Modelling of the social engineering attacks based on social graph of employees communications analysis

Abstract

The article is aimed at solving the problem of constructing and analyzing a compressed social graph, taking into account the estimates of the probability of a transition of the intruder's influence from user to user. The subject of the study are user accounts in the social network VKontakte as the basis for building user interaction and social graphs on them. Due to the peculiarities of the graphs under consideration, processing them is a very time-consuming task, requiring large computational resources. In order to optimize the algorithm of the analyzer being developed, the authors of this article conducted a study aimed at modifying the analysis module, with the purpose of diluting the social graph of staff interaction to optimize the process of its further processing. In this paper, we investigate various approaches to diminishing the dimension of edges of a graph, describe their weak and strong sides for the purpose of choosing the optimal graph dilution algorithm. The construction and analysis of the social graph will make it possible to calculate estimates of the security of users of the information system from social engineering attacks and also to analyze the trajectories of the spread of social engineering attacks.