Security against sybil attack in social network
Narendra M. Shekokar, Krishna B. Kansara · 2016
Sybil (fake) accounts penetrate the OSN security by hosting multiple threats. Multiple social-graph-based defence schemes have been proposed till date which can effectively detect and isolate the sybils based on the ground truth of limited social connections between non-sybil (honest) and sybil users. In real world scenario, sybils may elude these defenses by imploring many social connections to real users. This situation may degrade the performance of social graph based sybil detection schemes. So, existing social graph based sybil identification further requires new solutions. A new approach would be more convincing if it enriches the structure of a social graph with more information about the behavioural analysis of its users. In this paper, we have proposed a sybil node identification (SNI) method based on classical social graph based sybil detection scenario and SNI-B which is the extension to SNI by incorporating user behavioural aspects. Our Simulation results show that social graph based SNI outperform if combined with behavioural aspects (SNI-B), of users even in case of implored sybil connections over the time.