ADVANCED TECHNIQUES FOR SPAMMER DETECTION AND FAKE USER IDENTIFICATION ON SOCIAL NETWORKS

Abstract

Social networks are used by millions of drug addicts around the world. Transactions on social sites such as Twitter and Facebook have a huge impact on daily life and sometimes lead to unwanted consequences. Popular social networks have become targeted platforms for spammers who spread a huge amount of useless and dangerous information. For example, Twitter has become one of the most used platforms of all time, which can send a disproportionate amount of spam. Fake drug addicts send unwanted tweets to drug addicts to promote services and websites that not only influence real drug addicts but also disrupt their resource consumption. It also increases the chances that junkies will use fake identities to obtain invalid information, leading to the spread of malicious content. Recently, detecting spammers and associating fake junkies on Twitter has become a popular research area in modern Internet Social Networks (OSNs). In this article, we discuss methods to identify spammers on Twitter. We also present a number of methods for detecting spam on Twitter and evaluate the methods according to their ability to detect (i) fake content, (ii) URL-based spam, (iii) trending content spam, and (iv) fake junkies. The presented methods are compared based on certain features such as stoner features, happiness features, graph features, structural features, and temporal features. We hope that the presented work will be a useful resource for experimenters to find the most important recent developments in Twitter spam detection on a single platform.