The Implementation Of Crime-Related Telecommunication Activities Finding Based On Outlier Data Mining

With the development of modern telecom technology, we have found that there are lots of telecommunication activities associated with crime, and in many crime investigation practices, these information are often very helpful .Many investigators call these information "criminal telecom trace" . The traditional access to obtain these information has been become more and more difficult. In fact, like many other telecommunication activities, while crime-related telecommunication activities happening, they have been recorded in company's database .Finding criminal telecom trace from company's database by data analysis technology, will open up a new way for crime investigation.Data Mining is a well-developed analysis technology of finding useful information from database .According to criminal telecom trace 's features, we have found outlier detecting technique may very useful on probing criminal telecom trace . Having analyzed main outlier-detecting techniques , the paper assigned a new outlier-detecting algorithm called double K-factor outlier-relation mining algorithmBased on crime investigation practices , the paper developed a special method on identifying overall features of telecommunication activities .After programmed double K-factor outlier- relation mining algorithm, the paper then implemented this criminal telecom trace detecting technique on a experimental data set obtained from a company's database through SQL2005 platform. Finally the paper traced criminal telecom trace detecting techniques future research and application prospect.