Social context based naive bayes filtering of spam messages from online social networks

Abstract

Nowadays, online social networking (OSN) sites are an inevitable way of communication. Almost all the OSNs have attained an explosive growth in the last few years. Spammers or hackers found that OSN is an important and easy way to spread spam messages over the network because of its popularity. Spammers use different strategies to spread spam. So, spam detection must be strong enough to detect spam effectively. Though several spam detection techniques are available, it is necessary to increase the accuracy of spam detection techniques. In this paper, a spam detection technique is proposed to detect and prevent spam messages. In addition to the usage of basic classifiers, social context features such as shares, likes, comments (SLC) are also done. Experimental evaluations and comparisons prove that the proposed system with SLC factors provides a higher accuracy than accuracy of basic classifier.