Personalized language-independent music recommendation system
N. Radhika, Syam Sankar · 2017
Social media is a place where users present themselves and share all type of information, ideas, and experiences with world. In recent years sentiment analysis has been explored by several internet services to recommend contents accordance with human emotions, which expressed through social network. Many user's on social network write sentence in variety of languages. This paper presents language independent sentiment analysis in music recommendation system. This system suggests music to user depending upon the current emotional state of that person. Where the current emotional state of a person is calculated by measuring adjusted sentiment intensity value, which is an association of sentiment intensity value with a correction factor. This correction factor is based on user profile information and used to adjust final sentiment intensity value. Traditional classifiers are language specific and require much work to be applied to different languages. Proposed system uses a supervised sentiment classification approach for language independent sentiment intensity calculation. We train Naive bays classifier on our data and evaluate it on over 1000 posts in 2 languages, English and Malayalam. The supervised sentiment classifier with a correction factor can improve the performance of a proposed language independent music recommendation system.