GML Temporal Clustering And Temporal Sequence Similarity Search Key Research Questions

With the rapid development of modern information technology, GIS is an importantcomponent of modern information technology. The problem of data sharing and interoperabilitystill exist. This makes the lack of good communication between GIS and exchanges, for which,OGC introduced the GML specification, makes it possible in a variety of GIS as a bridge betweenthe data and realize GIS community in all directions.GML（Geography Markup Language）is an encoding standards of geography informationunder network environment which is widely used in various areas with the development ofcomputer technique, network technique and database technology. Also, a large number oftemporal data in GML format emerged with the expansion of the market of location basedservices. A series of problems are produced through convenience are brought. The mostprominent problem is information overload, the low use ratio and the process beyond the people’scapacity. Traditional data mining techniques for structured data, can’t solve the change, withGML data, hierarchical structure for this, this paper focuses on the research of GML problemspace clustering.Temporal and spatial relationship as the basic frame of reference of all things, the time seriesdata exist widely in real life, and the data show a "geometric" growth. Behind these massive datacontains many valuable information. How to extract knowledge from spatial data, the analysis ofthe results, provides useful suggestions for the decision makers, has become the urgent problemof spatial data mining. The GML time series similarity of queries is valuable space, especially forGML data.In view of the GML spatio-temporal clustering and temporal sequence similar to currentresearch status of inquiry, this article has done the research work mainly in the following aspectsFirstly, In detail introduced the GML spatiotemporal data model. A variety of models forspatio-temporal data analysis is described, and according to the massive data storage on GMLspatio-temporal data model based on HBase.Secondly, The GML spatial clustering algorithm are studied, the classical clusteringalgorithm (partitioning methods, hierarchical methods, density-based algorithm based on themodel, the algorithm based on grid algorithm,), and on the basis of the classical algorithm isproposed based on the K-clustering algorithm of spatial neighborhood relation and spatialclustering algorithm based on GML space the neighborhood, the corresponding algorithmverification respectively on the experiments, the K-means clustering algorithm of spatialneighborhood relation of regional economic development, spatial correlation verification of spatial clustering of regional economic development and regional economic development analysisof spatial clustering analysis.Thirdly, Research on similar queries on GML time series is done in-depth study, especiallythe study of query similarity GML time sequence space based on adjacent relation, the mainlandof China,31provinces and municipalities from1997to2012a total of16years of nationaleconomic statistics, respectively for GDP1per, GDP2per and GDP3per before similarity themeasure to standardization, analysis to reflect the level of regional economic development, reflectthe regional industrial structure three.