Learning Personal Agents with Adaptive Player Modeling in Virtual Worlds

There has been growing interest in creating intelligent agents in virtual worlds that do not follow fixed scripts predefined by the developers, but react accordingly based on actions performed by human players during their interaction. In order to achieve this objective, previous approaches have attempted to model the environment and the user´s context directly. However, a critical component for enabling personalized virtual world experience is missing, namely the capability to adapt over time to the habits and eccentricity of a particular player. To address the above issue, this paper presents a cognitive agent with learning player model capability for personalized recommendations. Specifically, a self-organizing neural model, named FALCON (Fusion Architecture for Learning and Cognition), is deployed, which enables an autonomous agent to adapt and function during the players´ interaction. We have developed personal agents with adaptive player models as tour guides in a virtual world environment. Our experimental results show that we are able to learn user models that evolve and adapt with players in real time. Furthermore, the virtual tour guides with player models outperform those without adaptive player modeling in terms of recommendation accuracy.