Research On Attribute Reduction And Constructing Algorithm Of Concept Lattice

Concept lattice is a concept hierarchical structure based on binary relations. It essentially describes the relations between objects and attributes, embodies the unification of intension and extension of concepts, and is an effective tool of data analysis and rule extraction. Rough set is a new valid mathematical theory developed in recent years, which has the ability to deal with imprecise, uncertain, and vague information. It can find valid and potentially useful knowledge in data. The relation between concept lattice and rough set has been focused by many scholars and some algorithms become predigested and improved owing to their efficient combination.The thesis is composed of five parts. The first part introduces the background and development of the research issue in the paper. Then the second part presents the thesis of concept lattice and rough set and the relation of them. Those establish the foundation for studying on the attribute reduction and constructing algorithm of concept lattice. The third part analyses the approach of attribute reduction of concept lattice base on discernible matrixes and presents its algorithm. And another approach is proposed on the basis of formal context. In the fourth part, a constructing concept lattice algorithm based equivalence class in the rough set is put forward. Compared with the incremental structuring algorithm, this algorithm saves much time about finding concept nodes and is more efficient due to the equivalence class. The fifth part summarizes the thesis, and points out the direction of research on concept lattice further.The algorithms we proposed are tested under different experiment and compared with associated algorithms. A great deal of simulation results shows that the algorithms are feasible and effective.