A Novel Approach For Generating Rules For SMS Spam Filtering Using Rough Sets

摘要

Abstract: Spam is defined as unwanted commercial messages to many recipients. Email Spamming is a universal problem with which everyone is familiar. This problem has reached to the mobile networks also now days to a great extent which is referred to as SMS Spamming. A number of approaches are used for SMS spam filtering like blacklist-white list filter, Content based filter, Bayesian filtering, checksum filter, heuristic filter. The most common filtering technique is content based spam filtering which uses actual text of messages to determine whether it is spam or not. Bayesian method represents the changing nature of message using probability theory. Bayesian classifier can be trained very efficiently in supervised learning. We have used a new mathematical approach Rough set Theory. Rough Set Theory is a new methodology which is used to cluster the objects of a decision system with a large data set. In this dissertation, the Na've Bayes and the RST method are implemented.