Automatic answer validation system on english language

Partha Pakray, Santanu Pal, Sivaji Bandyopadhyay, Alexander Gelbukh

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

6 Scopus citations

Abstract

We present an Answer Validation System (AV) based on Textual Entailment and Question Answering. The important features used to develop the AV system are Named Entity Recognition, Textual Entailment, Question-Answer type Analysis and Chunk Boundary and Dependency relations. Separate AV modules have been developed for each of these features. We first combine the question and the answer into Hypothesis (H) and the Supporting Text as Text (T) to check the entailment relation as either "VALIDA TED" or "REJECTED". The important lexical features used for the Textual Entailment Module in the present system are: WordNet based unigram match, bigram match and skip-gram. The results obtained from the four answer validation modules are integrated using a voting technique. Evaluation scores obtained on the AVE 2008 test set show 67% precision for "VALIDATED" decisions.

Original languageEnglish
Title of host publicationICACTE 2010 - 2010 3rd International Conference on Advanced Computer Theory and Engineering, Proceedings
PagesV6329-V6333
DOIs
StatePublished - 2010
Event2010 3rd International Conference on Advanced Computer Theory and Engineering, ICACTE 2010 - Chengdu, China
Duration: 20 Aug 201022 Aug 2010

Publication series

NameICACTE 2010 - 2010 3rd International Conference on Advanced Computer Theory and Engineering, Proceedings
Volume6

Conference

Conference2010 3rd International Conference on Advanced Computer Theory and Engineering, ICACTE 2010
Country/TerritoryChina
CityChengdu
Period20/08/1022/08/10

Keywords

  • Answer validation exercise (AVE)
  • Chunk boundary and dependency relations
  • Named entity
  • Question type
  • Textual entailment (TE)

Fingerprint

Dive into the research topics of 'Automatic answer validation system on english language'. Together they form a unique fingerprint.

Cite this