Trained Trigger Language Model for Sentence Retrieval in QA: Bridging the Vocabulary Gap
Saeedeh Momtazi and Dietrich Klakow
In: CIKM 2011, 24 Oct - 28 Oct 2011, Glasgow, UK.

Abstract

We propose a novel language model for sentence retrieval in Question Answering (QA) systems called trained trigger language model. This model addresses the word mismatch problem in information retrieval. The proposed model cap- tures pairs of trigger and target words while training on a large corpus. The word pairs are extracted based on both unsupervised and supervised approaches while dierent no- tions of triggering are used. In addition, we study the im- pact of corpus size and domain for a supervised model. All notions of the trained trigger model are nally used in a language model-based sentence retrieval framework. Our ex- periments on TREC QA collection verify that the proposed model signicantly improves the sentence retrieval perfor- mance compared to the state-of-the-art translation model and class model which address the same problem.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.

[Edit]