项目名称: 大规模在线社会网络社区发现及隐私保护研究
项目编号: No.61462003
项目类型: 地区科学基金项目
立项/批准年度: 2015
项目学科: 自动化技术、计算机技术
项目作者: 陈本辉
作者单位: 大理大学
项目金额: 46万元
中文摘要: 大规模在线社会网络分析和隐私保护是当前的研究热点,针对大规模社会网络的问题特征及应用环境,需要对传统社会网络分析方法和隐私保护方法进行发展及改进。本项目拟基于Web数据挖掘技术及智能计算方法进行大规模在线社会网络分析和隐私保护研究,主要包括:(1)大规模在线社会网络的社区发现方法,针对传统社区发现算法处理大规模社会网络效率较低的问题,基于分布估计算法(EDA)、概率模型优化方法、聚类等智能计算方法,引入和设计新的社区挖掘计算方法来获得更好的社区挖掘效果。(2)针对大规模在线社会网络数据发布中的隐私保护问题,研究有效防范去匿名化攻击且适用于大规模网络的高效隐私保护模型。(3)探索个性化隐私保护技术,在大规模在线社会网络隐私保护模型中融合个性化隐私保护技术,研究新的度量函数更好地平衡个性化服务质量和隐私保护的效果,同时探索提高个性化隐私保护算法的效率,使之适应于大规模在线社会网络系统应用环境。
中文关键词: 大规模在线社会网络;社区发现;数据发布;隐私保护;匿名化技术
英文摘要: The community detection and privacy preserving for large-scale online social network are two current research hotspots. Due to the different characteristics and application environments of large-scale online social network, the traditional social network analysis methods and privacy preserving models should be developed and improved. This research aims to develop community detection method and privacy preserving model based on Web data mining technologies and computational intelligence methods. (1) In order to improve the efficiency of community detection model for large-scale online social network, the hybrid methods are explored by integrating some strategies including Estimation of Distribution Algorithm (EDA), probability model based optimization algorithms and clustering methods. (2) Consider to the data publishing privacy preserving for large-scale online social network, the improved privacy protection models are explored to prevent De-Anonymization attacks. (3) In order to deal with the personalized privacy preserving problem on large-scale online social network, some new metric functions and preserving strategies for personalized privacy preseving model are explored to balance the personalized service quality and privacy protection efficiency
英文关键词: Large-scale Online Social Network;Community Detection;Data Publishing;Privacy Preserving;Anonymization Technology