Workshop on the uses of Agents for Education, Games and Simulations


at the Taipei International Convention Centre (TICC)




Taipei, Taiwan on May 2, 2011

Training for complex situations in human societies such as in education, business transactions, military operations, medical care and crisis management can be provided effectively using serious games and simulations. In these types of games and simulations the role of agents to model and simulate naturally behaving characters becomes more and more important. Especially in situations where the games are not just meant to provide fun, but are used to support the learning process it is important that the games achieve their goal and do not just distract (or entertain) the trainee.


A major aim of this workshop is to discuss how to model rational (or non-rational, but natural) behaving agents who are embedded in a social context with other characters and humans. This is especially important when both characters and humans can be pro-active but also have to react to the behaviour of others in their environment.  Thus these characters should have some social conscience of themselves and others and base their decisions for actions on this knowledge. Of course social knowledge may consist of detailed knowledge such as that some person has been your long time friend and thus can be trusted to help you, but also general knowledge such as that society looks bad at people that cheat but adores people that grasp opportunities.  Thus we aim to model also different levels of action and interactions. Both the operational ones such as gestures and general way of animating characters, the tactical decisions such as negotiation tactics when trying to get some help and long term strategies such as behaving cooperative towards your boss in order to secure a promotion.  One of the interesting questions is how these should be modelled and how they interact? And how do current agent architectures support these models?


In general the technologies used in game engines and multi-agent platforms are not readily compatible due to some inherent differences of concerns. Where game engines focus on real-time aspects and thus propagate efficiency and central control, multi-agent platforms assume autonomy of the agents. And while the multi agent platforms offer communication facilities these can or should not be used when the agents are coupled to a game. So, although increased autonomy and intelligence may offer benefits for a more compelling game play and may even be necessary for serious games, it is not clear whether current multi agent platforms offer the facilities that are needed to accomplish this.