Hybrid multi-criteria decision-making algorithm for music composition evaluation using T-spherical fuzzy sets

This dissertation will introduce a robust way of rating the music composition, dealing with the subjectivity and vagueness inherent in such an evaluation. A hybrid multi-criteria decision-making (MCDM) algorithm based on T-spherical fuzzy analytic hierarchy process (T-SF-AHP) for weighing the criteria and T-spherical fuzzy TOPSIS (T-SF-TOPSIS) for ranking compositions is proposed. Expert and listener input identified five key criteria - musicality, creativity, emotional impact, technical complexity, and audience appeal. The method was based on a dataset of 1,000 diverse compositions and realised high alignment with expert (ρ = 0.92) and listener (ρ = 0.88) rankings. Compared to traditional fuzzy and crisp MCDM approaches, it yields more accurate and efficient results in 32 minutes while processing the dataset. Integrating T-spherical fuzzy sets improves the model's competence in resolving ambiguity and conflicting criteria. We provide a scalable evaluation framework that is applicable to music competitions as well as streaming and educational platforms and, potentially, to other types of problems in which subjective ratings need to be assessed.