}
],
"year": 1997,
"venue": "Proceedings of the Compression and Complexity of Sequences",
"volume": "",
"issue": "",
"pages": "21--29",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Andrei Broder. 1997. On the resemblance and containment of documents. In Proceedings of the Compression and Complexity of Sequences, pages 21-29, Salerno, Italy.",
"links": null
},
"BIBREF1": {
"ref_id": "b1",
"title": "Some approaches to best-match file searching",
"authors": [
{
"first": "A",
"middle": [],
"last": "Walter",
"suffix": ""
},
{
"first": "Robert",
"middle": [
"M"
],
"last": "Burkhard",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Keller",
"suffix": ""
}
],
"year": 1973,
"venue": "Communications of the ACM",
"volume": "16",
"issue": "4",
"pages": "230--236",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Walter A. Burkhard and Robert M. Keller. 1973. Some ap- proaches to best-match file searching. Communications of the ACM, 16(4):230-236, April.",
"links": null
},
"BIBREF2": {
"ref_id": "b2",
"title": "Similarity estimation techniques from rounding algorithms",
"authors": [
{
"first": "Moses",
"middle": [
"S"
],
"last": "Charikar",
"suffix": ""
}
],
"year": 2002,
"venue": "Proceedings of the 34th",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Moses S. Charikar. 2002. Similarity estimation techniques from rounding algorithms. In Proceedings of the 34th",
"links": null
},
"BIBREF3": {
"ref_id": "b3",
"title": "Annual ACM Symposium on Theory of Computing",
"authors": [],
"year": null,
"venue": "",
"volume": "",
"issue": "",
"pages": "19--21",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Annual ACM Symposium on Theory of Computing, pages 380-388, Montreal, Quebec, Canada, 19-21 May.",
"links": null
},
"BIBREF4": {
"ref_id": "b4",
"title": "Improvements in automatic thesaurus extraction",
"authors": [
{
"first": "James",
"middle": [],
"last": "Curran",
"suffix": ""
},
{
"first": "Marc",
"middle": [],
"last": "Moens",
"suffix": ""
}
],
"year": 2002,
"venue": "Proceedings of the Workshop of the ACL Special Interest Group on the Lexicon",
"volume": "",
"issue": "",
"pages": "59--66",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "James Curran and Marc Moens. 2002. Improvements in au- tomatic thesaurus extraction. In Proceedings of the Work- shop of the ACL Special Interest Group on the Lexicon, pages 59-66, Philadelphia, PA, USA, 12 July.",
"links": null
},
"BIBREF5": {
"ref_id": "b5",
"title": "From Distributional to Semantic Similarity",
"authors": [
{
"first": "James",
"middle": [],
"last": "Curran",
"suffix": ""
}
],
"year": 2004,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "James Curran. 2004. From Distributional to Semantic Simi- larity. Ph.D. thesis, University of Edinburgh.",
"links": null
},
"BIBREF6": {
"ref_id": "b6",
"title": "Improved approximation algorithms for maximum cut and satisfiability problems using semidefinite programming",
"authors": [
{
"first": "X",
"middle": [],
"last": "Michel",
"suffix": ""
},
{
"first": "David",
"middle": [
"P"
],
"last": "Goemans",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Williamson",
"suffix": ""
}
],
"year": 1995,
"venue": "Journal of Association for Computing Machinery",
"volume": "42",
"issue": "6",
"pages": "1115--1145",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Michel X. Goemans and David P. Williamson. 1995. Improved approximation algorithms for maximum cut and satisfiability problems using semidefinite program- ming. Journal of Association for Computing Machinery, 42(6):1115-1145, November.",
"links": null
},
"BIBREF7": {
"ref_id": "b7",
"title": "Approximate searching for distributional similarity",
"authors": [
{
"first": "James",
"middle": [],
"last": "Gorman",
"suffix": ""
},
{
"first": "James",
"middle": [],
"last": "Curran",
"suffix": ""
}
],
"year": 2005,
"venue": "ACL-SIGLEX 2005 Workshop on Deep Lexical Acquisition",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "James Gorman and James Curran. 2005a. Approximate searching for distributional similarity. In ACL-SIGLEX 2005 Workshop on Deep Lexical Acquisition, Ann Arbor, MI, USA, 30 June.",
"links": null
},
"BIBREF8": {
"ref_id": "b8",
"title": "Augmenting approximate similarity searching with lexical information",
"authors": [
{
"first": "James",
"middle": [],
"last": "Gorman",
"suffix": ""
},
{
"first": "James",
"middle": [],
"last": "Curran",
"suffix": ""
}
],
"year": 2005,
"venue": "Australasian Language Technology Workshop",
"volume": "",
"issue": "",
"pages": "9--11",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "James Gorman and James Curran. 2005b. Augmenting ap- proximate similarity searching with lexical information. In Australasian Language Technology Workshop, Sydney, Australia, 9-11 November.",
"links": null
},
"BIBREF9": {
"ref_id": "b9",
"title": "Explorations in Automatic Thesaurus Discovery",
"authors": [
{
"first": "Gregory",
"middle": [],
"last": "Grefenstette",
"suffix": ""
}
],
"year": 1994,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Gregory Grefenstette. 1994. Explorations in Automatic The- saurus Discovery. Kluwer Academic Publishers, Boston.",
"links": null
},
"BIBREF10": {
"ref_id": "b10",
"title": "Fast approximate similarity search in extremely high-dimensional data sets",
"authors": [
{
"first": "Michael",
"middle": [
"E"
],
"last": "Houle",
"suffix": ""
},
{
"first": "Jun",
"middle": [],
"last": "Sakuma",
"suffix": ""
}
],
"year": 2005,
"venue": "Proceedings of the 21st International Conference on Data Engineering",
"volume": "",
"issue": "",
"pages": "619--630",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Michael E. Houle and Jun Sakuma. 2005. Fast approximate similarity search in extremely high-dimensional data sets. In Proceedings of the 21st International Conference on Data Engineering, pages 619-630, Tokyo, Japan.",
"links": null
},
"BIBREF11": {
"ref_id": "b11",
"title": "Navigating massive data sets via local clustering",
"authors": [
{
"first": "Michael",
"middle": [
"E"
],
"last": "Houle",
"suffix": ""
}
],
"year": 2003,
"venue": "Proceedings of the 9th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining",
"volume": "",
"issue": "",
"pages": "547--552",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Michael E. Houle. 2003. Navigating massive data sets via local clustering. In Proceedings of the 9th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 547-552, Washington, DC, USA.",
"links": null
},
"BIBREF12": {
"ref_id": "b12",
"title": "Approximate nearest neighbors: towards removing the curse of dimensionality",
"authors": [
{
"first": "Piotr",
"middle": [],
"last": "Indyk",
"suffix": ""
},
{
"first": "Rajeev",
"middle": [],
"last": "Motwani",
"suffix": ""
}
],
"year": 1998,
"venue": "Proceedings of the 30th annual ACM Symposium on Theory of Computing",
"volume": "",
"issue": "",
"pages": "24--26",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Piotr Indyk and Rajeev Motwani. 1998. Approximate near- est neighbors: towards removing the curse of dimension- ality. In Proceedings of the 30th annual ACM Symposium on Theory of Computing, pages 604-613, New York, NY, USA, 24-26 May. ACM Press.",
"links": null
},
"BIBREF13": {
"ref_id": "b13",
"title": "Random indexing of text samples for latent semantic analysis",
"authors": [
{
"first": "Pentti",
"middle": [],
"last": "Kanerva",
"suffix": ""
},
{
"first": "Jan",
"middle": [],
"last": "Kristoferson",
"suffix": ""
},
{
"first": "Anders",
"middle": [],
"last": "Holst",
"suffix": ""
}
],
"year": 2000,
"venue": "Proceedings of the 22nd Annual Conference of the Cognitive Science Society",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Pentti Kanerva, Jan Kristoferson, and Anders Holst. 2000. Random indexing of text samples for latent semantic anal- ysis. In Proceedings of the 22nd Annual Conference of the Cognitive Science Society, page 1036, Mahwah, NJ, USA.",
"links": null
},
"BIBREF14": {
"ref_id": "b14",
"title": "Sparse distributed memory and related models",
"authors": [
{
"first": "Pentti",
"middle": [],
"last": "Kanerva",
"suffix": ""
}
],
"year": 1993,
"venue": "Associative Neural Memories: Theory and Implementation",
"volume": "",
"issue": "",
"pages": "50--76",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Pentti Kanerva. 1993. Sparse distributed memory and re- lated models. In M.H. Hassoun, editor, Associative Neu- ral Memories: Theory and Implementation, pages 50-76. Oxford University Press, New York, NY, USA.",
"links": null
},
"BIBREF15": {
"ref_id": "b15",
"title": "From words to understanding",
"authors": [
{
"first": "Jussi",
"middle": [],
"last": "Karlgren",
"suffix": ""
},
{
"first": "Magnus",
"middle": [],
"last": "Sahlgren",
"suffix": ""
}
],
"year": 2001,
"venue": "Foundations of Real-World Intelligence",
"volume": "",
"issue": "",
"pages": "294--308",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Jussi Karlgren and Magnus Sahlgren. 2001. From words to understanding. In Y. Uesaka, P. Kanerva, and H Asoh, ed- itors, Foundations of Real-World Intelligence, pages 294- 308. CSLI Publications, Stanford, CA, USA.",
"links": null
},
"BIBREF16": {
"ref_id": "b16",
"title": "A solution to plato's problem: The latent semantic analysis theory of acquisition, induction, and representation of knowledge",
"authors": [
{
"first": "K",
"middle": [],
"last": "Thomas",
"suffix": ""
},
{
"first": "Susan",
"middle": [
"T"
],
"last": "Landauer",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Dumais",
"suffix": ""
}
],
"year": 1997,
"venue": "Psychological Review",
"volume": "104",
"issue": "2",
"pages": "211--240",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Thomas K. Landauer and Susan T. Dumais. 1997. A solution to plato's problem: The latent semantic analysis theory of acquisition, induction, and representation of knowledge. Psychological Review, 104(2):211-240, April.",
"links": null
},
"BIBREF17": {
"ref_id": "b17",
"title": "Discovering word senses from text",
"authors": [
{
"first": "Patrick",
"middle": [],
"last": "Pantel",
"suffix": ""
},
{
"first": "Dekang",
"middle": [],
"last": "Lin",
"suffix": ""
}
],
"year": 2002,
"venue": "Proceedings of ACM SIGKDD-02",
"volume": "",
"issue": "",
"pages": "23--26",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Patrick Pantel and Dekang Lin. 2002. Discovering word senses from text. In Proceedings of ACM SIGKDD-02, pages 613-619, 23-26 July.",
"links": null
},
"BIBREF18": {
"ref_id": "b18",
"title": "Randomized algorithms and NLP: Using locality sensitive hash functions for high speed noun clustering",
"authors": [
{
"first": "Deepak",
"middle": [],
"last": "Ravichandran",
"suffix": ""
},
{
"first": "Patrick",
"middle": [],
"last": "Pantel",
"suffix": ""
},
{
"first": "Eduard",
"middle": [],
"last": "Hovy",
"suffix": ""
}
],
"year": 2005,
"venue": "Proceedings of the 43rd Annual Meeting of the ACL",
"volume": "",
"issue": "",
"pages": "622--629",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Deepak Ravichandran, Patrick Pantel, and Eduard Hovy. 2005. Randomized algorithms and NLP: Using locality sensitive hash functions for high speed noun clustering. In Proceedings of the 43rd Annual Meeting of the ACL, pages 622-629, Ann Arbor, USA.",
"links": null
},
"BIBREF19": {
"ref_id": "b19",
"title": "Automatic bilingual lexicon acquisition using random indexing of parallel corpora",
"authors": [
{
"first": "Mangus",
"middle": [],
"last": "Sahlgren",
"suffix": ""
},
{
"first": "Jussi",
"middle": [],
"last": "Karlgren",
"suffix": ""
}
],
"year": 2005,
"venue": "Journal of Natural Language Engineering, Special Issue on Parallel Texts",
"volume": "11",
"issue": "3",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Mangus Sahlgren and Jussi Karlgren. 2005. Automatic bilin- gual lexicon acquisition using random indexing of parallel corpora. Journal of Natural Language Engineering, Spe- cial Issue on Parallel Texts, 11(3), June.",
"links": null
},
"BIBREF20": {
"ref_id": "b20",
"title": "Co-occurance retrieval: A flexible framework for lexical distributional similarity",
"authors": [
{
"first": "Julie",
"middle": [],
"last": "Weeds",
"suffix": ""
},
{
"first": "David",
"middle": [],
"last": "Weir",
"suffix": ""
}
],
"year": 2005,
"venue": "Computational Linguistics",
"volume": "31",
"issue": "4",
"pages": "439--475",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Julie Weeds and David Weir. 2005. Co-occurance retrieval: A flexible framework for lexical distributional similarity. Computational Linguistics, 31(4):439-475, December.",
"links": null
},
"BIBREF21": {
"ref_id": "b21",
"title": "Data structures and algorithms for nearest neighbor search in general metric spaces",
"authors": [
{
"first": "N",
"middle": [],
"last": "Peter",
"suffix": ""
},
{
"first": "",
"middle": [],
"last": "Yianilos",
"suffix": ""
}
],
"year": 1993,
"venue": "Proceedings of the fourth annual ACM-SIAM Symposium on Discrete algorithms",
"volume": "",
"issue": "",
"pages": "311--321",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Peter N. Yianilos. 1993. Data structures and algorithms for nearest neighbor search in general metric spaces. In Pro- ceedings of the fourth annual ACM-SIAM Symposium on Discrete algorithms, pages 311-321, Philadelphia.",
"links": null
}
},
"ref_entries": {
"FIGREF0": {
"uris": null,
"text": "A SASH, where p = 2, c = 3 and k = 2",
"num": null,
"type_str": "figure"
},
"TABREF0": {
"type_str": "table",
"num": null,
"html": null,
"content": "<table/>",
"text": "Extracted Context Information"
},
"TABREF2": {
"type_str": "table",
"num": null,
"html": null,
"content": "<table/>",
"text": "INVR vs frequency cut-off"
},
"TABREF4": {
"type_str": "table",
"num": null,
"html": null,
"content": "<table/>",
"text": "Full thesaurus extraction the scale of the problem."
},
"TABREF6": {
"type_str": "table",
"num": null,
"html": null,
"content": "<table><tr><td>: Average number of comparisons per term</td></tr><tr><td>considering that different tasks may require differ-</td></tr><tr><td>ent weights and measures</td></tr></table>",
"text": ""
}
}
}
} |