Robust question answering for speech transcripts using minimal syntactic analysis

Pere R. Comas, Jordi Turmo, Mihai Surdeanu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Scopus citations

Abstract

This paper describes the participation of the Technical University of Catalonia in the CLEF 2007 Question Answering on Speech Transcripts track. For the processing of manual transcripts we have deployed a robust factual Question Answering that uses minimal syntactic information. For the handling of automatic transcripts we combine the QA system with a novel Passage Retrieval and Answer Extraction engine, which is based on a sequence alignment algorithm that searches for "sounds like" sequences in the document collection. We have also enriched the NERC with phonetic features to facilitate the recognition of named entities even when they are incorrectly transcribed.

Original languageEnglish (US)
Title of host publicationAdvances in Multilingual and Multimodal Information Retrieval - 8th Workshop of the Cross-Language Evaluation Forum, CLEF 2007, Revised Selected Papers
PublisherSpringer-Verlag
Pages424-432
Number of pages9
ISBN (Print)3540857591, 9783540857594
DOIs
StatePublished - 2008
Externally publishedYes
Event8th Workshop of the Cross-Language Evaluation Forum, CLEF 2007 - Budapest, Hungary
Duration: Sep 19 2007Sep 21 2007

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5152 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other8th Workshop of the Cross-Language Evaluation Forum, CLEF 2007
Country/TerritoryHungary
CityBudapest
Period9/19/079/21/07

Keywords

  • Phonetic Distance
  • Question Answering
  • Spoken Document Retrieval

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Robust question answering for speech transcripts using minimal syntactic analysis'. Together they form a unique fingerprint.

Cite this