Towards Dynamically Crafted Affective Experiences Through Emotional Response Player Modelling

Abstract

Designing adaptive games for individual emotional experiences is an elaborate task due to the necessary dedicated hardware, complex signal processing and unique emotional bonds each player may forge with the game. The research work I propose focuses on developing a novel affective interaction scheme where players’ emotional reactions are learned by a computational model that leverages this information to influence their future emotional states. To this end I aim at interpreting the causality relations between physiologically measured emotional response variations and their eliciting events. I then plan using these reactions to build each player’s individual emotional reaction model, which will guide the game engine’s logic on how to plan future game events, in order to elicit a target affective experience. Ultimately, I expect this technique to allow game designers to focus on defining high-level rules for generating their envisioned affective gameplay experience, rather than in manually tuning it in a repetitive and time-consuming fashion.