File size: 68,681 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
{
    "paper_id": "O08-4001",
    "header": {
        "generated_with": "S2ORC 1.0.0",
        "date_generated": "2023-01-19T08:02:36.113820Z"
    },
    "title": "Multiple Document Summarization Using Principal Component Analysis Incorporating Semantic Vector Space Model",
    "authors": [
        {
            "first": "Om",
            "middle": [],
            "last": "Vikas",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Akhil",
            "middle": [
                "K"
            ],
            "last": "Meshram",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Girraj",
            "middle": [],
            "last": "Meena",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Amit",
            "middle": [],
            "last": "Gupta",
            "suffix": "",
            "affiliation": {},
            "email": ""
        }
    ],
    "year": "",
    "venue": null,
    "identifiers": {},
    "abstract": "Text Summarization is very effective in relevant assessment tasks. The Multiple Document Summarizer presents a novel approach to select sentences from documents according to several heuristic features. Summaries are generated modeling the set of documents as Semantic Vector Space Model (SVSM) and applying Principal Component Analysis (PCA) to extract topic features. Pure Statistical VSM assumes terms to be independent of each other and may result in inconsistent results. Vector space is enhanced semantically by modifying the weight of the word vector governed by Appearance and Disappearance (Action class) words. The knowledge base for Action words is maintained by classifying the words as Appearance or Disappearance with the help of Wordnet. The weights of the action words are modified in accordance with the Object list prepared by the collection of nouns corresponding to the action words. Summary thus generated provides more informative content as semantics of natural language has been taken into consideration.",
    "pdf_parse": {
        "paper_id": "O08-4001",
        "_pdf_hash": "",
        "abstract": [
            {
                "text": "Text Summarization is very effective in relevant assessment tasks. The Multiple Document Summarizer presents a novel approach to select sentences from documents according to several heuristic features. Summaries are generated modeling the set of documents as Semantic Vector Space Model (SVSM) and applying Principal Component Analysis (PCA) to extract topic features. Pure Statistical VSM assumes terms to be independent of each other and may result in inconsistent results. Vector space is enhanced semantically by modifying the weight of the word vector governed by Appearance and Disappearance (Action class) words. The knowledge base for Action words is maintained by classifying the words as Appearance or Disappearance with the help of Wordnet. The weights of the action words are modified in accordance with the Object list prepared by the collection of nouns corresponding to the action words. Summary thus generated provides more informative content as semantics of natural language has been taken into consideration.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Abstract",
                "sec_num": null
            }
        ],
        "body_text": [
            {
                "text": "With the advent of the information revolution, electronic documents are becoming a principal media of business and academic information. The Internet is being populated with hundreds of thousands of electronic documents each day. In order to fully utilize these on-line documents effectively, it is crucial to be able to extract the main idea of these documents. Having a Text Summarization system would thus be immensely useful in serving this need. Multiple Document Summarization System aids to provide the summary of a document set that \uff0a Indian Institute of Information Technology and Management, Gwalior, India-474010 E-mail: {omvikas, akhil, girrajmeena, amitgupta}@iiitm.ac.in contains documents which belong to same topic. It can also be used to generate the summary of a single document.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1."
            },
            {
                "text": "In the present work, we propose a method of text summarization that uses semantics of data in order to form efficient and relevant summary. Summary is generated by constructing Statistical Vector Space Model (3.1) and then modifying it using the concept of Action words to form Semantic Vector Space Model (3.2). Action Words are identified using the Action Word Classifier which makes use of Wordnet [Kedar et al.] in order to analyze the semantics of word.",
                "cite_spans": [
                    {
                        "start": 401,
                        "end": 415,
                        "text": "[Kedar et al.]",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1."
            },
            {
                "text": "Principal Component Analysis (3.3) is then applied on SVSM to reduce the dimension of multidimensional data sets. Singular Value Decomposition (SVD) is carried out on SVSM as a part of PCA to yield singular values and eigen vectors. Backprojection is then performed to project the documents onto the eigen space yielding projected values of documents which are henceforth compared with the singular values to yield the most relevant document/topic. Sentence Extraction (3.4) from multiple document sets has been assigned weight on the basis of keywords obtained from the most important document/topic. Sentences with higher weight are taken to form a summary.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1."
            },
            {
                "text": "Various multiple document summarization systems already exist. This document summarizer is based on Kupeic 95 [Kupeic et al. 1995] which is a method of training a Bayesian classifier to recognize sentences that should belong in a summary. The classifier estimates the probability that a sentence belongs in a summary given a vector of features that are computed over the sentence. It identifies a set of features that correspond to the absence/presence of certain words or phrases and avoids the problem of having to analyze sentence structure. Their work focused on analyzing a single document at a time. Since then, there has been lot of work on the related problem of Multiple-document Summarization [Regina et al. 1999; Radev et al. 1998 ], where a system summarizes multiple documents on the same topic. For example, a system might summarize multiple news accounts of the recent massacre in Nepal; into a single document. Our hypothesis is that the similarities and differences between documents of the same type (e.g. bios of CS professors, earnings releases, etc.) provide information about the features that make a summary informative. The intuition is that the 'information content' of a document can be measured by the relationship between the document and a corpus of related documents. To be an informative summary, an abstract has to capture as much of the 'information content' as possible. To gain a handle on the problem of capturing the relationship between a document and a corpus, we examined several papers on Multiple-Document Summarization [Regina et al. 1999; Radev et al. 1998 Radev et al. , 2000 Radev et al. , 2004 Otterbacher et al. 2002] . However, we found most of their approaches were not applicable to Multiple Document Summarization Using 143 Principal Component Analysis Incorporating Semantic Vector Space Model our problem since they are mostly trying to match sentences of the same meaning to align multiple documents. The MEAD summarizer [Radev et al. 2000 [Radev et al. , 2001 , which was developed at the University of Michigan and at the Johns Hopkins University 2001 Summer Workshop on Automatic Summarization, produces summaries of one or more source articles (or a 'cluster' of topically related articles).",
                "cite_spans": [
                    {
                        "start": 110,
                        "end": 130,
                        "text": "[Kupeic et al. 1995]",
                        "ref_id": null
                    },
                    {
                        "start": 671,
                        "end": 723,
                        "text": "Multiple-document Summarization [Regina et al. 1999;",
                        "ref_id": null
                    },
                    {
                        "start": 724,
                        "end": 741,
                        "text": "Radev et al. 1998",
                        "ref_id": "BIBREF11"
                    },
                    {
                        "start": 1562,
                        "end": 1582,
                        "text": "[Regina et al. 1999;",
                        "ref_id": null
                    },
                    {
                        "start": 1583,
                        "end": 1600,
                        "text": "Radev et al. 1998",
                        "ref_id": "BIBREF11"
                    },
                    {
                        "start": 1601,
                        "end": 1620,
                        "text": "Radev et al. , 2000",
                        "ref_id": "BIBREF14"
                    },
                    {
                        "start": 1621,
                        "end": 1640,
                        "text": "Radev et al. , 2004",
                        "ref_id": "BIBREF16"
                    },
                    {
                        "start": 1641,
                        "end": 1665,
                        "text": "Otterbacher et al. 2002]",
                        "ref_id": "BIBREF9"
                    },
                    {
                        "start": 1976,
                        "end": 1994,
                        "text": "[Radev et al. 2000",
                        "ref_id": "BIBREF14"
                    },
                    {
                        "start": 1995,
                        "end": 2015,
                        "text": "[Radev et al. , 2001",
                        "ref_id": "BIBREF15"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Work",
                "sec_num": "2."
            },
            {
                "text": "Our Summarizer works on the documents belonging to same topic. It is strongly motivated by the analogy between this problem and the problem of face identification, where a system learns features for facial identification by applying PCA to find the characteristic eigenfaces [Turk et al. 1991; Pentland et al. 1994; Moon et al. 2001] .",
                "cite_spans": [
                    {
                        "start": 275,
                        "end": 293,
                        "text": "[Turk et al. 1991;",
                        "ref_id": "BIBREF19"
                    },
                    {
                        "start": 294,
                        "end": 315,
                        "text": "Pentland et al. 1994;",
                        "ref_id": "BIBREF10"
                    },
                    {
                        "start": 316,
                        "end": 333,
                        "text": "Moon et al. 2001]",
                        "ref_id": "BIBREF8"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Work",
                "sec_num": "2."
            },
            {
                "text": "Any set of documents dealing with the same subject is decomposed using Vector Space model. The important keywords can be extracted from the Vector Space Model using a threshold. Such keywords are called thematic keywords which are based on statistics. Important sentences can be extracted and a summary can be made using thematic keywords. We propose a new methodology for multiple document summarization by enhancing the VSM using semantics and identifying topic features based keywords to make the multiple document summary. The approach is:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "New Methodology",
                "sec_num": "3."
            },
            {
                "text": "1. Statistical VSM construction from Multiple Document Set.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "New Methodology",
                "sec_num": "3."
            },
            {
                "text": "Wordnet.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Semantic VSM generation using the concept of Contextual Action Words using",
                "sec_num": "2."
            },
            {
                "text": "3. Application of PCA on Semantic VSM to reduce the dimension of the multidimensional data set yielding the most important Keywords.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Semantic VSM generation using the concept of Contextual Action Words using",
                "sec_num": "2."
            },
            {
                "text": "4. Score Sentences based on several features such as sentence length cut-off feature, position feature, keyword weight, etc.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Semantic VSM generation using the concept of Contextual Action Words using",
                "sec_num": "2."
            },
            {
                "text": "5. Generation of Summary extracting the sentences with high Score.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Semantic VSM generation using the concept of Contextual Action Words using",
                "sec_num": "2."
            },
            {
                "text": "The Multiple Document Summarizer models the set of documents related to the same topic as the Statistical Vector Space Model based on several heuristics. The simplest way to transform a document into a vector is to define each unique word as a feature. The weight of a feature being decided based on the contribution of various parameters such as Cue-phrase Keywords, topic keywords and term frequency in document. The weight of the feature is being termed as Feature Combination.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Statistical VSM Construction",
                "sec_num": "3.1"
            },
            {
                "text": "The vector representations of the documents; collectively define an n-dimensional vector space (where each document is an nx1 vector). The m document vectors taken as the columns of an nxm matrix D, define a linear transformation into the vector space.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Statistical VSM Construction",
                "sec_num": "3.1"
            },
            {
                "text": "The existing vector space model is statistical in nature. This vector space is input to a number of tools and processes like a summarizer and information retrieval system. PCA/SVD Technique has been applied earlier for Summarization based on statistical vector space [Gong et al. 2001] . Some times this statistically generated model is unable to define the context.",
                "cite_spans": [
                    {
                        "start": 267,
                        "end": 285,
                        "text": "[Gong et al. 2001]",
                        "ref_id": "BIBREF4"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Semantic VSM Construction",
                "sec_num": "3.2"
            },
            {
                "text": "Keywords identified by a statistical model can be non-contextual in nature. Therefore, an effort is to be made in the direction of identification of contextual keywords and modification of existing model so that it can be more helpful and contextual for various applications like text summarization and text retrieval.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Semantic VSM Construction",
                "sec_num": "3.2"
            },
            {
                "text": "To identify the contextual keywords, we try to exploit human psychology. In any article, we identify that those words are important which either give a sense of either appearance or disappearance of any object/event. Thus, after we have the pure statistical vector space we need to enhance the vector space semantically by modifying the weights of the word vector by identifying the Appearance and Disappearance (ACTION class) words. To do so, we need to have a knowledgebase (KB) with some seed wordlist which belongs to appearance or disappearance. Following, are the steps involved in the semantic vector space model. 1. Get the tf matrix, T from existing document D.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Semantic VSM Construction",
                "sec_num": "3.2"
            },
            {
                "text": "2. Identify the set of action words, A from the given tf matrix, T, (number of action words =n).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Semantic VSM Construction",
                "sec_num": "3.2"
            },
            {
                "text": "A i ; A i \u03b5 A, 0<i<n.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Find the associated object list O i for action word",
                "sec_num": "3."
            },
            {
                "text": "4. Find contextual objects Co from Object list O1,O2,\u2026.,On.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Find the associated object list O i for action word",
                "sec_num": "3."
            },
            {
                "text": "5. Modify weight of contextual objects in T to form semantic vector space S{T}.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Find the associated object list O i for action word",
                "sec_num": "3."
            },
            {
                "text": "Action words are the backbone of the semantic vector space model.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Identification of Action words",
                "sec_num": "3.2.1"
            },
            {
                "text": "Definition: Action words are verbs that are used to strengthen the way experiences are presented whether it is expressing positive or negative experience.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Identification of Action words",
                "sec_num": "3.2.1"
            },
            {
                "text": "With the help of Wordnet, the terms from the tf (term frequency) matrix which belong to the ACTION class can be easily classified. The algorithm uses a seed word list to identify the action words.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Identification of Action words",
                "sec_num": "3.2.1"
            },
            {
                "text": "Whenever a term from the tf matrix is fetched, it is matched against seed word list. If it is matched, then the fetched term is action word; otherwise, synonyms of the fetched terms are matched.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Multiple Document Summarization Using 145 Principal Component Analysis Incorporating Semantic Vector Space Model Definition: Seed Word List is the collection of action words. (Appendix A)",
                "sec_num": null
            },
            {
                "text": "Given Input: T = {t1,t2,\u2026..,tn}. List type: A = { }, integer type: depth Do: for every t \u03b5 T depth = 0; match(t,seedwrdlst) if found then A = A U t. else, not found if(depth == 0) match(extractsynonym (t),seedwordlist) depth = 1 else continue endif endif endfor Output : A = {t1,t2,\u2026.tm}.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Multiple Document Summarization Using 145 Principal Component Analysis Incorporating Semantic Vector Space Model Definition: Seed Word List is the collection of action words. (Appendix A)",
                "sec_num": null
            },
            {
                "text": "To decide whether the word belongs to action list or not, we have to build a seed wordlist and compare them with standard meaning. For example, let 'devastation' be the word to be decided as action or not. After searching in WordNet, the following meanings were obtained:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Figure 1. ACTION Word Classifier",
                "sec_num": null
            },
            {
                "text": "\u2022 desolation (an event that results in total destruction) \u2022 destruction, (the termination of something by causing so much damage to it that it cannot be repaired or no longer exists)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Figure 1. ACTION Word Classifier",
                "sec_num": null
            },
            {
                "text": "From the first and last meaning, it clearly lies in the phenomenon of appear/disappear so it will be appended into the seed list along with its Synset.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Figure 1. ACTION Word Classifier",
                "sec_num": null
            },
            {
                "text": "Merely acquiring the ACTION words doesn't provide the semantic to the vector space. We have to find whether these words are really important. The importance of the word can be estimated by the application of the word in the article. Objects corresponding to the Action Words and their weight in Statistical VSM have to be identified in order to determine the extent of relevancy of Action Words. The Objects are the Nouns or Adjectives for the Action. The nearest Noun for Verb is identified using POS Tagger and termed as Object of Action. Only those sentences are to be chosen which contain action words. ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Finding the Objects of the Action",
                "sec_num": "3.2.2"
            },
            {
                "text": "Contextual words are being defined as those action words which are applied to the important object. The Weight of Action Word is being taken as the maximum weight amongst all the objects corresponding to the given Action Word. The weight obtained is added to the weight of the corresponding Action Words in Statistical VSM yielding Semantic Vector Space Model for the given set of Documents.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Classification of Contextual Words",
                "sec_num": "3.2.3"
            },
            {
                "text": "If we take an example of a single document:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Classification of Contextual Words",
                "sec_num": "3.2.3"
            },
            {
                "text": "Today broke fire in Delhi. Mass Destruction of material happens due to this fire. Many suffered from the broken glass in the road. The authority arrives here soon. Till now there is report of any casualties in these fire except from few injures. Thanks for the local communities for help.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Classification of Contextual Words",
                "sec_num": "3.2.3"
            },
            {
                "text": "the Vector Space generated on the basis of term frequency feature is The Action Word List obtained corresponding to the above example is: broke, destruction, and arrives. Now, the Action-Object list is prepared by identifying the Object words in which the ACTION words are acted. The Statistical VSM is now modified and the Semantic VSM is being generated as follows ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Classification of Contextual Words",
                "sec_num": "3.2.3"
            },
            {
                "text": "Principal Component Analysis (PCA) [Michael et al. 2003 ] is used to reduce the multidimensional datasets to lower dimensions for analysis. Singular Value Decomposition (SVD) [Michael et al. 2003 ] is carried out on Semantic VSM to find the principal components of Vector Space. The singular value decomposition (SVD) of matrix A mxn is the factorization A=U\u2211V T , where U and V are orthogonal, and \u2211= diag (\u03c3 1,\u2026, \u03c3 r ), r= min (m,n), with \u03c3 1 \u2265 \u03c3 2 \u2265 \u2026\u2026\u2265 \u03c3 r \u22650 . The columns of V are the 'hidden' dimensions that we are looking for. The diagonal of \u2211 are the singular values which are the weights for the new set of basis vectors. \u2211 is symmetric, its singular values are its eigen values and its basis vectors are the eigen vectors.",
                "cite_spans": [
                    {
                        "start": 35,
                        "end": 55,
                        "text": "[Michael et al. 2003",
                        "ref_id": null
                    },
                    {
                        "start": 175,
                        "end": 195,
                        "text": "[Michael et al. 2003",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Principal Component Analysis",
                "sec_num": "3.3"
            },
            {
                "text": "Given an eigen vector e, we can find the corresponding dimension in document space.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Principal Component Analysis",
                "sec_num": "3.3"
            },
            {
                "text": "Multiple Document Summarization Using",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "d= D.e",
                "sec_num": null
            },
            {
                "text": "After determining out the dimension of eigen vector in document space, backprojection of d * is carried out. Commonly, composing a vector in terms of the principal components is called backprojection. Since our principal components or eigen documents are all orthogonal vectors, this is easy to accomplish. Let E be the matrix formed from the eigen documents then vector p is the document projected onto the eigenspace.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Principal Component Analysis Incorporating Semantic Vector Space Model",
                "sec_num": "149"
            },
            {
                "text": "p = E T d",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Principal Component Analysis Incorporating Semantic Vector Space Model",
                "sec_num": "149"
            },
            {
                "text": "Relevance of the topic/document is calculated by dividing projected component by the corresponding Singular Value. Metrics thus obtained is arranged in decreasing order excluding out the negative metrics. Main topic/Document is the one with highest metric value.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Principal Component Analysis Incorporating Semantic Vector Space Model",
                "sec_num": "149"
            },
            {
                "text": "After selecting the main topic, we now need the topic keywords. We simply take the eigen document vector corresponding to main document and select the words with high weight. These are the set of Keywords which are of high relevance in summary.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Principal Component Analysis Incorporating Semantic Vector Space Model",
                "sec_num": "149"
            },
            {
                "text": "To identify sentences that should belong to summary, several features have been taken into consideration.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Sentence Extraction",
                "sec_num": "3.4"
            },
            {
                "text": "\u2022 Sentence-Length Cut off Feature -If the sentence length is greater than 4 words, only then it is taken into consideration. \u2022 Position Feature -Sentences have been given some weight based on their position in the paragraph whether it is in initial, middle or final.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Sentence Extraction",
                "sec_num": "3.4"
            },
            {
                "text": "\u2022 Keywords -Sentence weight also depends not only on the number of keywords present in it but also the weight of each keyword.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Sentence Extraction",
                "sec_num": "3.4"
            },
            {
                "text": "\u2022 Upper Case Feature -Sentences containing upper case words have been given additional weight as it is probable that they may contain proper nouns.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Sentence Extraction",
                "sec_num": "3.4"
            },
            {
                "text": "Sentences with higher weight are taken as the relevant sentences for the summary and arranged in the order they appear in the document yielding the required summary. The rearrangement becomes a challenge in the case of multiple documents. In that case, sentences are kept at the position at which they appear in original document (initial/middle/final). This rearrangement technique provides fair results.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Sentence Extraction",
                "sec_num": "3.4"
            },
            {
                "text": "The Multiple Document Summarization System is implemented in Java using JAMA (Java Matrix Package) and WVTool (Word Vector Tool) packages. JAMA is used to perform all the matrix operations as computing SVD, eigen vector, Backprojection, etc. WVTool is used to generate the Statistical Vector Space Model taking input as the Multiple Document Set or a Single document based on user requirement.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Implementation",
                "sec_num": "4."
            },
            {
                "text": "The present section will focus on the accuracy of the proposed summarization method. The accuracy of the method was examined on both single as well as multiple document summaries:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Evaluation of Summarizer",
                "sec_num": "5."
            },
            {
                "text": "Text belonging to different areas was taken. Summaries to the same texts were made by sentence extractions by different people. Based on the set of the summaries, we ranked sentences of the texts.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Single Document summary",
                "sec_num": "5.1"
            },
            {
                "text": "We then carried out the summarization process using our algorithm, the Auto Summarizer in MS Word, and the Gnome Summarizer and compared their agreement on the extracted sentences with the human sentence extractions.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Single Document summary",
                "sec_num": "5.1"
            },
            {
                "text": "The results are given in the following table. ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Single Document summary",
                "sec_num": "5.1"
            },
            {
                "text": "On an average, we get an average accuracy of 61.85% and improvement of 39.17% with respect to MS Word Summarizer.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "44.44% 47%",
                "sec_num": null
            },
            {
                "text": "The set of Documents belonging to \"Introduction to Web crawler\" were taken and then summary was generated using the proposed algorithm, and it was observed that the summary thus generated was in coherence with most of the documents. The input documents set consisting of documents related to the topic for summarization has been shown in Table 7 . The cause of the low contribution of Doc4 to the summary generated was observed to be sentences with fewer keywords in them with respect to sentences from other documents, resulting in a low score of sentence.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 338,
                        "end": 345,
                        "text": "Table 7",
                        "ref_id": "TABREF4"
                    }
                ],
                "eq_spans": [],
                "section": "Multiple Documents Summary",
                "sec_num": "5.2"
            },
            {
                "text": "As seen from the results, the proposed method works better for various domains and, by using Semantic VSM instead of Statistical VSM; the summary obtained has become more informational and meaningful. Moreover, this method can be used to generate single as well as multiple document summaries.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Conclusion and Future Work",
                "sec_num": "6."
            },
            {
                "text": "The following areas in Multiple Document Summarization System require improvement: 1 Documents for Summary % age in sum m ary Doc1 Doc2 Doc3 Doc4 1. Rearrangement of Extracted Sentences in the case of Multiple Documents Summarization to form an effective summary.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Conclusion and Future Work",
                "sec_num": "6."
            },
            {
                "text": "2. Enhance Flexibility of the system to generate a summary of multiple documents not necessarily belonging to the same topic.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Conclusion and Future Work",
                "sec_num": "6."
            },
            {
                "text": "3. Develop better methodology to incorporate the ACTION word score into Statistical VSM.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Conclusion and Future Work",
                "sec_num": "6."
            },
            {
                "text": "4. Evaluation of the system on large data samples. ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Conclusion and Future Work",
                "sec_num": "6."
            }
        ],
        "back_matter": [],
        "bib_entries": {
            "BIBREF0": {
                "ref_id": "b0",
                "title": "Information fusion in the context of multi-document summarization",
                "authors": [
                    {
                        "first": "R",
                        "middle": [],
                        "last": "Barzilay",
                        "suffix": ""
                    }
                ],
                "year": 2003,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Barzilay, R., \"Information fusion in the context of multi-document summarization,\" Phd. Thesis, Columbia University, 2003.",
                "links": null
            },
            "BIBREF1": {
                "ref_id": "b1",
                "title": "Generic Text Summarization using WordNet",
                "authors": [
                    {
                        "first": "K",
                        "middle": [],
                        "last": "Bellare",
                        "suffix": ""
                    },
                    {
                        "first": "A",
                        "middle": [],
                        "last": "Sarma",
                        "suffix": ""
                    },
                    {
                        "first": "A",
                        "middle": [],
                        "last": "Sarma",
                        "suffix": ""
                    },
                    {
                        "first": "N",
                        "middle": [],
                        "last": "Loiwal",
                        "suffix": ""
                    },
                    {
                        "first": "V",
                        "middle": [],
                        "last": "Mehta",
                        "suffix": ""
                    },
                    {
                        "first": "G",
                        "middle": [],
                        "last": "Ramakrishnan",
                        "suffix": ""
                    },
                    {
                        "first": "P",
                        "middle": [],
                        "last": "Bhattacharya",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Bellare, K., A. Das Sarma, A. Das Sarma, N. Loiwal, V. Mehta, G. Ramakrishnan, and P. Bhattacharya, Generic Text Summarization using WordNet, http://i.stanford.edu/ ~anishds/publications/lrec04/lrec04.ps.",
                "links": null
            },
            "BIBREF2": {
                "ref_id": "b2",
                "title": "Using linear algebra for intelligent information-retrieval",
                "authors": [
                    {
                        "first": "M",
                        "middle": [
                            "W"
                        ],
                        "last": "Berry",
                        "suffix": ""
                    },
                    {
                        "first": "S",
                        "middle": [
                            "T"
                        ],
                        "last": "Dumais",
                        "suffix": ""
                    },
                    {
                        "first": "G",
                        "middle": [
                            "W"
                        ],
                        "last": "Obrien",
                        "suffix": ""
                    }
                ],
                "year": 1995,
                "venue": "Siam Review",
                "volume": "37",
                "issue": "",
                "pages": "573--95",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Berry, M. W., S. T. Dumais, and G. W. Obrien, \"Using linear algebra for intelligent information-retrieval,\" Siam Review, 37, 1995, pp. 573-95.",
                "links": null
            },
            "BIBREF4": {
                "ref_id": "b4",
                "title": "Generic Text Summarization Using Relevance Measure and Latent Semantic Analysis",
                "authors": [
                    {
                        "first": "Y",
                        "middle": [],
                        "last": "Gong",
                        "suffix": ""
                    },
                    {
                        "first": "X",
                        "middle": [],
                        "last": "Liu",
                        "suffix": ""
                    }
                ],
                "year": 2001,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "19--25",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Gong, Y., and X. Liu, \"Generic Text Summarization Using Relevance Measure and Latent Semantic Analysis,\" SIGIR 2001, pp. 19-25.",
                "links": null
            },
            "BIBREF5": {
                "ref_id": "b5",
                "title": "A parallel algorithm for computing the singular-value decomposition of a matrix",
                "authors": [
                    {
                        "first": "E",
                        "middle": [
                            "R"
                        ],
                        "last": "Jessup",
                        "suffix": ""
                    },
                    {
                        "first": "D",
                        "middle": [
                            "C"
                        ],
                        "last": "Sorensen",
                        "suffix": ""
                    }
                ],
                "year": 1994,
                "venue": "Siam Journal on Matrix Analysis and Applications",
                "volume": "15",
                "issue": "",
                "pages": "530--548",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Jessup, E. R., and D. C. Sorensen, \"A parallel algorithm for computing the singular-value decomposition of a matrix,\" Siam Journal on Matrix Analysis and Applications, 15, 1994, pp. 530-548.",
                "links": null
            },
            "BIBREF6": {
                "ref_id": "b6",
                "title": "Principal Component Analysis",
                "authors": [
                    {
                        "first": "I",
                        "middle": [
                            "T"
                        ],
                        "last": "Jolliffe",
                        "suffix": ""
                    }
                ],
                "year": 1986,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Jolliffe, I. T., Principal Component Analysis, New York: Springer, 1986.",
                "links": null
            },
            "BIBREF7": {
                "ref_id": "b7",
                "title": "A Trainable Document Summarizer",
                "authors": [
                    {
                        "first": "J",
                        "middle": [],
                        "last": "Kupiec",
                        "suffix": ""
                    },
                    {
                        "first": "J",
                        "middle": [],
                        "last": "Pedersen",
                        "suffix": ""
                    },
                    {
                        "first": "F",
                        "middle": [],
                        "last": "Chen",
                        "suffix": ""
                    }
                ],
                "year": 1995,
                "venue": "Proceedings of the 18th ACM-SIGIR Conference",
                "volume": "",
                "issue": "",
                "pages": "68--73",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Kupiec, J., J. Pedersen, and F. Chen, \"A Trainable Document Summarizer,\" In Proceedings of the 18th ACM-SIGIR Conference, 1995, pp. 68-73.",
                "links": null
            },
            "BIBREF8": {
                "ref_id": "b8",
                "title": "Computational and Performance aspects of PCA-based Face Recognition Algorithms",
                "authors": [
                    {
                        "first": "H",
                        "middle": [],
                        "last": "Moon",
                        "suffix": ""
                    },
                    {
                        "first": "P",
                        "middle": [
                            "J"
                        ],
                        "last": "Phillips",
                        "suffix": ""
                    }
                ],
                "year": 2001,
                "venue": "Perception",
                "volume": "30",
                "issue": "",
                "pages": "303--321",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Moon, H., and P. J. Phillips, \"Computational and Performance aspects of PCA-based Face Recognition Algorithms,\" Perception, 30, 2001, pp. 303-321.",
                "links": null
            },
            "BIBREF9": {
                "ref_id": "b9",
                "title": "The Michigan Single and Multidocument Summarizer for DUC",
                "authors": [
                    {
                        "first": "J",
                        "middle": [
                            "C"
                        ],
                        "last": "Otterbacher",
                        "suffix": ""
                    },
                    {
                        "first": "A",
                        "middle": [
                            "J"
                        ],
                        "last": "Winkel",
                        "suffix": ""
                    },
                    {
                        "first": "D",
                        "middle": [
                            "R"
                        ],
                        "last": "Radev",
                        "suffix": ""
                    }
                ],
                "year": 2002,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Otterbacher, J. C., A. J. Winkel, and D. R. Radev, The Michigan Single and Multidocument Summarizer for DUC 2002, http://www-nlpir.nist.gov/projects/duc/pubs/2002papers/umich_otter.pdf.",
                "links": null
            },
            "BIBREF10": {
                "ref_id": "b10",
                "title": "View-Based and Modular Eigenspaces for Face Recognition",
                "authors": [
                    {
                        "first": "A",
                        "middle": [],
                        "last": "Pentland",
                        "suffix": ""
                    },
                    {
                        "first": "B",
                        "middle": [],
                        "last": "Moghaddam",
                        "suffix": ""
                    },
                    {
                        "first": "T",
                        "middle": [],
                        "last": "Starner",
                        "suffix": ""
                    }
                ],
                "year": 1994,
                "venue": "Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition",
                "volume": "",
                "issue": "",
                "pages": "84--91",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Pentland, A., B. Moghaddam, and T. Starner, \"View-Based and Modular Eigenspaces for Face Recognition,\" Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 21-23 June, 1994, Seattle, Washington, USA, pp. 84-91.",
                "links": null
            },
            "BIBREF11": {
                "ref_id": "b11",
                "title": "Generating natural language summaries from multiple on-line sources",
                "authors": [
                    {
                        "first": "D",
                        "middle": [
                            "R"
                        ],
                        "last": "Radev",
                        "suffix": ""
                    },
                    {
                        "first": "K",
                        "middle": [
                            "R"
                        ],
                        "last": "Mckeown",
                        "suffix": ""
                    }
                ],
                "year": 1998,
                "venue": "Computational Linguistics",
                "volume": "24",
                "issue": "3",
                "pages": "469--500",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Radev, D. R., and K. R. McKeown, \"Generating natural language summaries from multiple on-line sources,\" Computational Linguistics, 24(3), 1998, pp. 469-500.",
                "links": null
            },
            "BIBREF12": {
                "ref_id": "b12",
                "title": "Multiple Document Summarization Using 153",
                "authors": [],
                "year": null,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Multiple Document Summarization Using 153",
                "links": null
            },
            "BIBREF13": {
                "ref_id": "b13",
                "title": "Principal Component Analysis Incorporating Semantic Vector Space Model",
                "authors": [],
                "year": null,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Principal Component Analysis Incorporating Semantic Vector Space Model",
                "links": null
            },
            "BIBREF14": {
                "ref_id": "b14",
                "title": "Centroid-based summar-ization of multiple documents: sentence extraction, util-ity based evaluation, and user studies",
                "authors": [
                    {
                        "first": "D",
                        "middle": [
                            "R"
                        ],
                        "last": "Radev",
                        "suffix": ""
                    },
                    {
                        "first": "H",
                        "middle": [],
                        "last": "Jing",
                        "suffix": ""
                    },
                    {
                        "first": "M",
                        "middle": [],
                        "last": "Budzikowska",
                        "suffix": ""
                    }
                ],
                "year": 2000,
                "venue": "ANLP /NAACL Workshop on Summarization",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Radev, D. R., H. Jing, and M. Budzikowska, \"Centroid-based summar-ization of multiple documents: sentence extraction, util-ity based evaluation, and user studies,\" In ANLP /NAACL Workshop on Summarization, Seattle, WA, April 2000.",
                "links": null
            },
            "BIBREF15": {
                "ref_id": "b15",
                "title": "Evaluation challenges in large-scale multidocument summarization: the MEAD project",
                "authors": [
                    {
                        "first": "D",
                        "middle": [],
                        "last": "Radev",
                        "suffix": ""
                    },
                    {
                        "first": "S",
                        "middle": [],
                        "last": "Teufel",
                        "suffix": ""
                    },
                    {
                        "first": "H",
                        "middle": [],
                        "last": "Saggion",
                        "suffix": ""
                    },
                    {
                        "first": "W",
                        "middle": [],
                        "last": "Lam",
                        "suffix": ""
                    },
                    {
                        "first": "J",
                        "middle": [],
                        "last": "Blitzer",
                        "suffix": ""
                    },
                    {
                        "first": "A",
                        "middle": [],
                        "last": "Celebi",
                        "suffix": ""
                    },
                    {
                        "first": "H",
                        "middle": [],
                        "last": "Qi",
                        "suffix": ""
                    },
                    {
                        "first": "D",
                        "middle": [],
                        "last": "Liu",
                        "suffix": ""
                    },
                    {
                        "first": "E",
                        "middle": [],
                        "last": "Drabek",
                        "suffix": ""
                    }
                ],
                "year": 2001,
                "venue": "Johns Hopkins University CLSP Workshop Final Report",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Radev, D., S. Teufel, H. Saggion, W. Lam, J. Blitzer, A. Celebi, H. Qi, D. Liu, and E. Drabek, \"Evaluation challenges in large-scale multidocument summarization: the MEAD project,\" Johns Hopkins University CLSP Workshop Final Report, 2001.",
                "links": null
            },
            "BIBREF16": {
                "ref_id": "b16",
                "title": "Centroid-based summarization of multiple documents",
                "authors": [
                    {
                        "first": "D",
                        "middle": [
                            "R"
                        ],
                        "last": "Radev",
                        "suffix": ""
                    },
                    {
                        "first": "H",
                        "middle": [],
                        "last": "Jing",
                        "suffix": ""
                    },
                    {
                        "first": "M",
                        "middle": [],
                        "last": "Sty",
                        "suffix": ""
                    },
                    {
                        "first": "D",
                        "middle": [],
                        "last": "Tam",
                        "suffix": ""
                    }
                ],
                "year": 2004,
                "venue": "Information Processing and Management",
                "volume": "40",
                "issue": "",
                "pages": "919--957",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Radev, D. R., H. Jing, M. Sty, and D. Tam, \"Centroid-based summarization of multiple documents,\" Information Processing and Management, 40, 2004, pp. 919-38.",
                "links": null
            },
            "BIBREF17": {
                "ref_id": "b17",
                "title": "The Smart Retrieval System",
                "authors": [
                    {
                        "first": "G",
                        "middle": [],
                        "last": "Salton",
                        "suffix": ""
                    }
                ],
                "year": 1971,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Salton, G., The Smart Retrieval System, Prentice Hall, Englewood Cliffs, N.J. 1971.",
                "links": null
            },
            "BIBREF18": {
                "ref_id": "b18",
                "title": "Introduction to Linear Algebra",
                "authors": [
                    {
                        "first": "G",
                        "middle": [],
                        "last": "Strang",
                        "suffix": ""
                    }
                ],
                "year": 1998,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Strang, G., Introduction to Linear Algebra, Wellesley, MA: Wellesley Cambridge Press, 1998.",
                "links": null
            },
            "BIBREF19": {
                "ref_id": "b19",
                "title": "Eigenfaces for Face Detection/Recognition",
                "authors": [
                    {
                        "first": "M",
                        "middle": [],
                        "last": "Turk",
                        "suffix": ""
                    },
                    {
                        "first": "A",
                        "middle": [],
                        "last": "Pentland",
                        "suffix": ""
                    }
                ],
                "year": 1991,
                "venue": "Journal of Cognitive Neuroscience",
                "volume": "3",
                "issue": "1",
                "pages": "71--86",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Turk, M., and A. Pentland, \"Eigenfaces for Face Detection/Recognition,\" Journal of Cognitive Neuroscience, 3(1), 1991, pp. 71-86.",
                "links": null
            },
            "BIBREF20": {
                "ref_id": "b20",
                "title": "Singular Value Decomposition and Principal Component Analysis",
                "authors": [
                    {
                        "first": "M",
                        "middle": [
                            "E"
                        ],
                        "last": "Wall",
                        "suffix": ""
                    },
                    {
                        "first": "A",
                        "middle": [],
                        "last": "Rechtsteinen",
                        "suffix": ""
                    },
                    {
                        "first": "L",
                        "middle": [
                            "M"
                        ],
                        "last": "Rocha",
                        "suffix": ""
                    }
                ],
                "year": 2003,
                "venue": "A Practical Approach to Microarray Data Analysis",
                "volume": "",
                "issue": "",
                "pages": "91--109",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Wall, M. E., A. Rechtsteinen, and L. M. Rocha, \"Singular Value Decomposition and Principal Component Analysis,\" In A Practical Approach to Microarray Data Analysis (D.P. Berrar, W. Dubitzky, M. Granzow, eds.) Kluwer: Norwell, MA, 2003, pp. 91-109, LANL LA-UR-02-4001.",
                "links": null
            }
        },
        "ref_entries": {
            "FIGREF0": {
                "num": null,
                "uris": null,
                "text": "ravaging, (plundering with excessive damage and destruction)",
                "type_str": "figure"
            },
            "FIGREF1": {
                "num": null,
                "uris": null,
                "text": "broke fire in Delhi. (Action is verb) Today/NN broke/VBD fire/NN in/IN Delhi/NNP Destruction of material happens due to this fire. Destruction//NN of/IN material/NN happens/VBZ due/JJ to/TO this/DT fire/NN ./. Many suffered from the broken glass in the road. (Action is Adjective) Many/JJ suffered/VBD from/IN the/DT broken/JJ glass/NN in/IN the/DT road/NN. /. The authority arrives here soon.The/DT authority/NN arrives/VBZ here/RB soon/RB. /.",
                "type_str": "figure"
            },
            "FIGREF2": {
                "num": null,
                "uris": null,
                "text": "has been given weight as per the contextual word obtained corresponding to it.",
                "type_str": "figure"
            },
            "FIGREF4": {
                "num": null,
                "uris": null,
                "text": "Figure 2. Contribution percentage",
                "type_str": "figure"
            },
            "FIGREF6": {
                "num": null,
                "uris": null,
                "text": "Schemed Screened Set goals Shaped Skilled Solicited Solved Specialized Spoke Stimulated Strategized Streamlined Strengthened Stressed Studied Substantiated Succeeded Summarized Synthesized",
                "type_str": "figure"
            },
            "TABREF0": {
                "num": null,
                "text": "",
                "html": null,
                "content": "<table><tr><td>broke</td><td>fire, glass</td></tr><tr><td>arrives</td><td>Authority</td></tr><tr><td>destruction</td><td>material</td></tr></table>",
                "type_str": "table"
            },
            "TABREF1": {
                "num": null,
                "text": "",
                "html": null,
                "content": "<table><tr><td>Broke</td><td>0.1889 Injuries</td><td>0.1889</td></tr><tr><td>Fire</td><td>0.5669 Thanks</td><td>0.1889</td></tr><tr><td>Delhi</td><td>0.1889 Local</td><td>0.1889</td></tr><tr><td>Mass</td><td>0.1889 communities</td><td>0.1889</td></tr><tr><td>Destruction</td><td>0.1889 Help</td><td>0.1889</td></tr><tr><td>Material</td><td>0.1889 Authority</td><td>0.1889</td></tr><tr><td>Happens</td><td>0.1889 Arrives</td><td>0.1889</td></tr><tr><td>Suffered</td><td>0.1889 Report</td><td>0.1889</td></tr><tr><td>Broken</td><td>0.1889 Casualties</td><td>0.1889</td></tr><tr><td>glass</td><td>0.1889 Road</td><td>0.1889</td></tr></table>",
                "type_str": "table"
            },
            "TABREF2": {
                "num": null,
                "text": "Similarly, the model is extended for multiple documents. This Semantic Vector Space Model is used further to determine important Keywords and henceforth, the summary.",
                "html": null,
                "content": "<table><tr><td/><td>0.7558 Injuries</td><td>0.1889</td></tr><tr><td>Fire</td><td>0.5669 Thanks</td><td>0.1889</td></tr><tr><td>Delhi</td><td>0.1889 Local</td><td>0.1889</td></tr><tr><td>Mass</td><td>0.1889 communities</td><td>0.1889</td></tr><tr><td colspan=\"2\">Destruction 0.3778 Help</td><td>0.1889</td></tr><tr><td>Material</td><td>0.1889 Authority</td><td>0.1889</td></tr><tr><td>Happens</td><td>0.1889 Arrives</td><td>0.3778</td></tr><tr><td>Suffered</td><td>0.1889 Report</td><td>0.1889</td></tr><tr><td>Broken</td><td>0.1889 Casualties</td><td>0.1889</td></tr><tr><td>glass</td><td>0.1889 Road</td><td>0.1889</td></tr></table>",
                "type_str": "table"
            },
            "TABREF3": {
                "num": null,
                "text": "",
                "html": null,
                "content": "<table><tr><td>Science (789 words)</td><td>60.0%</td><td>50.0%</td><td>70.0%</td></tr><tr><td>Geography (725 words)</td><td>55.56%</td><td>33.33%</td><td>22.5%</td></tr><tr><td>History (557 words)</td><td>70.0%</td><td>50.0%</td><td>48.5%</td></tr><tr><td>Average accuracy</td><td>61.85%</td><td/><td/></tr></table>",
                "type_str": "table"
            },
            "TABREF4": {
                "num": null,
                "text": "",
                "html": null,
                "content": "<table><tr><td>Doc No.</td><td>Title of Doc</td><td>Doc Length</td></tr><tr><td>Doc1</td><td>Introduction to Crawler Architecture</td><td>1076 words</td></tr><tr><td>Doc2</td><td>Developing Web Search Engine</td><td>890 words</td></tr><tr><td>Doc3</td><td>Overview of Web Crawler</td><td>945 words</td></tr><tr><td>Doc4</td><td>Future of Search Engines</td><td>970 words</td></tr></table>",
                "type_str": "table"
            }
        }
    }
}