融合信任传播和混合相似性度量的推荐算法  

Recommendation Algorithm Combining Trust Propagation and Hybrid Similarity Measurement

作　　者：杨磊[1] 刘美枝 YANG Lei;LIU Mei-zhi(School of Physics and Electronic Science,Shanxi Datong University,Datong Shanxi,037009)

机构地区：[1]山西大同大学物理与电子科学学院,山西大同037009

出　　处：《山西大同大学学报（自然科学版）》2022年第4期44-49,共6页Journal of Shanxi Datong University(Natural Science Edition)

基　　金：山西省自然科学基金[201901D211441];山西大同大学科研基金[2020Q4]、[2020CXZ2]。

摘　　要：针对传统推荐算法中存在的冷启动及稀疏性问题,提出一种融合信任传播和混合相似性度量的推荐算法TPHS。首先,在社交网络推荐算法的基础上,融入信任传播机制,计算用户之间的显性和隐性信任度,进一步挖掘用户之间的信任关系;其次,在衡量用户相似性时,采用混合相似性度量,更好地描述用户之间的相似性;最后,综合考虑用户之间的信任关系和相似性关系,采用Top-n方法和相似性阈值法进行预测评分,得出推荐列表。在Epinions数据集和Movielens数据集上的实验结果表明,该文提出的推荐算法在推荐精度和召回率方面具有更好的推荐性能。Aiming at the problems of cold start and sparsity in traditional recommendation algorithms,a recommendation algorithm TPHS,which combines trust propagation and mixed similarity measures,is proposed.First,on the basis of social network recommendation algorithm,the trust propagation mechanism is integrated to calculate the explicit and implicit trust between users,and further excavate the trust relationship between users;secondly,when measuring the similarity of users,the mixed similarity measure is used to better describe the similarity between users;finally,considering the trust relationship and similarity relationship between users,the Top-n method and similarity threshold method are used to predict and score,and the recommendation list is obtained.The experimental results on the Epinions dataset and the Movielens dataset show that the proposed recommendation algorithm has better recommendation performance in terms of recommendation accuracy and recall.

关 键 词：推荐系统 社交网络 信任关系 信任传播 相似性度量 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]