Formal Concept Analysis and Set-valued Information Systems

Formal concept analysis (FCA) and rough set theory (RST) are two effective tools for knowledge discovery, and both have been successfully applied to various fields. This paper indicates that each formal context can be trans- formed into a set-valued information system. It is proved that the object granular reduction in a formal context is equiv- alent to the attribute reduction in the set-valued information system induced by the formal context. The characteristics of three types of object granular attributes in a formal context are analyzed. Finally the conception and the conclusion of the attribute granular are given.