Gml Spatial And Temporal Clustering Mining

GML(Geography Markup Language)is an encoding standards of geography information under network environment which is widely used in various areas with the development of computer technique , network technique and database technology. Also, a large number of temporal data in GML format emerged with the expansion of the market of location based services. A series of problems are produced through convenience are brought. The most prominent problem is information overload, the low use ratio and the process beyond the people's capacity. How to extract the hidden knowledge from the mass of GML data or database has become the advanced and challenging topic and GML data mining comes into being.Traditional data mining techniques is about structured data and can't solve the GML data which is vriational and hierarchical. The paper focuses on clustering mining are as the followings:Firstly, the theory of spatial data mining, XML data mining and GML theory is expounded in detail. The key technology of GML clustering mining such as analysis, visualization, the methods and the procedures of mining, the assessment of clustering quality and so on are studied.Secondly, architecture of GML space-time clustering mining is designed. The architecture is divided into data sources, data mining and user interface. The data source is the GML temporal data document which describes the route at different times. The digger is used to completing the task according to the number of cluster which is supplied by the user. The user interface is the intermediary of human and the computer. The tasks the user deliverd are disposed by the computer and then the computer deliver the result in the way of graph, text and so on.Thirdly, a clustering algorithm called L-Kmeans combined the classic clustering algorithm and the expand query language of XML document is proposed. The algorithm is an effective solution to the GML clustering mining and the validity and practicability are proved with the experiments.Fourthly, a system of GML clustering mining which including GML data analysis, visualization, clustering mining and other aspects is achieved with the component technology and the development environment of .NET.In brief, the study on GML clustering mining has a strong significance of practical and theoretical. It plays a key role in directing the spatial distribution of the spatial objects. On the other hand, it will further enrich and improve the data mining theory and technology system.