融合社交网络特征的协同过滤推荐算法  

Collaborative Filtering Recommendation Algorithm Based on Characteristics of Social Network

作　　者：郭宁宁 王宝亮[1] 侯永宏[1] 常鹏 

机构地区：[1]天津大学电子信息工程学院,天津300072 [2]天津大学信息与网络中心,天津300072

出　　处：《计算机科学与探索》2018年第2期208-217,共10页Journal of Frontiers of Computer Science and Technology

基　　金：国家自然科学基金No.61571325~~

摘　　要：为了解决传统协同过滤算法中存在的严峻的数据稀疏性问题,提出了一种融合社交网络特征的协同过滤推荐算法。该算法在传统矩阵分解模型基础上,通过融合社交网络特征与用户评分偏好程度得到信任和被信任特征矩阵,然后利用社交特征矩阵、商品特征矩阵和用户评分偏好相似性共同预测用户对商品的评分值。为了验证该算法的可靠性,使用Epinions公开数据集对算法性能进行对比分析。实验结果显示,相比现有的社交推荐算法,所提算法有更小的平均绝对误差和均方根误差,同时算法的时间复杂度与数据集的数量之间为线性关系。因此,该算法可以有效缓解数据稀疏性对推荐结果的影响,并提高推荐准确率。在现实推荐中,该算法可以考虑作为大规模数据集进行商品推荐的一个选择方式。To solve the severe sparseness problem of traditional collaborative filtering recommendation algorithm,this paper proposes a novel collaborative filtering recommendation algorithm based on the characteristics of social network. On the basis of traditional matrix decomposition model, the algorithm obtains the trust and trusted characteristic matrix by integrating the characteristics of social network and users preference degree, and then, predicts the rating of the commodity by the social identity matrix, the commodity characteristic matrix and the user rating preference similarity in common. In order to verify the reliability of the proposed algorithm, this paper uses the Epinions open dataset to compare the algorithm performance. The experimental results show that compared with the existing social recommendation algorithms, the proposed algorithm has smaller average absolute error and root mean square error. Meanwhile, there is a linear relationship between the time complexity of the proposed algorithm and the number of the dataset. Therefore, the proposed algorithm can effectively reduce the impact of data sparseness on recommendation results and improve the recommendation accuracy rate. In practice, the proposed algorithm can be considered as an alternative and development of the large-scale data set recommendation.

关 键 词：推荐系统 社交网络 协同过滤 用户评分偏好 评分预测 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]