An Approach to Detect Fake Profiles in Social Networks Using Cellular Automata-Based PageRank Validation Model Involving Energy Transfer

Social Network analysis has been considered as an important and promising research area in today’s digital era as many human users in their real lives are found to be directly influenced by such Online Social networks, and unfortunately, very often such human users have been found to be influenced in mostly negative ways caused by fake users of such Social Networks. For this reason, detection of user(s) influence and or, fake user detection in any Social Network scenario may play a crucial role in the reduction of vulnerability in Social Networks. To facilitate enhanced trustworthiness of Social Networks, a novel approach involving an energy transfer-based PageRank validation using Cellular Automata (CAs) has been presented and further has been investigated in this paper to explore its scope for possible usage in Social Network scenarios. For this purpose, a new metric, i.e., energy-based influence score (influence score), has been introduced in this paper to facilitate an easy and effective computation of user influence, in which the user influence is measured based on energy transfer among users of the Social Networks during sending and or, receiving connection requests. Furthermore, a detailed discussion towards possible fake profile(s) detection in any social network scenario facilitating proposed energy-based influence scores has been presented in the presented research. Several analyses of different data as collected from computer simulation and real-time web traffic have confirmed the efficiency and cost-effectiveness of our proposed approach.