Improvement of information filtering by independent components selection

We propose an improvement of an information filtering process with independent components selection. The independent components are obtained by Independent Components Analysis and considered as topics. Selection of independent components is an efficient method of improving the accuracy of the information filtering for the purpose of extraction of similar topics by focusing on their meanings. To achieve this, we select the topics by Maximum Distance Algorithm with Jensen-Shannon divergence. In addition, document vectors are represented by the selected topics. We create a user profile from transformed data with a relevance feedback. Finally, we sort documents by the user profile and evaluate the accuracy by imputation precision. We carried out an evaluation experiment to confirm the validity of the proposed method considering meanings of components used in this experiment. (C) 2008 Wiley Periodicals, Inc.