Music Genre Classifier and Recommender
1Dr. Geetha Arjunan, 2Bammidi Lakshmi Deekshitha, 3Bayamgari Shaik Sayouf,
4C Bala Gangadhar Reddy, 5Kakani Sri Anshu Reddy
1Associate Professor, 2Student BCA,3Student BCA,4Student BCA, 5Student BCA,
School of Computer Science and Engineering and Information Science,
Presidency University, Bengaluru, Karnataka, India
________________________________________________________________________________________________________
Abstract: Reclassification of music genres plays an important role in organizations and is a recommendation for audio content in large music collections. Traditionally, manual listening and tagging was required for this genre classification, which was time-consuming and subjective. To improve this, we propose an automated system that categorizes music genres and provides personalized recommendations based on user preferences. Our system extracts related features from audio files and uses machine learning models to predict the corresponding music genre. Additionally, the categorized genres are integrated into recommended frames suggesting similar tracks and related events for users. The flask-based web interface supports user registration, admin registration, uploading audio files, and visualizing recommendations. This project works efficiently with commodity hardware and uses CSV and MySQL databases for data management. This approach highlights the user-friendly, scalability, and application possibilities for real-world application, particularly for music streaming platforms and event recommendation services. Experimental results provide accurate genre classification and user-specific recommendations. This improves the typical user experience.
IndexTerms - Music Genre Classification, Machine Learning, Audio Feature Extraction, Recommendation System, Music Information Retrieval, Flask Web Application, User Personalization, Audio Processing.
_______________________________________________________________________________________________________
