An automatic and proactive identity theft detection model in MMORPGs

Abstract

Identity theft happens frequently, especially in popular multiplayer games where cyber-assets can be monetized. In this work, we propose an automatic and proactive identity theft detection model in online games. We specify the identity theft process into exploration, monetization, and theft and pose identity theft detection as a multi-class classification problem. We propose an automatic and proactive detection model utilizing rich features, along with appropriate problem-specific domain knowledge regarding the unique properties of identity theft. The proposed model based on process specification and automatic learning will reduce financial losses to game users and game companies through early detection.