Live Music popularity prediction using genre and clustering based classification system: A machine learning approach

• Published in: 2023 9th International Conference on Smart Computing and Communications (ICSCC) pp. 67 - 71
• Main Authors: Agarwal, Saket, Goyal, Jayant, Thapa, Sneha, Deshpande, Akshada, Aryan, Kumari, Deepa
• Format: Conference Proceeding
• Language: English
• Published: IEEE 17.08.2023
• Subjects: Clustering Base Prediction; Clustering methods; Extra tree Regressor; Feature Base Prediction; Genre&Clustering Base Prediction; Industries; Live Song Popularity Prediction; Machine learning; Music; Predictive models; Training; Visualization
• DOI: 10.1109/ICSCC59169.2023.10334940

This research project aims to develop a machine-learning model that can predict the popularity of a song before its release. Predicting a song's popularity is critical in the music industry for making informed decisions on marketing and promotion strategies, which can increase the chances of success. To achieve this goal, we propose a novel approach that combines genre-based prediction and clustering methods using machine learning algorithms. While previous research has explored the use of genre-based prediction or clustering individually, this combined approach has not been explored yet. Our model shows the highest accuracy in predicting song popularity to date, and the prediction labels for live datasets are reliable and efficient. Our methodology involves feature extraction, clustering of datasets, combining the datasets with various prediction models of genre-based and cluster-based approaches, and using the Extra Tree regressor to further improve accuracy. This innovative approach will revolutionise the music industry by providing composers with an illustrative idea of their song's potential popularity and enabling them to modify their composition to meet trending needs, thereby increasing their song's likability among listeners.