Updated Spam Filtering Algorithm Based on K-Means Clustering
LI Xiang-lin · Journal of Lanzhou Polytechnic College · 2010
The disadvantages of anti-spam technigue which is based on content filtering are analysed.This kind of methods often suffer from low recall rate due to concept drift and skewed class distribution.An updated spam filtering algorithm is proposed based on spam detection method put forward by Minoru Sasaki and Hiroyuki Shinnou,which improves the feature selection algorithm and adds the automatic learning mechanism.The results show that it can effectively deal with the issues of concept drift and skewed class distribution.