A SURVEY ON EMOTION BASED MUSIC RECOMMENDED SYSTEM USING MACHINE LEARNING

Abstract :

This design aims to produce a recommendation system that utilizes facial expressions as a base for suggesting reality like songs to druggies. Facial expression recognition technology will be employed to descry and classify druggies' feelings, mapping them to standard emotional countries. The term" recommender system" refers to a system that may be used to suggest products to a stoner grounded on information or criteria similar as previous stoner commentary or other stoner patterns. This design is aimed at not using the stoner past feedbacks or other pattern, it'll rather use the stoner’s facial expression to recommend druggies like songs. thus, it's important to develop a recommendation system that uses lower stoner data while still performing well since a stoner's requirements couldn't be tied to his or her history but rather to the present, which is denoted by the stoner's expressions. The success of this recommendation system will depend on its capability to acclimatize to individual and artistic variations in facial expressions while furnishing real- time, sequestration-conscious, and ethical recommendations. The algorithm used in this systems are CNN( Convulotion neural network), KNN( K- nearest neighbourhood algorithm). Content grounded filtering and cooperative filtering are used in recommendation system.