ABSTRACT
Applications of text classification technology are becoming widespread. In the defense against spam email, suspect messages are flagged as potential spam and set aside to facilitate batch deletion. News articles are automatically sorted into topic channels and conditionally routed to individuals based on learned profiles of user interest. In content management, documents are categorized into multi-faceted topic hierarchies for easier searching and browsing. Shopping and auction Web sites do the same with short textual item descriptions. In customer support, the text notes of call logs are categorized with respect to known issues in order to quantify trends over time [3].These are but a few examples of how text classification is finding its way into applications. Readers are referred to the excellent survey by Sebastiani [13].
