关键词: LDAMURE模型, 推荐算法, 协同过滤, Gibbs抽样, LDA模型

Abstract: Aiming at the lack of personalized music recommendation and collaborative filtering method based on music information, t hrough the analysis of the user’s listening to music data and download data, co mbined with LDA(latent Dirichlet allocation) theme mining model, we proposed a recommendation algorithm based on the LDAMURE model. Experimental results show that, compared with collabora tive filtering algorithm based on user of music works and collaborative filterin g algorithm based on music attribute item, the LDAMURE algorithm can be more effecti ve to music users recommend music works of interest.

Key words: LDA-MURE model； recommendation algorithm, collaborative filtering; , Gibbs sampling, LDA model