Construction And Reduction Of Concept Lattice And Application Of Formal Concept Analysis

Formal concept analysis (FCA) was elaborated by Professor Wille of German in the eighties of the twentieth century. With the development of the research about FCA, it has become a notable research topic in computer science field. Concept lattice is also called as Galois lattice, which is the core data structure of FCA and represents the order relationship between the concepts concretely and iconically. Also, it is an important knowledge expression structure. This paper mainly studies some basic points of concept lattice, whose creative research works are as follows:(1) The constructing method of concept lattice. This paper presents an algorithm of generating concepts rapidly which is based on the consecutive matrix. We compare the time complexity of the new methods with the traditional one. For using the existent efficiently, this paper also develops an algorithm that is merging concept lattices based on direct product operation.(2) The reduction method of concept lattice. This paper presents a concept lattice reduction method that is based on the user's requirements and different from the attribute reduction theory of concept lattice presented by Zhang Wenxiu. Users are allowed to reorder the attributes according to their interesting levels. Then the method will do reduction operation on the concept lattice based on the new order. After this course, the most valuable concept lattice for users is got.(3) The application of concept lattice in the recognition of handwriting figures. The thesis combines formal concept analysis theory with rough geometry theory and applies them in handwriting figures recognition tentatively, which provides new thought for handwriting figure recognition.