Research On Criminal Cases Association Analysis And Crime Prevention And Control Policing Mode

With the quick development of society's politics, economy and technology, the social security situation has become more complex. It is necessary for both the public security organization and relevant scholars to study the regular patterns of criminal cases in such new situations so as to improve the preventive and regulating power most effectively and make full use of the limited police resources. Only by doing this can the security of people's life and production be guaranteed and a harmonious society be established. To explore the regular patterns of criminal cases in order to distribute policing force scientifically and efficiently and to prevent crimes before hand and attack crimes precisely, the dissertation has done the following work:1. Based on demarcating criminal cases data and setting the principles for categorizing them, these data have been sorted out and their characteristics analyzed and problems in their use summed up.2. Criminal cases ontology is established in order to best present the various indexes and factors in criminal cases and to support the multi-sourced and different-structured data integration and deeper data mining and analysis. Based on the ontology established, interviews are carried out at two stages of leaders and policemen at the local stations. The results of the interviews help further clarify the general situation and the existing problems, especially the indexes which need more attention. At last, an ontology-based integration method is proposed to do further analysis and data mining.3. Based on the attribute-of relationship among different categories in the criminal cases ontology, a multi-dimensional associative rule model is put forward together with its algorithm. A multi-layered associative analysis model is presented together with its algorithm based on the kind-of relationship in the ontology. An associative rule filtering method based on semantic rules is proposed in order to solve the problems of redundancy or being too far from reality. The above models and methods have been applied to the associative rule mining and data analysis in criminal cases of Dalian city.4. Based on the analysis of the existing policing models, a new model of crime prevention and control is established with the constructed data ontology and analysis indexes aided by statistical analysis and data mining methods. In applying the new model to the work of crime prevention and control in Dalian, four new modes of police control are carefully worked out: the "effective patrolling" mode according to the time periods of high crime occurrences, the "key points monitoring" mode according to the places and areas of high crime occurrences, the "precise management" mode for the high risk groups and the "dynamic mode" for crimes involving property loss such as burglary, robbery, cheating etc. The effects resulting from these new modes show that they can actually stop crime growth and improve the rate of cracking criminal cases.