A hybrid approach for spam filtering using support vector machine and artificial immune system
Kunal Jain, Sanjay Agrawal · 2014
Internet is a very powerful tool for information sharing; it provides email, chat, and audio/video talk for communication. All these email are widely used for official and non official communication because it is freely available to users and it also provides file transfer up to some limit. Hence use of Email (Electronic mail) services increasing rapidly due to higher dependency of organizations and individuals. Whenever Emails are sent/received unnecessarily then it is known as spam mail (Unsolicited Bulk Email). Spam Email is major issue for internet community because it causes wastage of resources and also pollutes our environment. Hence spam filtering is essential task. There are many existing techniques and algorithms available which focuses on individual parameters of the malicious content. Many times effectiveness of filtering algorithm gets significantly decreased, whenever spammer attacks on limitations of individual filtering mechanism. In this paper we have introduced an approach which includes advantages of Support Vector Machine and Artificial Immune System. In this approach we are trying to combine various positive properties of these filtering techniques at different level by deploying them in a hybrid approach. We have also discussed shortcomings of traditional spam filtering techniques in comparison of our proposed work.