| 摘 要: 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。 |
| 关键词: 协同过滤；标签；音乐推荐；推荐系统 |
|
中图分类号: TP312
文献标识码: A
|
| 基金项目: 2020年度辽宁省社会科学规划基金项目“健全重大舆情和突发事件舆论引导机制研究”(L20BXW003)；2020年辽宁省教育厅科学研究经费项目“东北地区人口流动与经济发展协同演化机制研究”(SYDR202014). |
|
| Research on Hybrid Music Recommendation Algorithm based on Collaborative Filtering and Tags |
|
HUANG Chuanlin, LU Yanxia
|
(School of Information and Business Management, Dalian Neusoft University of Information, Dalian 116023, China )
huangchuanlin@neusoft.edu.cn; luyanxia@neusoft.edu.cn
|
| Abstract: 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. |
| Keywords: collaborative filtering; tags; music recommendation; recommendation system |
