Object-oriented Concept Lattice With Attribute-oriented Concept Lattice Compression

Rough set theory and formal concept analysis are two related methods for knowledge discovery. They discover knowledge from different perspectives. As two potential tools in knowledge discovery, rough set theory and formal concept analysis have been concerned by artificial intelligence researchers and have been applied to various research areas, such as soft engineering, pattern recognition, information retrieval, data mining, and so on.However, with the rapid development of network technology, data and concept become very complex, and thus knowledge discovery also become more difficult. How to extract the effective information from a large number of complex data has become a new problem, and one way to solve such problems is compressing the concept lattices. Therefore this paper considers the reduction of the object-oriented concept lattices and the property-oriented concept lattices. The main results are summarized as follows:1. To propose a compression method based on object covering and attribute covering directly from the formal context, then compress the corresponding concept lattices through adjusting the object neighborhood's size or the attribute neighborhood's size by the similarity degree.2. To propose a compression method based on attribute fusion directly from the formal context, and compress the formal context through attribute fusion by the similarity degree between attribute sets. Then compress the corresponding concept lattices.