基于协同过滤和标签的混合音乐推荐算法研究

Research on Hybrid Music Recommendation Algorithm based on Collaborative Filtering and Tags

结合音乐这一特定的推荐对象,针对传统单一的推荐算法不能有效解决音乐推荐中的准确度问题,提出一种协同过滤技术和标签相结合的音乐推荐算法。该算法先通过协同过滤技术确定相似用户,再通过相似用户对某一歌手的标签评分预测另一用户对该歌手的偏好程度,从而选择更符合用户喜好的音乐进行推荐,以此提升个性化推荐效率,为优化音乐推荐系统提供参考方法。

Traditional single recommendation algorithm cannot effectively solve the accuracy problem in music recommendation.In view of music,a specific recommendation object,this paper proposes a music recommendation algorithm combining collaborative filtering technology and tags.First,collaborative filtering technology is used to identify similar users.Then,another user’s preference for a singer is predicted through similar users’tag ratings for the singer.Thus,recommended music is more in line with the user’s preference,which enhances personalized recommendation efficiency and provides a reference method for optimizing music recommendation system.