大数据在音乐推荐质量提升中的实践及应用

Big Data Improving Recommendation Quality in Music Applications

基于音乐推荐的应用实例,探索和实现了大数据如何提高推荐质量的过程及方法。提出通过建立基于RFM模型的用户歌曲综合评分体系,在推荐算法中引入项目稀疏度、重叠度、可信度概念作为调整因子,在混合推荐时引入飙升词、内容标签和二次规则过滤等组合方法以解决推荐系统面临的常见问题,为大数据应用提供具体的参考和指导。

The process and methods of big data improved recommendation quality in music applications werefocused. The individual music score system referring RFM model was described. The sparsity and overlapping andreliability was applied to adjust recommendation algorithm, combined soaring words and content labels and filteringrules with mixing recommendation to resolve the common problems of recommendation system.