Exploring domain-specific term weight in archived question search

Abstract

Community Question Answering services, e.g., Yahoo! Answers, have accumulated large archives of question answer (QA) pairs for information and answer retrieval. An effective question retrieval model is essential to increase the accessibility of the QA archives. QA archives are usually organized into categories and question search can be performed within the whole collection or within a certain category. In this paper, we explore domain-specific term weight for archived question search. Specifically, we propose a novel light-weighted term weighting scheme that exploits multiple aspects of the domain information. We also introduce a framework to seamlessly integrate domain-specific term weight into the existing retrieval models. Extensive experiments conducted on real Archived QA data demonstrate the utility of the proposed techniques. © 2010 ACM.