| Summary: | 碩士 === 中央警察大學 === 資訊管理研究所 === 106 === Along with the development of communication technology, people could hardly get rid of using mobile communication. Because of the basic technical principle of mobile communication involves the connection with the base station nearby, the maxim “Every contact leaves a trace” become the characteristic of communication technology.
The crime investigation almost depends on the video surveillance systems nowadays, but the set up of these video surveillance systems comes to some problems, such as position、coverage、angle、image resolutions and the costs. Consequently, lots of criminal case could not arrest the suspect by survey and investigate the video surveillance system.
However, consider the phenomenon of most criminal behavior committed by minority and the geographic characteristics of mobile communications. This study purpose to use the communication database in the specific time and place for Big Data analysis, to find the core criminal population and its relationships, and save the result of analysis as a database, it will become a assistant tool of crime investigation. Once a crime committed, in the absence of evidences or related traces, the database could become a well source for the investigation, and it would be a useful tool to assist and enhanced the traditional crime investigation method.
We utilize the WEKA data mining software as analysis tool in this study, We take the user database of mobile communication as sources, after the data preprocess, clustering the different levels of data sets by Farthest first traversal algorithm, and try to fine the clustering phenomenon of the holders. After the experiments, compare the clustering results between different parameters, and the result can be applied in crime investigation and prevention.
|
|---|
