Indian Journal of Science and Technology

Abstract

The growth of the World Wide Web in the last two decades has posed a lot of challenges to the field of Information Retrieval. The way in which the information is collected, shared and searched is changed drastically. Searching information has never been so easy because of the search engines. Any Information Retrieval application has different components such as query handling where the user enters the user information need, Indexing part where the document representation is stored and maintained, the Ranking part which arranges the documents based on the relevance and the matching part which compares the query representation with the document representation. Most of the time since the user information need is not specified correctly, the documents that are retrieved may not be relevant or the relevant links may be less. Hence it is the challenge to be addressed by the search applications which can transform the original query into another representation which will be more responsive for the information retrieval. In this work we propose a hybrid framework that can be used to transform the original query representation to another representation which helps to retrieve more relevant results than the original representation. We further validate our point with an experiment we conducted.
Keywords: Information Retrieval, Ontology, Query Expansion, Query Refinement, Search Engine