Journal of Intelligent Computing Luyin Shao, Binshan Zhao 15 4 2024 https://doi.org/10.6025/jic/2024/15/4/142-148 https://www.dline.info/jic/fulltext/v15n4/jicv15n4_4.pdf Music can be heard everywhere in the fast-paced and high-stress modern society. With the development of time, people’s demand for music has shifted from simple leisure entertainment to seeking emotional resonance. To address this, the big data Boruta algorithm has emerged as an alternative to random forest classification algorithms, enabling more precise classification results. By re-implementing it, we applied the LightGBM algorithm to a Turkish music dataset and used random forest, XGBoost, and LightGBM algorithms for classification prediction. We found their better applicability to these datasets by evaluating these algorithms’ accuracy, Kappa coefficient, and Hamming distance.