一种新的犯罪团伙挖掘算法

A New Criminal Gang Mining Algorithm

为了利用图模式挖掘犯罪情报网络中的核心团伙和核心人物,提高犯罪网络威胁预测和识别的效率,提出一种新的核心团伙挖掘算法(Core Gang Mining Algorithm,CGMA).对海量的犯罪情报网络数据集建立相应的无向简单图模型,通过改进图挖掘方式,构建候选核心团伙集的数据结构,并提出由k-团伙通过连接和扩展2种操作得到(k+1)-团伙,从各个不同的图数据中统计其频度,最后在模拟数据集和真实数据集上验证算法CGMA的准确性和时间复杂度.该算法避免了传统的图模式挖掘中的子图同构问题,同时也优于其他常用的犯罪团伙挖掘算法.试验结果表明:该算法能对犯罪核心团伙信息进行有效预测.

In order to utilize graph patterns to mine core gangs and key characters in criminal intelligence networks and improve the efficiency of threat prediction and identification in criminal networks,a new core gang mining algorithm(CGMA)is proposed.Data set based on vast amounts of criminal intelligence network,the undirected simple graph model is established,by improving the mining method and building candidate core gangs set data structure.It obtains the k-gangs by join and extension of two(k+1)-gang from different graph data statistics of the frequency.Finally,the accuracy and time complexity of the algorithm CGMA are verified on the real data set,the algorithm avoids the problem of subgraph isomorphism in traditional graph pattern mining,which is better than other common mining algorithms of criminal gangs.Experimental results show that the algorithm can effectively predict the information of criminal core groups.