Abstract: Although community detection can deeply reveal the underlying structural characteristics of the network and the relationships between nodes, it also raises privacy leakage issues. The community hiding methods can effectively counteract community detection algorithms, thereby achieving privacy protection of network nodes information. However, most of the traditional community hiding methods focus on the privacy protection of a single target or community in the network, there is still a lack of a method in current research that can effectively hide any arbitrary target set. In order to solve the above problems, a Base on Permanence loss Max multiple target Nodes Hiding (BPMNH) method was proposed, which can freely configure the set of target nodes to be hidden, and adaptively provide a permanence loss maximization scheme according to the network scale. Moreover, BPMNH method can achieve hiding of multiple target nodes in different communities with minimal network topology disturbance cost. Experimental results show that the proposed method can effectively achieve multiple target nodes community hiding.

Key words: community hiding, community detection, multiple target nodes, permanence, complex network

摘要： 社区检测尽管能深度揭示网络潜在的结构特征以及节点之间关系，但也产生了隐私泄露问题。社区隐藏方法能够有效对抗社区检测算法，从而实现网络节点信息的隐私保护。但传统的社区隐藏方法大多关注网络中单一目标或单一社区的隐私保护，当前研究中仍然缺乏一种能够针对任意目标集合进行隐藏的方法。针对上述问题，提出一种持久性损失最大化的多目标节点隐藏（BPMNH）方法，该方法可以自由配置拟隐藏的目标节点集合，并根据网络规模自适应地给出持久性损失最大化方案，在最小的网络拓扑扰动代价下，实现不同社区的多个目标节点隐藏。实验结果表明，该方法能够有效实现多目标节点社区隐藏。

关键词: 社区隐藏, 社区检测, 多目标节点, 持久性, 复杂网络