New method for evaluating anti-SPIT in VoIP networks

Unsolicited bulk email or Spam is a menace to the usability of Internet email. The possibility of cheap Voice over IP (VoIP) services have introduced the threat of Spam over Internet Telephony (SPIT). In order to detect SPIT efficiently we need to extract some features which help us in categorizing the incoming calls. In recent researches various features have been introduced but some of these features have less effect in detecting SPIT. In this paper we propose an approach that assigns weight to features and determines which feature is more important than the others in classification of suspicious incoming calls. Once we select the best features, we can detect SPIT efficiently in less time.