基于Pregel的分布式保护节点影响力匿名算法  

Distributed protection node influence anonymity algorithm based on Pregel

作　　者：李健 张晓琳[1] 刘娇 高鹭[1] 张换香[1,2] Li Jian;Zhang Xiaolin;Liu Jiao;Gao Lu;Zhang Huanxiang(Dept.of Information Engineering,Inner Mongolia University of Science&Technology,Baotou Inner Mongolia 014010,China;Dept.of Computer Engineering&Science,Shanghai University,Shanghai 200000,China)

机构地区：[1]内蒙古科技大学信息工程学院,内蒙古包头014010 [2]上海大学计算机工程与科学学院,上海200000

出　　处：《计算机应用研究》2020年第11期3428-3432,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(61562065);内蒙古自然科学基金资助项目(2019MS06001)。

摘　　要：针对当前社会网络图数据规模不断增加,现有匿名算法大多只考虑匿名隐私强度,忽略匿名后节点影响力变化的问题进行了研究。基于Pregel模型提出分布式保护节点影响力的匿名算法(anonymous protecting influence of nodes,APIN)。算法分解社会网络图得到k-核图,核数代表节点影响力,分裂节点匿名的同时保证原节点核数不变,从而保证节点影响力不变。为了提高APIN算法隐私保护强度,针对社区结构提出保护社区中节点影响力的社会网络匿名算法(anonymous protecting influence of nodes in community,APINC),基本思想是在社区中实现δ-shell安全分组,从而达到δ-核匿名。在真实社会网络数据实验表明,所提出的算法在保持节点影响力的同时很好地保护了图结构性质;最后展望了下一步研究方向。With the rapid increment of the size of the graph in the social network,most of the existing anonymous algorithms only consider the anonymous privacy intensity,and ignore the change of node influence after anonymity.This paper proposed a distributed APIN.The algorithm decomposed the social network graph to obtain the k-core graph.The core number represented the influence of the node.The split node anonymity also ensured that the node core number in the original graph was unchanged,thus ensuring that the node’s influence was unchanged.In order to improve the privacy protection strength of the APIN algorithm,this paper proposed an APINC for the community structure.The basic idea was to implementδ-shell security grouping in the community to achieveδ-core anonymity.Experiments on real social network data show that the proposed algorithm can protect the structure of the graph well while maintaining the influence of the node.Finally,the paper forwarded to the next research direction.

关 键 词：社会网络 影响力 Pregel k-核 社区结构 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]