A Method of Detecting Bot Networks Based on Graph Clustering in the Recommendation System of Social Network

This paper proposes a method for detecting a network of bots in the recommendation system based on graph clustering and analysis of user actions. This method is proposed to be used if there are signs of an attack on the recommendation system and a set of probable targets of the attack is revealed. A series of experiments was carried out to test the effectiveness of the proposed method. The input data for the experiments were generated in the developed software simulation model of users and items of the recommendation system that made it possible to simulate the influence of external destabilizing factors on the system such as information attacks. A subsystem of information security of the recommendation system has been developed. Such subsystem consists of a method of detecting an information attack on the recommendation system based on trend analysis in item ratings and a method of detecting bot networks in the recommendation system based on graph clustering and user action analysis. Also, appropriate algorithms have been developed.