Research And Application Of Data Mining Techniques In E-Government Based-on ODS

With development of e-Government in China,the information resources of e-Government are increasing rapidly.However,Compare to the fast-growth of these information resources,less helpful knowledge we can get from.It becomes an obstacle to e-Government upgrade from data management-based and information management-based to knowledge management-based.How to make use of these huge and complicate e-Government information resources efficiently and supply more knowledge to Government daily management is becoming an important and valuable research theme in our country's information industry.As a very crucial technology of today's information field,the decision support technology can provide the decision maker more helpful and valuable knowledge from e-Government information resources to make a better decision.By now,decision support technology is widely applied in many fields,however,in e-Government,it is not so widespread applied,so it has a broad research and develop prospect.In this paper,according to the features of e-Government information resources and government structure,a decision support platform which combines DB(database), ODS(operational data store),DW(data warehouse) three levels data storage solution and the OLAM(online analysis data mining) information and knowledge discovery technology is proposed. Then,based-on data mining algorithm component theory,a flexible framework of data mining algorithms construction is proposed to improve the OLAM technology.First,for the analysis conclusion of recently decision support system,which are all commonly associate with data warehouse,data mining and online analysis process technologies, and the defect of DB-DW conventional two level data storage framework,an appropriate customized data storage framework which is comprised of DB,ODS and DW is proposed to solve the problems of complicity,redundant and data nonconformity in the management and sharing of e-government information resources.Based-on ODS,OLAM technology is applied to provides the detail data analysis and real-time decision support for the government.Then based on this data storage framework and OLAM technology,a method of using components of some typical data mining algorithms to build-up a customized data mining algorithm is proposed to improve the OLAM technology,and also can provide flexible and fast decision support in different circumstances.Finally,an application is developed to implement this analysis platform,which includes three modules,OLAP data analysis and manipulate modules,data mining algorithm constructions modules and data visualization modules.Though an analysis of the water quality monitoring data, this platform shows more flexible for information management and knowledge discovery,which provide e-government a new development direction.