Automatic Identification of Human Strategies by Cognitive Agents

So far most cognitive modeling approaches have concentrated on modeling and predicting the actions of an “average user” – a user profile that in reality often does not exist. User performance is highly dependent on psychological factors like working memory, planning depth, search strategy etc. that differ between users. Therefore, we propose a combination of several AI methods to automatically identify user profiles. The proposed method assigns each user a set of cognitive agents which are controlled by several psychological factors. Finally, this method is evaluated in a case study on preliminary user data on the PSPACE-complete planning problem Rush-Hour.