【摘要】 通过消息监控识别罪犯是一个十分有意义的实际问题。采用Markov模型,将整个消息传递网络看作一个犯罪传递的Markov链,根据所收集到的消息估计出两个节点(人)之间犯罪传递的概率,得到一个Markov概率转移矩阵,并求出网络长期运行的稳定解,作为各节点(人)参与犯罪程度的度量。通过实例说明了该方法的有效性。更多还原

【Abstract】 It’s very important to identify crime through monitoring message.In this paper,based on Markov model,the whole message-transmitting network is regarded as a Markov chain.According to the message,the transfer probability between two nodes is estimated first.Then the transfer matrix of Markov is obtained.The stable solution of the network is solved and the suspicious degree of every person is obtained.At last,the method is demonstrated by a real illustration.更多还原