A Visual Analytics Approach for User Behaviour Understanding through Action Sequence Analysis

Abstract

Analysis of action sequence data provides new opportunities to understand and model user behaviour. Such data are often in the form of timestamped and labelled series of atomic user actions. Cyber security is one of the domains that show the value of the analysis of these data. Elaborate and specialised models of user-behaviour are desired for effective decision making during investigation of cyber threats. However, due to their complex nature, activity sequences are not yet well-exploited within cyber security systems. In this paper, we describe the initial phases of a visual analytics approach that aims to enable a rich understanding of user behaviour through the analysis of user activity sequences. First, we discuss a motivating case study and discuss a number of high level requirements as derived from a series of workshops within an ongoing research project. We then present the components of a visual analytics approach that constitutes a novel combination of ''action space'' analysis, pattern mining, and the interactive visual analysis of multiple sequences to take the initial steps towards a comprehensive understanding of user behaviour.