www.ai-center.com


Reference: goh-10-pleasurepropagation

Reference Goh, C. K., and Nareyek, A. 2010.
Pleasure Propagation to Reward Predictors.
In Proceedings of the 2010 IEEE Conference on Computational Intelligence and Games (CIG 2010), 62-68.
Abstract

Learning has always been one of the core mechanics in computer games. Players learn specific battle mechanics, control schemes, and much more, which enables them to progress further through the game and experience pleasure. Previous studies on learning often focused on the learning of predictors (cues) for reward and their motivational properties, but fail to address the impact on pleasure. For example, the cue of a smell of chocolate will already produce a pleasure feeling, which means that apart from the motivational learning of the smell cue, something affected pleasure generation as well. In this paper, we perform an experimental study to investigate this relation of learning and pleasure. The study, in which 38 test subjects participated, used smile reactions to rewards of showing funny pictures as a primary measure of pleasure. The findings of the study demonstrate that experiment repetition leads to an increase in pleasure at the cue, a decrease in pleasure at the reward, and a potential relation between motivation and pleasure change at the cue.

Download Rules and Regulations     [PDF; 889 KB]     [download PDF viewer]