Reduction Based On Rough Set Concept Lattice And Tectonic Research

The Concept lattice theory and the Rough Sets theory are two kind of different knowledge express and the knowledge discovery tools, have succeeded applies in many domains. The rough sets theory's data performance way is the information system, its research foundation is between the object equivalent relations; the concept lattice theory's data performance way is the formal context, its research foundation is one kind of order hierarchical structure. Although two kinds of theories have the differences, but the concept lattice and the rough sets are all establishment in the classified foundation, so it also has its general character. Therefore we can use the rough sets theory and methods to study concept lattice theory, in turn we also can use concept lattice theory research rough sets, uses for reference mutually and promoted the concept lattice and the rough sets theory development mutually.First, the concept lattice and the rough sets are the knowledge discovery and the information processing important methods, each one has some advantages and disadvantages. If the concept lattice has big formal context, the part useful information will be covered by this kind of complex structure. This paper discusses relationships between the rough sets and concept lattice in reduction of objects and attribute and then give us a new methods that in the formal concept analysis we can take advantage of rough sets in reduction of attribute and construct the lattice after reduction. This methods'innovation is reduced the concept lattice based on rough sets and construct it at the same time. If the concept lattice has big formal context, this methods can let the information more stand out and the concept lattice more condensed.Second, in the existing concept lattice constructing algorithm, most of the algorithms are objects based approaching algorithms, namely increases one by one through the objects constructs the concept lattice. But the attribute—based algorithm for constructing and maintenance of concept lattice structure are few. However if the concept lattice which attribute need increases, it must needs maintenance and constructs, in order to adapt the new data; on the other side if the reduction concept lattice need return to the original concept lattice, we also need constructs and repairs the lattice. For these reasons we must design an attribute—based algorithm for constructing and maintenance of concept lattice. This paper proposed an algorithm solves these questions.