Question analysis model using user modelling and relevance feedback for question answering

Accessing vast volume of information quickly and easily through the Internet has become a major challenge. One way to access information on the web is through a question answering (QA) mechanism. A Question Answering system aims to provide relevant answers to users’ (natural language) questions (queries) by consulting its knowledge base. Providing users with the most relevant answers to their questions is an issue. Many answers returned are not relevant to the questions and this issue is due to many factors. One such factor is the ambiguity yield during the semantic analysis of lexical extracted from the user’s question. The existing techniques did not consider some of the terms, called modifier terms, in the user’s question which are claimed to have a significant impact of returning correct answer. The objective of this research is to propose the question analysis model of the QA system that would correctly interpret all the modifier terms in the user’s question in order to yield correct answers. In question analysis model, a combination of user modelling (UM) and relevance feedback (RF) is used to increase the accuracy of the returned answer. On the one hand, UM helps the QA system to understand the user’s question, manage for question adjustment and increase robustness of the question. Additionally,RF provides an extended framework for QA system to avoid or remedy the ambiguity of the user’s question. The proposed model which utilizes Vector Space Model (VSM) is able to semantically interpret and correctly convert modifier terms into a quantifiable form. These modifier terms in user’s question may involve with evaluation, computation and/or comparison. The proposed model is implemented in a prototype of QA system called QAUF (Question Answering system with User Modelling and Relevance Feedback). The Answer Retrieval Module of QAUF is adopted from FREyA QA system. The experiments are conducted by using the Raymond Mooney gold standard dataset of Geoquery and run on QAUF. The results are then compared with those of the previous existing QA systems, namely Aqualog and FREyA. The proposed model shows a relatively increased in the F-measure where QAUF is 94.7%, FREyA is 92.4% and Aqualog is only 42%.

Preview Text FSKTM 2015 45IR.pdf Download (1MB) | Preview

Actions (login required)

View Item