The Research On Methods For Detecting Dynamic Change Of Covert Network Based On ARIMA Model

Detecting dynamic change of covert network is an important direction in the research of counter-terrorism or anti-crime.It is an important research on anomaly monitoring of parameters in covert network based on dynamic time series.The monitoring of dynamic change detecting in covert network is an important part of early warning in anti-crime.The results of monitoring can provide concrete method support and evaluation reference for the warning of crime actions.In this study,we will establish time-series network,and extracting the key subgroups,to calculate the parameters of each model.Then,we introduce the ARIMA to build model based on the parameters.Finally,the purpose of established model is to monitor the dynamic changes of cover network.The main contents of the paper include:1.First,we proposed internal weighted algorithm in calculating parameters of covert network and analyzed the characteristics of changes in time series.Based on the traditional social network analysis,the proposed algorithm improves the effectiveness by simple average.And then we expand the static analysis to dynamic detecting.By analysis the changing characteristics of parameters in covert networks,we can provide theoretical and methodological support for detecting the dynamic changes of covert network.2.Second,we develop the technique for change detection in covert network.Based on the principle of ARIMA,the feasibility and constraint condition of ARIMA model in the dynamic monitoring of hidden network are verified and clarified by the mathematically matured method of ARIMA.The ARIMA model is used to establish the differential autoregressive moving average model,that is,to determine the value of the model parameters(p,d,q),and to calculate the key subgroups of the covert organization network by means of the feature extraction of the subgroup key subgroup parameters of the covert organization time series.The results of different monitoring methods such as traditional statistical process control(SPC)method are compared and the time points of abnormal change of terrorist network are identified and analyzed.3.The third is Al-Qaeda network dynamic monitoring empirical analysis.First,we construct the Al-Qaeda event network with event as the unit.There will be a process,that is,time series,for each event from planning to maturity.The parameters of the key subgroups of the time-series network are calculated and analyzed.Finally,the comprehensive indexes of the time-series are obtained by comprehensive calculation.Based on this,the ARIMA model is used to model the comprehensive parameters,and the dynamic model of the organization network is monitored.The feasibility of the monitoring method is verified by the analysis of the monitoring results.4.Finally,we work at analysis of monitoring results and counter-terrorism,anti-crime measures recommendations.Based on the results of empirical analysis and the characteristics of the methods and the characteristics of the current anti-terrorism and anti-crime situation in China,this paper puts forward some feasible countermeasures and suggestions for the monitoring and early-warning of terrorist and criminal networks in China.