Mood-Based Music Discovery: A System for Generating Personalized Thai Music Playlists Using Emotion Analysis

Abstract

This study enhances the music-listening experience and promotes Thai artists. It provides users easy access to Thai songs that match their current moods and situations, making their music journey more enjoyable. The system analyzes users’ emotions through text input, such as typing their current feelings, and processes this information using machine learning to create a playlist that resonates with their feelings. This study focuses on building a tool that caters to the preferences of Thai music listeners and encourages the consumption of a wider variety of Thai songs beyond popular trends. This study develops a tool that successfully creates personalized playlists by analyzing the listener’s emotions. Phrase and keyword recognition detect the listener’s emotions, generating playlists tailored to their feelings, thus improving their music-listening satisfaction. The classifiers employed in this study achieved the following accuracies: random forest (0.94), XGBoost (0.89), decision tree (0.85), logistic regression (0.79), and support vector machine (SVM) (0.78).