Combining Multiple Term Selection Methods for Automatic Query Expansion in Pseudo Relevance Feedback using Rank Score Method

In Information Retrieval (IR) System, the retrieval of relevant documents is a challenging task. The Automatic Query Expansion (AQE) technique helps IR system to retrieve the relevant documents. In AQE the selection of candidate terms for query expansion plays a critical in retrieval of documents. The individual term ranking methods produces a set of candidate terms for expansion. In this paper we propose an approach to combine results of the individual term selection methods in to a single list of candidate terms using rank score algorithm. We have compared the performance of individual term selection methods over the combined approach, in which the experimental results shows that our approach is significantly improve the performance of the IR system.