VIS UALIZATIONOF THE ANOMALOUS ACTIVITYIN TRAJECTORIES OF THE EMPLOYEES’ OF THE CRITICAL INFRASTRUCTURE
Thepaperpresentsanapproachtoanalysisofthemovementsofthecriticalinfrastructurestuff characterized by usage of the data mining algorithms and interactive visualization techniques. The groups of the employees with similar behavior are determined using Kohonen self-organizing maps, which are set up using special BandView visualization model designed by the authors. To detect anomalies in employees’ behavior, a special mechanism for rating deviations based on assessment of their spatiotemporal attributes is proposed. The approach his test edusing data set given within VAST MiniChallenge-2 2016, which describes employees’ movement in the organization building.
Authors: I. N. Murenin, E. S. Novikova
Direction: Informatics and Computer Technologies
Keywords: Anomaly detection, visual analytics, heat maps, self-organizing maps, behavior patterns, anomaly rating
View full article