Attribute-oriented Concept Lattice And The Partition Relationship (object)

Formal concept analysis (FCA), proposed by R. Wille in 1982, is a method for data analysis used in finding, ordering and displaying of concepts. Rough set theory (RST), proposed by Z. Pawlak in 1982, is a mathematical theory used in data analysis. As two valid and potential tools in knowledge discovery, rough set theory and concept lattice theory have attracted much attention of artificial intelligence researchers. Now, they have been applied to various research areas, such as machine learning, expert system, computer network, decision analysis, data mining, and so on.Concept lattice theory and rough set theory, study and show the hidden knowledge of data form different viewpoints. The data base of concept lattice theory is formal context and corresponding concept lattice; the data base of RST is information system and equivalent relation between objects on the universe. Although they are different theories, they have a lot in common as for goals and methods. It is better for us to analyze data information by studying the relationship between FCA and RST and combining them.The relationship between FCA and RST is mainly stuied in this paper, the main results and innovations of the thesis are summarized as follows:1. The relationship between the property-oriented concept lattice and the partition is studied, the approach to change the property-oriented concept lattice into partition and the approach to construct the extensions of the property-oriented concept by partition are proposed. Simply, the relationship between object-oriented concept lattice and this partition is discussed.2. Method of Multi-valued concept lattice construction by combing finite concept lattices is proposed.