Inferencing Emotions through the Triangulation of Pupil Size Data, Facial Heuristics and Self-Assessment Techniques

This paper presents a proposal for the introduction of the affective dimension in online learning applications. The paper focuses especially on affective data assessment by presenting a study of `emotional inference´ through the triangulation of three techniques: facial expression interpretations, pupil size and students as self-evaluators. In order to validate the combination of these techniques as a specific methodology, we planned a test with seven students interacting with a virtual learning environment. At the present time we are ready to present the first results and the correlations between the used techniques. The study is aimed to contribute to the development of truly affective computer-based learning applications. More specifically, we believe it can contribute to telemedicine as it promotes a methodology for subject emotional data analysis, being computer-based affective analysis a young field that needs new developments.