Improved social network aided personalized spam filtering approach using RBF neural network
Shatabdi M. Bhalerao, Madhuri A. Dalal · 2017
In present day to day life the important communications are going on through the emails and social networking websites. Whatever the spam is main issue into the email systems. To save against unsolicited e-mails there are number of techniques presented with goal of efficient, accurate spam filtering. We studied the recent technique called social network known as the Personalized & focused spam filter (SOAP) using Bayesian spam filtering technique. SOAP showing better results as compared to existing methods; however this method further improved in terms of accuracy, efficiency and complexity by this paper. We proposed extension to SOAP method by using RBF (Radial Basis Function) RBF neural network rather than naïve Bayes method for spam filtering. This new technique is named as I SOAP (Improved SOAP). Into the ISOAP, each node to their social mates, as example, the nodes which is from the distributed overlay through direct use of the social network links like the overlay links. Into the distributed manner every node are utilizing the ISOAP for collecting the data & checking the spam. The last spam filters which is concentrating on parsing keywords or the blacklist building was unlinked. Into the every node, Into the RBF Neural Network filtering ISOAP has been built the four elements that was management of adaptive trust, social nearness based spam filtering, notification of friend & spam filtering of social interest based. Into the final output of experiment were shows that for the spam filter our proposed approach is very efficient.