社交复杂网络中基于无向图和聚类的社区检测关键技术研究（英文）

Research on key technologies of community detection based on undirected graph and clustering in social complex networks

社区检测过程是大数据分析时代的重要挑战之一,特别是在社交复杂网络领域。为了提高社区检测的准确性和效率,提出了一种基于无向图和聚类的社交复杂网络社区检测算法。首先采用了两个新的度量指标以便实现社区检测,即聚类系数和共同的邻居相似性。然后基于高效模块化的概念将社区检测的复杂度减少,并通过平衡二叉树来更新无向图中的边和节点,从而减少了计算的工作量。采用社会网络数据集对提出算法进行了验证分析,实验结果表明:相比其它两种算法,提出算法的运行效率和准确性更高。

The community detection process is one of the major challenges in the era of big data analytics, especially in the area of socially complex networks. In order to improve the accuracy and efficiency of community detection, a social complex network community detection algorithm based on undirected graph and clustering is proposed. Two new metrics were first used to achieve community detection, namely clustering coefficients and common neighbor similarities. Then the complexity of the conceptual community detection is reduced to x based on efficient modularity, and the edges and nodes in the undirected graph are updated by balancing the binary tree, thereby reducing the computational workload. The experimental results show that the proposed algorithm has higher operating efficiency and accuracy than those of the other two algorithms.