Exploiting question concepts for query

Abstract

In this paper, we present an efficient semantic query expansion methodology based on a question concept list comprised of terms that are semantically close to concepts represented in a query. The proposed system first constructs a concept list for each question concept and then learns the concept list for each question concept. When a new query is given, the question is classified into the question concept, and the query is expanded using the concept list of the classified concept. In the question answering experiments on 42,654 Wall Street Journal documents of the TREC collection, the traditional system showed in 0.223 in MRR and the proposed system showed 0.50 superior to the traditional question answering system. © Springer-Verlag Berlin Heidelberg 2005.