Construction And Application Research Of Spatio-temporal Data Warehouse Of The Criminal Cases In Fuzhou

With the continuous operation of the public security business system, public security departments have accumulated more and more cases. A great deal of data which is isomerous, debased, incomplete and inaccurate brings difficulties to integrate and share the data of public security.75%-85% of the public security cases which we contact are related to the spatial location. However, the common data warehouse cannot meet the requirements. In this context, this article will introduce spatio-temporal data warehouse technology into the public security system, analyze and map the crime basing on the spatio-temporal data warehouse of cases. In essence, it aims at improving the efficiency and precision of all the police in emergency prevention and decision-making. The main contents of the paper are showed as follows:Firstly, this thesis summarizes the differences and relations of OLAP (On-line Analytical Processing) and OLTP (On-Line Transaction Processing), the advantages and disadvantages of MOLAP (Multidimensional OLAP), ROLAP (Relational OLAP), HOLAP (Hybrid OLAP). Then, it discusses the four methods of crime Spatio-temporal analysis.Secondly, it confirms the themes, measures, and particle size of the cases of spatio-temporal data warehouse. The particle size design uses the design method of double the granularity of the grain size, which designs a multidimensional data analysis in mild synthetic data and real data is most suitable for the public security cases. On this basis, the Fuzhou Public Security cases of 2005-2010 spatial data are taken to be extracted, converted, synthesized, refined, cleaned and loaded. Then the data of pre-processing operation is used to build the conceptual model, logical model, and physical model of the cases of spatio-temporal data warehouse. Moreover, this thesis displays the report and the graph of the data cube.Finally, this thesis takes the Fuzhou Burglary cases in 2010 for example, using kernel density estimation method to analyze and map the crime hotspots. It can reveal the causes and the criminal rules of every county. This result is combined with the thematic map, and compared the crime number with different counties’ level of economic development and their geographic location. It can be seen that the number of crimes is greater in more developed regions. Because the more developed regions have more traffic and people. Moreover, the more developed regions always have a large gap between the rich and the poor than the less developed regions. However, the-gap between rich and poor is smaller in economic conditions of medium area, so the crime number is relatively small. In the end, the thesis employs the text table and dashboard of the different dimensions to visualization analysis in time and space.