Date Added: Jun 2012
With the increasing popularity of a E-mail users, E-mail spam problem growing proportionally. Spam filtering with near duplicate matching scheme is widely discussed in recent years. It is based on a known spam database formed by user feedback which cannot fully catch the evolving nature of spam and also it requires much storage. In view of above drawbacks, the authors proposed an effective spam detection scheme based on Bayesian approach. First, they use Bayesian filter based on single keyword sets and then they improve the scheme by using multiple keyword sets and satisfactory results obtained.