Music Recommendation System Based on Preference Classification of Real-time User Brainwave

In order to predict user-favorite songs, managing user preferences information and genre classification are necessary. In this paper, we propose a preference classification about content based on real-time user brainwave and a music recommendation system based on it. We focused on classifying real-time user preferences by analyzing the user’s brainwaves. The brainwaves are acquired using a wireless consumer Electroencephalography (EEG) device with small-sized pins in order to enhance the system’s usability for mobile devices. The performance of preference classification accuracy is nearly equal as that of one of the best EEG-based preference analyzer, despite the use of a comparatively lesser number of feature dimensions. In our study, a very short feature vector, obtained from low dimensional projection and already developed audio features, is used for music genre classification problem. Keywords— Preference classification; Real time; EEG (Electroencephalography); Music recommendation component;