Study Of Data Mining Methold Based On Concept Lattice Theory

The concept lattice theory, also named formal concept analysis (FCA), introduced by professor Wille, is based on the philosophy of concept which is made up by two parts of extension and intension. Over the past twenty years, concept lattice has been widely studied and become a powerful tool for machine learning, software engineering and information retrieval. In addition to being a technique for classifying and defining concepts from data, concept lattice can also be exploited to discover implications among the objects and the attributes. Concept lattice and applied it in knowledge discovery have important significance. This thesis mainly deals with the basic theory and technology of knowledge based on concept lattice including construction algorithm and attributes reduction of concept lattices. The main content is as follows:Firstly, concept lattice and rough set have many common characteristics in data analysis. In this paper, the association and their relation between concept lattice and rough set are described, have proved that it can be interconvert between concept and partition.Secondly, ameliorate Bordat algorithm by analyzing it deeply to conquer the question of many repeated nodes appears; suggests a new algorithm based on margin concept which constructs the concept lattice with the hierarchical structure of the lattice through computing the intersection and union of the intent and extent of the already concepts; presents a new definition of the fuzzy concepts of a fuzzy formal context, discusses its properties and gives an algorithm to compute all fuzzy concepts.Finally, give two kinds of arithmetic of attribute reduction by using distinguishable attribute matrix in rough set theory.