Passage relevancy through semantic relatedness

Luis Tari, Phan Huy Tu, Barry Lumpkin, Robert Leaman, Graciela Gonzalez, Chitta Baral

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

Abstract

Questions that require answers in the form of a list of entities and the identification of diverse biological entity classes present an interesting challenge that required new approaches not needed for the TREC 2006 Genomics track. We added some components to our automatic question answering system, including (i) a simple algorithm to select which keywords extracted from natural language questions should be treated as essential in the formation of queries, (ii) the use of different entity recognizers for various biological entity classes in the extraction of passages (iii) determining relevancy of candidate passages with the use of semantic relatedness based on MeSH and UMLS semantic network. We present here an overview of our approach, with preliminary analysis and insights as to the performance of our system.

Original languageEnglish (US)
Title of host publicationNIST Special Publication
StatePublished - 2007
Event16th Text REtrieval Conference, TREC 2007 - Gaithersburg, MD, United States
Duration: Nov 6 2007Nov 9 2007

Other

Other16th Text REtrieval Conference, TREC 2007
CountryUnited States
CityGaithersburg, MD
Period11/6/0711/9/07

ASJC Scopus subject areas

  • Engineering(all)

Fingerprint Dive into the research topics of 'Passage relevancy through semantic relatedness'. Together they form a unique fingerprint.

  • Cite this

    Tari, L., Tu, P. H., Lumpkin, B., Leaman, R., Gonzalez, G., & Baral, C. (2007). Passage relevancy through semantic relatedness. In NIST Special Publication