基于RT-RBM协同过滤的图书馆个性化推荐系统的研究

Research on Library Personalized Recommendation System Based on RT-RBM Collaborative Filtering

为了解决目前图书馆推荐系统存在的无法挖掘出读者个性化信息的深层次特征、数据稀疏、“冷启动”等问题,设计实现了基于R-RBM及Top N算法(RT-RBM)的协同过滤的图书馆个性化推荐系统.R-RBM协同过滤推荐算法可以有效解决数据稀疏的问题、提高个性化推荐的精准度.Top N协同过滤推荐算法可以较好地为新用户进行推荐.实验结果表明,RT-RBM协同过滤推荐系统相较传统协同过滤算法具有更好的应用效果.

At present,in order to solve the problems existing in recommendation systems in libraries,such as the inability to dig out the deep-seated characteristics of readers personalized information,data sparsity,and“cold boot”,a recommendation system for library based on R-RBM algorithm and Top N algorithm(RT-RBM)is designed and implemented.The Top N collaborative filtering recommendation algorithm can provide a better recommendation effect for new users.The experimental results show that the RT-RBM collaborative filtering recommendation system rivals thetraditional collaborative filtering algorithm by providing a better recommendation effect.