Profile Similarity Communication Matching Approaches for Detection of Duplicate Profiles in Online Social Network
Revathi S A, Dr.M. Suriakala · 2018
In today’s world, the social life of people has close connection with the social networking Sites, for example, Facebook, Twitter and LinkedIn. Most of the social networks have weak user authentication technique, which mainly depends on fundamental details such as displayed user name and photo. Fake or unauthorized users abuse the authorized users’ details such as name, text messages, photographs and videos with or without the approved user’s consent. Profile cloning is a major security problem in social networks as it creates a profile which is homogeneous or similar to the existing ones. Profile cloning detection facilitates the chance to detect frauds in social networks, which can attract people’s trust to collect social data. In this paper an effort is accomplished to give an idea of profile cloning recognition in Online Social Networks (OSN) utilizing Network Theory. This study examines Node Similarity Communication Matching algorithm utilizing profile cloning recognition in Online Social Network depending on malicious user’s latest activities in the social network. In this proposed method, the various activities to be studied includes the activities such as Updates, Wall posts and comments, By recent activities etc. Malicious which hacks users’ identity is identified based on the comparison of threshold values of user’s personal profile attributes and network similarity analysis. The processes used in the research include Creating Account, user operation, monitoring, searching recent activity, Detecting cloned profile process, selecting profile to be examined, and deciding Real/Fake profile. The processes are additionally examined with experiments and the outcomes clearly revealed that the proposed algorithms are beneficial and effective when compared with the known methods. It is exhibited with large number of profile data that the method used can find the cloning profile with about 93.87% accuracy.