Improving query expansion using link analysis

Abstract

Query expansion techniques aim to improve a userpsilas search by adding new query terms to an existing query. This can be done in two ways: by having a human user refines the first result set or by using automatic extraction of terms from the first result set. Both techniques usually depend on terms frequency in Web pages only. This paper proposes to improve upon query expansion technique using a link analysis technique called hypertext induce topic selection algorithm (HITS). The preliminary result using questions set from TREC is presented.