Research And Implement On The Key Technology Of Analysis And Dataming On The Social Basic Data

Data mining is an important way to solve the Pattern Recognition of mass data as well a kind of application oriented technology. In the current information society, it's unprecedented that the leader from all walks of life who make decisions desire for the hidden information in huge amounts of data. The combination of data mining with mass data become more and more popular, and more obvious benefits are obtained. It is limited that the research on the application of data mining technology in the field of national security and public security.And we can not find much actual results or progress on such work, no matter at home or abroad. This paper propose the method of 'loose relationship' and combine it in investigative work with the association rule mining method creatively, and designs a prediction social basic data mining system based on association rule mining algorithm. The combination the requirement in actual investigation work and data mining technology is obtained in this study,and exploration is executed in mass social basic data. Effective experiments are achieved by using data mining to solve actual problem in security work,which is significant in deep utilization of theory and practice based on social basic data in security system.According to the actual needs of the investigation work,this study propose a method to measure the relationship among the records in social basic data by defining 'loose relationship'. Through the method the investigator can distinguish the abnormal part from the numerous 'loose relationship' by setting a threshold to definition some unusual characteristics,and then elaborate the connection between 'loose relationship' and investigation work further.This paper put forward and discuss a way to find out the crowd which has 'loose relationship'obviously implicit in the social basic data.The social basic data mining system based on association rule method mainly consists of three modules: data extraction and preprocessing module,association rule data mining module and loose relationship analysis module. The main function of the data extraction and preprocessing module is extracting and converting data for mining. Then the association rule data mining module complete mining process and organize the mining results. Thus the investigators can analyze the mining results and make judge on the effectiveness and availability based on the results by using the analysis module.The social basic data is very complex and has huge amounts of data, so the mining efficiency according to the specific problems is facing a severe challenge. In this paper,according to the characteristics of mining task, the Apriori Algorithm mining method is improved. The improved method apply item sets combined to each other directly within each transaction to generate candidate set, thus it reduced the generation of invalid candidate sets effectively and improve the efficiency of mining indeed. This paper make excavation experiment on real social basic data in the Data Center of the institute by using this data mining system and confirms the validity of the mining result in real case, and achieve the expectant study purpose.