Abnormal user detection of malicious accounts in online social networks using cookie based cross verification

Abstract

Malicious account detecting is a serious problem on the Internet today. Online social media services like Facebook, LinkedIn, and Instagram, these services include good quality service like opinions, comments as well as poor quality services like rumors, spam, and other malicious activity. In this paper, we review the existing research work done on Facebook, Instagram and LinkedIn, study the techniques used to identify and analyze the poor quality content on Facebook, and other social networks, and we proposed a combined technique like dynamic user profile verification and cookie-based cross-verification to detect malicious activity in an online social network by using random forest machine learning algorithm We also attempt to understand the limitations posed by Facebook in terms of availability of data for collection, and analysis, and try to understand if existing techniques can be used to identify and study poor quality content on Facebook and other social networks.