An Emotional Space Modeling for the Adaptive Emotional Model Design Based on Sugeno Fuzzy Inference

To maximize the quality improvement and tangibility of emotion-based personalized services, a lot of efforts are put into researches on emotional expression languages, measurement of emotions, emotional transference and expression model, personalized emotional space model, emotion-based personalized services, and so forth. To maximize quality improvement and tangibility of emotion-based personalized services, research on emotional modeling for quantitative and structural expression of human emotions needs to precede the others. In addition, a high level of inference on human emotions as well as an emotional model with learning capabilities is necessary for personalized emotion modeling. To this end, this study defines the 12 emotional expression languages, which are defined in Thayer’s Valence-Arousal emotion model, with the fuzzy membership function. For emotional transference and inference modeling based on valence and arousal input, Mamdani and Sugeno Fuzzy Inference Methods are applied and evaluated. In this manner, this study provides the basis for an adaptive emotional inference system based on the personalized emotional model and neuro-fuzzy system required for personalized services.