Font Size: a A A

The Application Research Of Data Mining In CBR And GIS

Posted on:2004-11-03Degree:MasterType:Thesis
Country:ChinaCandidate:P ZhaoFull Text:PDF
GTID:2168360092486548Subject:Computer applications
Abstract/Summary:PDF Full Text Request
Data Mining is the process of extracting hidden , unknown but potential useful information and knowledge from vast, incomplete, noisy, fuzzy and random datum. Data Mining technology is oriented to application. It not only aims at simple search and query, but also makes a microcosmic and macroscopical statistic , analysis , synthesis and reasoning of datum to tutor the solving of the practical problem, to manage to find out the interrelation of the events, and even to make forecast by using the known datum.The research of AI( Artificial Intelligence) has always been one of the front ofthe computer theory and application research. But the bottleneck of knowledge acquisition cumbers the research progress of AI researcher. CBR( Case Based Reasoning) can solve this problem with better results and is widely applied to various fields of problem solving. The prospect of the application is very well. But it also needs vast work of knowledge acquisition in its own construction. Can we find potential knowledge by using Data Mining technology and reduce the reliance on the field expert? In this thesis, the application of Data Mining in CBR is investigated.GIS( Geographic Information System) was developed from 1960s'. It integrates the data collection , storage , management and analysis. It can describe the information of earth surface (including aerosphere ) and the spatial information of space and geography distribution. With the development of computer technology and social demand, GIS technology is going up and the range of its application are widening continuously. People don't want GIS only to show map simply and to make map automatically. People expect to get more knowledge from it. So DM (Data Mining) and IDSS (Intelligent Decision Support System) are imported into GIS. In this thesis, the application of Data Mining in GIS is investigated.The thesis consists of six chapters. In the first chapter, the thesis makes a summarize of the technology of Data Mining , CBR and GIS, illustrates the foundation and significance of this thesis and puts forward the research direction and emphases.In the second chapter, the thesis introduces some primary concepts of Data Mining, investigates and discusses the key technology of Data Mining, including outlier analysis , clustering and classification. And the main algorithms of Data Mining, which the later chapters of this thesis refer to, are also given in detail.In the third chapter, the thesis introduces some primary concepts and principlesof CBR, investigates and discusses the key technology of CBR.In the fourth chapter, the thesis investigates the application of Data Mining in CBR. Firstly, the main Data Mining technology and methods in CBR are investigated and discussed. Secondly, to meet different requirements of applications, two algorithms of case base construction and one algorithm of case base maintenance are put forward, to which association analysis , outlier analysis . clustering and classification are applied. We implement them and analyze the results. Our experiments show that our algorithms can effectively improve the automation degree of knowledge acquisition and performance of CBR system.In the fifth chapter, the thesis investigates the application of Data Mining in GIS. Firstly, some primary concepts of GIS are introduced. Secondly, the main technology of spatial data analysis are investigated and discussed. And then a system frame of Spatial Data Mining based on Expert System and CBR is put forword. At last, we develop a practical application system of GIS to which Data Mining technology is applied.The last chapter is the summarize of the whole thesis and also makes a prospect of our research.
Keywords/Search Tags:Data Mining, Outlier Analysis, Clustering, Classification, Case Based Reasoning, Geographic Information System
PDF Full Text Request
Related items