基于二维相关度的嫌疑人社交网络分析方法研究

An Algorithm of Two Dimensional Correlators for Information Mining from Suspect's Social Network

构建和分析嫌疑人的社交网络结构有助于深入研究嫌疑人信息传播的规律,从而获取更多嫌疑人作案线索。现有的社交网络的相关度设计多忽略数据属性,本文将数据属性引入相关度中,并改进相关度数量值计算模型,提出了一种适用于手机实联系数据,通过数据属性与数据数量二维度相结合来描述手机拥有人与手机中存在的联系人间的相关程度的模型,我们将其定义为二维相关度模型。最后结合数据可视化技术分析嫌疑人社交网络。研究表明此方法能够有效的反映嫌疑人与联系人之间的相关程度,形象直观地呈现社交网络图,并能更有效的挖掘嫌疑人社交网络的隐含信息,更有利于工作人员后期工作的展开。该研究为分析嫌疑人社交网络提供了新思路,具有一定的现实意义。

The structure of suspect＇s social network is helpful to analyze the regularity of his/her information transmission,thus more clues about him/her can be obtained.However,most of the existing social network performers relating to correlativity ignore the attribute type of data.Therefore,an algorithm of two dimensional correlators was put forward with inclusion of the data attribute so as to improve the calculation of correlated data.A model was set up to describe the correlating extent between the owner of one mobile phone and the contacts kept in the phone when the data dimensions of both attribute and quantity were combined.This model can be used to orient real contacts through mobile-phone-storing data like short messages,callings and name list of communication.Finally,with the visualization devices introduced,a suspect＇s social network can be visually revealed.Validity test showed that this manipulation can effectively measure the social intercourse between a suspect and his/her contacts by one intuitionistic social graph so that the implied information of the involved person will be further mined from his/her social network,more conducive for the following work to deploy.As a new idea to analyze suspect＇s social network,the method attempted here certainly holds its practical significance.