Information Retrieval via Truncated Hilbert-Space Expansions

Abstract

In addition to the frequency of terms in a document collection, the distribution of terms plays an important role in determining the relevance of documents. In this paper, a new approach for representing term positions in documents is presented. The approach allows an efficient evaluation of term-positional information at query evaluation time. Three applications are investigated: a function-based ranking optimization representing a user-defined document region, a query expansion technique based on overlapping the term distributions in the top-ranked documents, and cluster analysis of terms in documents. Experimental results demonstrate the effectiveness of the proposed approach.