Novel Answer Ranking Approach in Question Answering System Using Compositional Distributional Semantic Model

Abstract

A question answering (QA) system provides direct answers to user questions by consulting its knowledge base. In recent decade, there is an explosion in the freely available information, improvements in information technology and the increase in people's desire for the better information access. One question arises related to information “meaning” while giving the exact answer of the question instead of the large document freely available information, which has created the emergence of a wide variety of computational models for modeling the meaning of words based on the underlying assumption of the distributional hypothesis, that the meaning of word can be inferred from its use. This paper explores the role of distributional semantic models (DSM) in Question Answering (QA) system. The idea is to compute the relatedness between the user question and the candidate answer retrieved by the search module. One of the shortcomings of DSMs found while applying for larger linguistic unit is “bag-of-words” limitation.