The improvement of the stylometry-based cognitive assistant performance in conditions of big data analysis

Abstract

The current research is devoted to the problem of individual’s victim behavior in the Internet. The victim behavior is a complex one and can be characterized by numerous markers from the classical victimology point of view. Yet, there are no general approaches to identify the victim behavior in the Internet due to the novelty of this communication environment. In this paper the marker of numerous false pseudonyms is presented to identify the individual’s victim behavior, and the cognitive assistant is proposed as a facility to deal with this marker. The cognitive assistant is based on the stylometry analysis, yet this approach has some scalability and velocity issues under the big data conditions. So, the technique for cognitive assistant performance improvement is presented. The basic estimates are made to prove the efficiency of the proposed complex technique, based on the fog-computing and the distributed ledger technology.