On text mining techniques for personalization

Abstract

The popularity of the Web has made text mining techniques for personalization an increasingly important research topic. We first examine the problem on text mining for building categorization systems. Three different approaches which can be used for building categorization systems are discussed: Classification, clustering and partial supervision. We examine the advantages and disadvantages of each approach. Some Web specific enhancements are discussed. Applications of text mining techniques to collaborative filtering have then been examined. Specifically, a content-based collaborative filtering approach is considered.