基于重叠社区发现的社会网络推荐算法研究-《南京师范大学学报》（自然科学版）

文章信息/Info

Title:: Research of Social Recommendation Method Based onOverlapping Community Detection

作者:: 胡云¹; 张舒²; 佘侃侃¹; 李慧³; 4; 施珺³; (1.南京中医药大学信息技术学院,江苏南京 210023)(2.淮海工学院商学院,江苏连云港 222005)(3.淮海工学院计算机工程学院,江苏连云港 222005)(4.江苏省海洋资源开发研究院,江苏连云港 222005)

Author(s):: Hu Yun¹; Zhang Shu²; She Kankan¹; Li Hui³; 4; Shi Jun³; (1.College of Information Technology,Nanjing University of Chinese Medicine,Nanjing 210023,China)(2.Business School,Huaihai Institute of Technology,Lianyungang 222005,China)(3.Department of Computer Science,Huaihai Institute of Technology,Lianyungang 222005,China)(4.Marine Resources Development Institute of Jiangsu,Lianyungang 222005,China)

摘要:: 协同过滤算法已成为用来为用户提供个性化服务以处理海量信息最常用的方法之一. 本文提出一种基于重叠社区发现的社会网络推荐算法,该算法同时考虑了群组用户的兴趣以及他们复杂的内部关系,通过将重叠社区发现算法和基于模型的社会推荐算法进行创新融合,以实现重叠社区的发现、建立,和基于社区的智能推荐. 基于开放数据集,本文设计了一系列相关实验以验证算法的有效性和准确性. 实验结果表明本文提出的算法可以实现高效且准确的社会网络推荐.

Abstract:: Collaborative filtering algorithms have become one of the most popular approaches to provide personalized services for users to deal with large amounts of information. This paper proposes a new social recommendation methods based on overlapping community discovery. The algorithm considers both the interest of the group users and their complex internal relations. In order to achieve the detection,establishment coalition of overlapping community and intelligent recommendation based on community,it innovates and integrates overlapping community discovery algorithm and social recommendation algorithm based on the model. Based on the open data set,this paper designs a series of the related experiments to validate the accuracy and effectiveness of the algorithm. Experimental results show that the proposed algorithm can achieve highly efficient and accurate social network recommendation.