Abstract:
Sentiment analysis represents a contemporary approach in Natural Language Processing used to determine the sentiment of a user. Bangla music, with its unique rhythms, melodies, and lyrical depth, stands as a musical treasure that reflects the soul of the Bengali culture. This music serves as a powerful storyteller, narrating tales of love, longing, joy, and resilience. From the soulful tunes of Rabindra Sangeet to the electrifying beats of contemporary Bangla pop, this music genre captures the essence of Bengali life and spirit. It transcends borders, touching the hearts of listeners worldwide with its timeless beauty and profound lyrics, making it a cherished part of global music heritage. This research study has attempted to investigate on Bangla music using different machine learning classification algorithms from over 2224 data points. People of all ages are considered as data collection. After pre-processing and feature engineering the collected data, Random Forest, Decision Tree, Multinomial Naive Bayes, XGBoost, K-Nearest Neighbor and Support Vector Machine are used to train the model. These classifiers predict with high accuracy, with Random Forest having the highest accuracy of 63.68%.
