Application of fuzzy relational interval computing for emotional classification of music

Automated detection of emotions in a music piece is a multi-context problem due to the multiplicity of emotions and the overlapping hierarchies of physical factors. Classification of music based on human emotions becomes a complex computational task, which requires a simultaneous multidisciplinary approach. In this paper we propose a fuzzy relational interval computing based model for classification of music that works within the context of emotion depiction by its physical properties. We use interval based BK fuzzy relational products to factor hierarchy of physical properties in our computation for analysis and classification. We also generate a fuzzy interval based data model with the help of checklist paradigm, which limits its dependence on human perception.