The Application Of Data Mining In Household Registration Management

Population management in China has experienced the great changes from the manual management to the mono-computer management system later, and then to the network management today. It is obvious that the management level and methods have gained considerable improvement, but due to times development; the public security requirements for household registration are higher and higher. This complex work can not be done all by a simple query. At present all the analysis methods on data are directly dependent on experience. All the research by experts and the integration of all the data are only concluded by the method of artificial analysis and expert system, as looking for a needle in the ocean, through which some clues and valuable information maybe discovered. While this process, to some extent, just verifies some of our known knowledge. But the unknown, which has great potential value of "knowledge", is still a mystery for us.When Data mining works, the useful model and some potential, previous unknown available information will be found. Data mining is applied to make a reasonable and further analysis, and then found some useful clue, so that a further clue will be provided to practical units and they will discover the relation among different matters; at the same time population management level will be enhanced and offer more valuable assistance and support for the decision makers.The thesis will have an analysis on the practical situation of population and household register. Select data mining tool to analyze population data. A new judging module Miner would be developed on JAVA platform by improved K-means method; In this module K-means method is improved by introducing power concept and circular value. The cluster made by improved K-means method is much closer to practical situation than original K-means method.