Research Track 1: Cognitive Map – From Theory to Applications

In work on the theoretical foundations of AI, we have been leading research efforts on extending cognitive map (CM) and fuzzy cognitive map (FCM) theory with temporal, dynamic, probabilistic and learning properties to enrich the capability of modeling complex real time problems.

Track 2: Agent Mediated Interactive Digital Media

One of my main research goals in the next five years is to infuse intelligent agents into various new media (virtual, mixed, mobile and pervasive media, etc.) to create new experiences and dimensions in game design and interactive narrative. I believe that modeling and designing interactive storytelling/games as multi-agent systems represent an important new paradigm in game research. I am particularly interested in exploring player-centered interactive games/storytelling that enable situated community learning, transform the challenges of learning into playing and promote emotional maturity in young players who spend a lot of time on virtual/networked/mobile games to the detriment of their studies and risk losing their sense of reality. I am therefore interested in real time player/player behavior modeling including emotion and emotional behavior modeling, as well as trust and reputation modeling which is lacking in current games.

Research Track 3: Trust Modeling and Trust Agent

In psychology and sociology, trust measures how a member of a group is trusted by the other members. Attack resistance is an important property of trust metrics which reflects their ability not to be overly influenced by agents who try to manipulate the trust metric and who participate in bad faith. The first commercial computational trust model in computer software is eBay's Feedback Rating. We have led research efforts in proposing entropy-based distributed computational trust models, which have been proven to perform well in identifying and mitigating the adverse effects caused by dishonest referrers in a multi-agent system. The proposed model won the 2006 AAMAS trust agent competition. We have also applied the proposed model in web-service selection, virtual communities, online virtual worlds and other applications.

We have drawn up a road map for subsequent research in the area of trust and reputation management in multi-agent systems. Preliminary work has been done in the direction of formalizing trust ability as a core capability for the next generation of intelligent agents.

Research Track 4: Cognitive Web/Grid and Cognitive Radio

The growth of CR during the last 3 years starting with about 6 to 8 reports and conference papers to 1154 IEEE Papers and many others published by Springer, and Elsevier. CR is predicted to be the fifth generation of wireless communication systems. The CR network is a complex multiuser wireless communication system capable of emergent behavior. It embodies agent-like functions such as perceiving the radio environment and learning from the environment so as to dynamically allocate radio resources to facilitate efficient utilization of the radio spectrum in a fair-minded way.

One of my recent research areas is to apply agent technologies in CR. Together with my PhD students, we collaborate with leading researchers in this exciting field. Our work opens up new and important topics in CR, such as infusing trust modeling techniques to allocate radio resources dynamically based on reputation, proposing nature-inspired algorithms for adaptive radio resource allocation, and modeling cognitive radio network as a multi-agent system.