Internet Users Authentication via Artificial Intelligence

Abstract

The number of Internet users increases and the Internet is part of people’s daily lives, as a result, the behavior of the user becomes free and informal. This is the basis of the assumption that the manner of user actions on the Internet has become a factor that can be used by authentication using artificial intelligence. In turn, existing works related to users’ web browsing behavior-based authentication with using machine learning do not analyze some important behavioral user’s characteristics, such as patterns of behavior or user behavior on a frequently visited resource. It causes to suggest own features and check their contribution to the accuracy of the system. The aim of this work is to study the possibility of introducing a map of clicks, bigrams, trigrams of frequent web pages and their domains, evaluation of the contribution of added features. In this work, we replace the web pages’ genre classification by domain classification and don’t take into account the spikes in views. We have created a system based on artificial intelligence. As a work result, we have shown a significant improvement in the accuracy of the system using the click map and a slight improvement in the use of bigrams and trigrams.