A contribution to the method of automatic identification of human emotions by using semantic structures

Abstract

The study of human emotions has been significantly important in recent years, mainly due to its incidence in human behavior. Additionally, having semantic tools that infer emotions from multisensory sources is a crucial aspect, especially because the feelings or actions of a person might be identified through these semantic tools. In the present research, a methodology that uses semantic structures is proposed in order to identify complex emotions on the basis of simple emotions. For this purpose, the SHEO ontology was used. This ontology is designed to conceptualize simple emotions, combine them, and work with axioms and rules that infer complex emotions. SHEO takes simple emotions as instances. These emotions can be identified using computer algorithms. This is demonstrated in the testing phase in which the authors of this research designed the software called DetectionEmotion, which is used to identify simple emotions in video and text. The result of the authors' proposal proved the easiness to infer complex emotions by using SHEO. SHEO is not a final solution in this research, but rather a contribution to the semantic management of emotions.