Exploiting Structure and Content of Wikipedia for Query Expansion in the Context

International Conference RANLP 2009 - Borovets, Bulgaria, pages 103–106 Exploiting structure and content of Wikipedia for Query Expansion in the context of Question Answering Surya Ganesh, Vasudeva Varma Language Technologies Research Centre, IIIT-Hyderabad, India suryag@research.iiit.ac.in, vv@iiit.ac.in Abstract Retrieving answer containing passages is a chal- lenging task in Question Answering. In this pa- per we describe a novel query expansion method which aims to rank the answer containing pas- sages better. It uses content and structured in- formation (link structure and category informa- tion) of Wikipedia to generate a set of terms se- mantically related to the question. As Boolean model allows a fine-grained control over query expansion, these semantically related terms are added to the original query to form an expanded Boolean query. We conducted experiments on TREC 2006 QA data. The experimental results show significant improvements of about 24.6%, 11.1% and 12.4% in precision at 1, MRR at 20 and TDRR scores respectively using our query expansion method. 1 Introduction Question Answering (QA) aims at finding exact an- swers to natural language questions in a large collec- tion of documents (such as World Wide Web). Within a QA system, passage retrieval reduces the search space for finding the answer from such large collec- tion of documents to a fixed number of passages (say top 20). But this could lead to the case where the set of passages considered may not contain the an- swer. In such cases any QA system would not answer the question. One of the reasons for not retrieving answer containing passages is attributed to the prob- lem of vocabulary mismatch i.e., passages holding the answer to a question have semantic alterations of orig- inal terms in the question. Moldovan [5] showed that their system failed to answer 25.7% of questions solely because of vocabulary mismatch. In Information Re- trieval (IR) such a problem is addressed using the tech- nique of Query Expansion. It is the process of ex- panding the search query to match additional relevant documents. Query expansion techniques like adding synonyms of the query terms and relevance feedback have performed well in many IR applications. But most of these techniques as described by Derczynski [4] were not successful in the context of QA. In this paper we describe a novel query expansion method using Wikipedia. Wikipedia is the leading open encyclopedia with a wide coverage on diverse topics, events, entities, etc. It is a reliable data-source and has found its use in many applications [1]. An- other factor which motivated us to use it, is based on the simple experiment we conducted using TREC 2006 QA test set [3]. The test set consists of question series where each series asks for information regarding a par- ticular target. The targets in the test set include peo- ple, organizations, events and other entities. Because of low data redundancy in Wikipedia, the coverage of its articles is directly proportional to the size of the text content in them. So in this experiment, we search for size …

Content truncated to 500 words