一种解决协同过滤数据稀疏性问题的方法

面对信息量过载的问题,为了使用户尽快的从大量的数据中找到自己需要的信息,即运用协同过滤算法解决数据稀疏性问题,本文提出了一种基于径向基函数(RBF)神经网络的解决方法.首先构建RBF神经网络并提出了一种新的确定隐层节点方法.然后利用构建的RBF神经网络预测用户评价矩阵中的空缺值,提高用户相似度计算的准确性.最后通过与经典协同过滤算法的对比实验证明所提算法的实用性,实验结果表明,基于RBF神经网络的协同过滤算法可以有效的解决用户评分数据的稀疏性问题,提高推荐的准确度.

In Collaborative Filtering a Method of Alleviating the Sparsity Problem

This paper proposed a method using radial basis function(RBF) neural network to solve the sparsity problem in collaborative filtering.The RBF neural network was constructed using a new method.The missing data in users’ evaluation matrix was predicted by the RBF neural network,and the accuracy of the user similarity calculations was improved.The experimental results demonstrate that the proposed method can well-targeted recommend items for users compared with the classical collaborative filtering algorithm,and it can effectively alleviate the sparsity problem in collaborative filtering.