Motivation: Most of the affective gaming researches are focusing on using physiological measures to determine the emotional state of the user, while others try to create applications where the player can influence the game with his/her inner emotional states. We would like to mix these two approaches taking advantage of the psychological phenomena of anticipation of stress and using it up in games to predict user actions.
Research Approach: Our research focuses on the user action in the course of play, and tries to establish a link between physiological parameters reflecting on the user's emotional sate and the interaction he/she initiates in the game.
Research Design: In our experiments we are recording skin conductance response while playing a simple arcade game. We train artificial neural networks to learn when the user interacts (jumps).
Findings: In our paper we demonstrate that neural networks are not only capable of learning the exact time, but are also able to predict a jump 2 seconds before it is carried out only from the skin conductance data.
[1]
Jonathan Klein,et al.
This computer responds to user frustration
,
1999,
CHI Extended Abstracts.
[2]
Jonathan Klein,et al.
This computer responds to user frustration: Theory, design, and results
,
2002,
Interact. Comput..
[3]
Raúl Rojas,et al.
Neural Networks - A Systematic Introduction
,
1996
.
[4]
Rosalind W. Picard.
Affective Computing for HCI
,
1999,
HCI.
[5]
Regan L. Mandryk,et al.
A continuous and objective evaluation of emotional experience with interactive play environments
,
2006,
CHI.
[6]
Eszter Láng,et al.
Validating a New Method for Ergonomic Evaluation of Human Computer Interfaces
,
1999
.
[7]
James C. Lester,et al.
Predicting User Physiological Response for Interactive Environments: An Inductive Approach
,
2006,
AIIDE.