Collaborative Learning of Term-Based Concepts for Automatic Query Expansion

Abstract

Information Retrieval Systems have been studied in Computer Science for decades. The traditional ad-hoc task is to find all documents relevant for an ad-hoc given query but the accuracy of adhoc document retrieval systems has plateaued in recent years. At DFKI, we are working on so-called collaborative information retrieval (CIR) systems which unintrusively learn from their users search processes. In this paper, a new approach is presented called term-based concept learning (TCL) which learns conceptual description terms occurring in known queries. A new query is expanded term by term using the previously learned concepts. Experiments have shown that TCL and the combination with pseudo relevance feedback result in notable improvements in the retrieval effectiveness if measured the recall/precision in comparison to the standard vector space model and to the pseudo relevance feedback. This approach can be used to improve the retrieval of documents in Digital Libraries, in Document Management Systems, in the WWW etc.