Abstract

Abstract With the development of computer games, different game worlds and various game characters are found within them. Various Artificial Intelligence (AI) techniques are usually used to define behavior s of the characters within game worlds, which are controlled by AI algorithms in the computer as well as by the user. The AI techniques defined for these characters are generally developed by the game creators and cannot be changed without going to some effort, which means that if a user wished to control the behavior s of a character within a game, they could not easily do so. Being able to edit the behavior s of AI characters is beneficial as it gives the user extra control over their characters. Therefore a method for allowing a user to easily personalize the AI characters was needed. This goal was achieved by using an incremental knowledge acquisition method, called the MCRDR. The MCRDR allows the user to easily acquire new control knowledge of the AI characters by combining rule-based and case-based knowledge acquisition approach. Our experiment results showed that AI of a character could be personalized with this method of knowledge extraction.

Citations