Emotion Based Music Recommendation System Using Navie Bayes Classification
Along with the rapid extension of digital music system and formats, the searching and managing the music system based on the users preference has been very significant. Inorder to improve the users listening experience based on their music preference a recommender can be used. The number of choices of users also overwhelming, so it requires to filter, prioritize and effectively deliver relevant music based on the users choice. Recommending the music based on the user choice based on location, time of day, music listening history and their emotional state is a tedious task. Recommender system solves the problem by incisiving the large volume of spontenously generated information to provide users with personalized content and services.Though music retrieval system is evolved in recent years, but recommending songs based on the user interests is not evolved so far. In this paper, we proposed a emtion based Muisc Recommendation System (MRS) to find the correleation between the users data and music. To address such issue the proposed work provides a novel framework for recommending the music based on the user interest using Navie Bayes classification algorithm. The probability of an event with strong independence assumption between the features were applied to Navie Bayes algorithm. In this proposed system, the emotional state of a user is taken using the users search history and the results were evaluted with the exisiting recommender system.