Journal of Computer Applications ›› 2016, Vol. 36 ›› Issue (5): 1279-1283.DOI: 10.11772/j.issn.1001-9081.2016.05.1279

Previous Articles     Next Articles

Academic paper recommendation model based on community partition

HUANG Yonghang1, TANG Yong1, LI Chunying2,1, TANG Zhikang3, LIU Jiwei1   

  1. 1. School of Computer, South China Normal University, Guangzhou Guangdong 510631, China;
    2. Computer Network Center, Guangdong Polytechnic Normal University, Guangzhou Guangdong 510665, China;
    3. School of Computer, Guangdong Polytechnic Normal University, Guangzhou Guangdong 510655, China
  • Received:2015-10-29 Revised:2015-12-28 Online:2016-05-10 Published:2016-05-09
  • Supported by:
    This work is partially supported by the National High Technology Research and Development Program (863 Program) of China (2013AA01A212), the National Natural Science Foundation of China (61272067,61502180), the Natural Science Foundation of Guangdong Province (2015A030310509,2014A030310238), the Science Foundation of Guangzhou City (2014J4300033).

基于社区划分的学术论文推荐模型

黄泳航1, 汤庸1, 李春英2,1, 汤志康3, 刘继伟1   

  1. 1. 华南师范大学 计算机学院, 广州 510631;
    2. 广东技术师范学院 计算机网络中心, 广州 510655;
    3. 广东技术师范学院 计算机学院, 广州 510655
  • 通讯作者: 汤庸
  • 作者简介:黄泳航(1992-),男,广东佛山人,硕士研究生,主要研究方向:社区发现、信息搜索、数据挖掘;汤庸(1964-),男,湖南张家界人,教授,博士生导师,博士,CCF杰出会员,主要研究方向:信息搜索、数据挖掘、协同计算、移动互联网应用;李春英(1978-),女,黑龙江齐齐哈尔人,副教授,博士研究生,CCF会员,主要研究方向:社交网络、大数据应用、服务计算、社区发现。
  • 基金资助:
    国家863计划重大项目(2013AA01A212);国家自然科学基金资助项目(61272067,61502180);广东省自然科学基金资助项目(2015A030310509,2014A030310238);广州市科技计划项目(2014J4300033)。

Abstract: An academic paper recommendation model based on community partition was proposed according to sociability in social network. The model regarded the largest connected component in complex network as the logic unit in data processing, and divided up the largest connected component into non-intersect kernel sub-network. The labels would be established according to kernel sub-network by non-parameter control mode. Communities were divided in scholar social network through label propagation, and academic papers were recommended among the users in the communities by the results of the community partition. The proposed community partition method was compared with the classic community partition method in the experiments on artificial network. The experimental results show that the proposed method can achieve good community partition qualities on different characteristic artificial networks.

Key words: kernel sub-network, community partition, label propagation, self-adaptive threshold, academic paper recommendation

摘要: 针对学术社交网络独有的社交性,构建了基于社区划分的学术论文推荐模型。模型选择社区复杂好友关系网络图中最大连通分量作为数据处理逻辑单元,在此基础上进行核心关系网划分,并采用非参数控制的方式,在所建立的核心关系网内建立标签,在学术社交网络中通过标签传播进行社区划分,根据社区划分结果在社区内部的用户之间推荐学术论文。该社区划分算法与经典社区划分算法在人工网络上进行仿真实验,结果表明该算法在不同特征的人工网络上皆能取得良好的社区发现质量。

关键词: 核心关系网, 社区划分, 标签传播, 自适应阈值, 学术论文推荐

CLC Number: