| We present a novel hybrid approach for combining context-sensitive item-based Collaborative Filtering (CF) and Case Based Reasoning (CBR). Traditional CF methods involve the filtering of information along a single perspective and recommend the entire information item to the user. We develop an information filtering strategy that allows the use of recommendations by peers with similar interests and domain experts to guide the selection of information. Our framework, named PRECiSE---Personalized Recommendations in a Context-Sensitive Environment, consists of two phases: In phase I, we have developed a contextual item-based CF strategy in which personal interests initiate and shape the information retrieval criterion and serve as a precursor to peer based recommendations for pertinent information objects. We establish a rich context that takes into account multiple perspectives to support collaborative filtering. (Abstract shortened by UMI.)... |
