0.25
0.5
0.75
1.25
1.5
1.75
2
Anaphora and coreference resolution: still a hard nut to crack? How far has it gone, what is its impact on NLP and what are the ways forward?
Published on May 08, 20181185 Views
Anaphora and coreference resolution are arguably among the most challenging Natural Language Processing (NLP) tasks. Research in anaphora resolution and coreference resolution has focused almost exclu
Related categories
Chapter list
Anaphora and coreference resolution: still a hard nut to crack? How far has it gone, what is its impact on NLP and what are the ways forward?00:00
Outline of the presentation00:39
This presentation is partially based on...01:08
Anaphora and conference resolution: three pereninial questions01:25
Anaphora: basic notions and terminology - 101:30
Anaphora: basic notions and terminology - 202:18
Anaphora: basic notions and terminology - 302:52
Example03:11
Anaphora vs. coreference03:57
Anaphora (and coreference) resolution - 105:53
Anaphora (and coreference) resolution - 206:15
Anaphora and machine translation - 106:44
Anaphora and machine translation - 207:21
How far has anaphora resolution gone08:03
Intrinsic evaluation results08:49
The mystery of the original results - 111:55
Why are results so different?13:08
The mystery of the original results - 215:41
The issue of complexity of evaluation data15:48
Quantifying the complexity va the evaluation workbench16:19
Mysteries in evaluation16:54
Objectivity?17:48
Reluctance18:57
The mystery of the evaluation results20:10
Evaluation in anaphora resolution: status quo20:49
Objectives of study - 121:53
Objectives of study - 222:16
Study 1 - 122:32
Study 1 - 222:44
MARS - fully automatic pronoun resolution22:45
Antecedent indicators23:13
Example boosting indicator23:31
Example impeding indicator23:56
Study 1 - 324:23
Impovements in MARS 2002 - 124:24
Impovements in MARS 2002 - 224:40
Impovements in MARS 200624:58
Study 1 - 425:48
Evaluation data - 125:48
Evaluation data - 226:09
Study 1 - 526:39
Extrinsic evaluation26:39
Text summarisation26:42
Summarisation - 126:49
Summarisation - 227:12
Summarisation - 327:57
Summarisation - 428:03
Term extraction - 128:32
Term extraction - 228:42
Term extraction - 328:47
Term extraction - 429:00
Text categorisation - 129:09
Text categorisation - 229:14
Text categorisation - 329:19
Text categorisation - 429:35
Structure of the presentation29:55
Discussion29:56
Would dramatic improvement in anaphora resolution lead to a marked improvement of NLP applications? - 130:18
Would dramatic improvement in anaphora resolution lead to a marked improvement of NLP applications? - 230:46
Would dramatic improvement in anaphora resolution lead to a marked improvement of NLP applications? - 331:07
Study 2 - 131:45
The impact of coreference resolution on NLP applications31:59
BART toolkit - 132:07
BART toolkit - 232:30
Text summarisation32:32
The summarisation experiment - 132:37
The summarisation experiment - 232:43
Results and discussion - 132:47
Text categorisation33:09
Text classification experiments - 133:12
Text classification experiments - 233:17
Results and discussion - 233:22
Textual entailment33:53
Textual entailment experiments33:56
Results - Textual entailment experiments - 134:04
Results - Textual entailment experiments - 234:18
Final word - 134:20
Final word - 234:33
Anaphora and coreference are really a hard nut to crack...34:39
Ways forward36:23
My latest research - 137:38
My latest research - 238:12
My latest research - 1338:31
Anaphora and eye tracking39:07
Eye-tracking corpus39:24
Study 1 - 639:46
Study 1 - 739:54
Study 2 - 240:15
Study 2 - 340:46
Study 2 - 440:51
Conclusion41:41
Contact details41:47