Music recommendation system, music recommendation for computer programs and music recommendation methods

摘要

Without re-learning the basic learning unit, in the music recommendation system that can be registered immediately a new user, the new music, by monitoring the evaluation score storage unit 3, incremental learning section 21 of the first, change the evaluation score and whenever there is an addition of a new user, it is updated to remain likelihood defined by the basic learning unit 17, the genre selection probability for the new user and the user has changed the score is maximized or I want to add. By monitoring acoustic feature storage unit 5, each time with the addition of acoustic characteristics by the addition of music, incremental learning unit 21 of the second, defined by a fundamental learning unit 17, a music selection probability for music added Adding to remain likelihood has been maximized.