Towards Collaborative Information Retrieval: Three Approaches

The accuracy of ad-hoc document retrieval systems has plateaued in the last few years. At DFKI, we are working on so-called collaborative information retrieval (CIR) systems which unintrusively learn from their users’ search processes. As a first step towards techniques, we focus on a restricted setting in CIR in which only old queries and correct answer documents to these queries are available for improving on a new query. For this restricted setting we propose three initial approaches, called QSD, QLD, and TCL as well as combinations of these approaches with pseudo relevance feedback. The approaches are evaluated experimentally on standard Information Retrieval test collections. It turns out that in particular the hybrid approaches with pseudo relevance feedback give promising results. A bigger advantage of the proposed approaches is expected in real word test scenarios in which the overlap of user interests is larger than in our experimental set up.