We propose a methodology for improving the accuracy of models that predict self-reported player pairwise preferences. Our approach extends neuro-evolutionary preference learning by embedding a player modeling module for the prediction of player preferences. Player types are identified using self-organization and feed the preference learner. Our experiments on a dataset derived from a game survey of subjects playing a 3D prey/predator game demonstrate that the player model-driven preference learning approach proposed improves the performance of preference learning significantly and showcases promise for the construction of more accurate cognitive and affective models.
Hector P. Martinez, Kenneth Hullett, and Georgios N. Yannakakis. Extending Neuro-evolutionary Preference Learning through Player Modeling. IEEE Conference on Computational Intelligence in Games (CIG 2010), August 2010.
The final version of the paper can be found here.