A MUSIC BASED MOOD REGULATION SYSTEM USING SENTIMENT ANALYSIS, RUSSELL'S CIRCUMPLEX MODEL AND VECTOR DISTANCE CALCULATION TO IMPROVE THE PRODUCTIVITY OF THE USER
Aadil Fayas P., Alen Lawrance, Adarsh K. R., Abish B., Satheesh Kumar D.*
ABSTRACT
It has been recognized that music plays a significant role in altering and regulating the mood and emotion of a person. The various Music Information Retrieval models available today rarely factor in the user?s current mood based constraints. Making a model that sorts music with respect to these parameters can be utilized in many applications. With this realization, we have decided to implement a model and use it to build a software application that allows users to seamlessly retrieve music suggestions based on their mood and emotional state in order to help them reach their desired positive goal state. The underlying NLP-based Interactive Agent will map the user interaction data with a sentiment analysis model developed using the Russell?s Circumplex Model of Emotion developed by James Russell which suggests that emotions are distributed in a two-dimensional circular space, containing valence and arousal dimensions. This then is used as an input to the Valence-Arousal based song database we created earlier to find the best matching songs to enqueue for the user, and take them gradually to improve their mental state and reach the goal state.
[Full Text Article] [Certificate Download]
