The Rule Extraction Of Spatial Data Based On Concept Lattice

Data mining is a hot research issue in the filed of artificial intelligence, which is study about how to intelligently extract valuable information and knowledge from large amounts of data. Rough set and concept lattice are two effective methods in data mining analysis, which have obtained wide attention and application in the extraction of association rule mining.Rough set theory is to establish the equivalence of inner classes in the given data and classify the data sheets, which provides a new way of ideas for data mining analysis and an effective mathematical tool for processing the uncertainty problem. Concept lattice combined with the order theory, with its construction is a process of clustering and classification, which is easy for the discussion on the concept of hierarchical based on the data sheets.In recent years, with the development of geographic information systems, spatial data mining bred. As a hot issue in data mining, which is studied how to extract non-displayed stored knowledge, spatial association or other interesting patterns from spatial database.Spatial data mining use the capabilities of GIS to store, manage and analysis spatial data,which transfers spatial database into a similar relational database to extract rules by using spatial database technology, thus provides a powerful tool for the improvement of geographic information systems.For formal context in this paper, the use of the special relationship between the extension and connotation of the concept, combined with down approximation concept in rough set, we proposed a new algorithm for constructing rough concept lattice, using roughness mining reliable knowledge after attribute reduction. In the construction process, judge among the nodes’ properties, which can reduces the time complexity of the algorithm effectively.Practical case analysis demonstrate that by combining attribute reduction and roughness, the algorithm can effectively obtain reliable knowledge, which provides a workable thought and approach for data analysis mining knowledge.In spatial data mining, concept lattice is an effective way to extract association rules in data mining, this paper applies it to spatial databases for the extraction of association rules in spatial data mining. In order to improve the speed of extracting association rules, by comparing the extension of concept to build the concept lattice and introducing the support constraint in the process of building concept lattice, delete the nodes which don’t meet the conditions, thus reducing the rule extraction number of judgment, which have achieved practical application results.