基于概率矩阵分解的社交网络推荐算法研究

Research on Social Network Recommendation Algorithm Based on Probability Matrix Decomposition

推荐系统已经成为人们在网上寻找自己所需信息的常用工具之一。基于社交网络的推荐方法能够解决传统推荐算法存在的问题,例如新用户的冷启动问题。本文提出了一种基于矩阵分解的并且可以应用于社交网络的新模型。该模型将信任传播机制融入模型中,并使用Epinions.com数据集进行实验。试验结果表明,基于社交网络的新模型在推荐准确度方面相较于传统模型,针对评分较少的新用户所存在的冷启动问题有较好的解决。

The recommendation system has become one of the most common tools for people to find what they need online. The recommendation method based on the social network can solve the problems of traditional recommendation algorithm, such as new user＇s cold start. In this paper, we propose a new model based on matrix decomposition which can be applied to social networks. In the new model, we incorporate the trust propagation mechanism into the model and experiment with the Epinions.com dataset. The experimental results show that the new model based on social network has a better solution to the cold start problem of new users with fewer scores due to the traditional model.