计算机工程 ›› 2020, Vol. 46 ›› Issue (8): 72-77,84.doi: 10.19678/j.issn.1000-3428.0054954

• 人工智能与模式识别 • 上一篇    下一篇

一种融合社交关系的矩阵分解推荐模型

吴清春a,b, 贾彩燕a,b   

  1. 北京交通大学 a. 计算机与信息技术学院;b. 交通数据分析与挖掘北京市重点实验室, 北京 100044
  • 收稿日期:2019-05-20 修回日期:2019-07-22 发布日期:2019-07-26
  • 作者简介:吴清春(1993-),男,硕士研究生,主研方向为机器学习、推荐系统;贾彩燕,教授、博士。
  • 基金项目:
    国家自然科学基金(61876016);中央高校基础科研业务费专项资金(2017JBM023)。

A Matrix Factorization Recommendation Model Fusing Social Relationship

WU Qingchuna,b, JIA Caiyana,b   

  1. a. School of Computer and Information Technology;b. Beijing Key Laboratory of Traffic Data Analysis and Mining, Beijing Jiaotong University, Beijing 100044, China
  • Received:2019-05-20 Revised:2019-07-22 Published:2019-07-26

摘要: 推荐系统可以有效解决信息过载问题,为用户提供个性化的推荐服务。然而在评分矩阵稀疏的情况下,仅通过分析用户项目评分矩阵来生成预测结果的传统模型效果较差。针对该问题,利用用户评分以及用户社会信任关系两部分信息计算用户相似度,提出一种融合社交关系的矩阵分解型推荐模型SoRegIM。通过挖掘社交网络中用户的拓扑关系,利用目标用户的直接邻居与间接邻居信息构建加权社会信任网络,在充分使用用户社交关系信息的同时减少冗余的社交噪声。基于公开数据集的实验结果表明,与SoReg、SocialMF等7种经典模型相比,SoRegIM能够有效提高推荐准确性,且对于稀疏数据的提升效果明显。

关键词: 社交网络, 推荐系统, 拓扑关系, 信任网络, 矩阵分解

Abstract: Recommendation systems can effectively solve the problem of information overload and provide personalized recommendation service for users.However,traditional models which generate prediction results only by analyzing user project scoring matrix are not effective in the case of sparse scoring matrix.To address the problem,this paper uses the rating information and social trust relationships of users to calculate user similarity,and on this basis proposes a Matrix Factorization(MF) recommendation model named SoRegIM fusing social relationship.By mining the topological relationship of users in social network,the weighted social trust network is constructed based on the information of direct and indirect neighbors of the target users,so as to reduce the redundant social noise while making full use of the social relationship information of users.Experimental results on the open datasets show that,compared with classical models such as SoRec and SocialMF,SoRegIM has higher recommendation accuracy and demonstrates an obvious improvement on sparse data.

Key words: social network, recommendation system, topological relationship, trust network, Matrix Factorization(MF)

中图分类号: