A Parallel Algorithm For Fuzzy Concept Lattice Construction

With the rapid development of computer technology and the Internet, social networks,smart devices, sensor devices, cloud computing center to generate large amounts of real-time information data, how to extract valuable knowledge has become a huge challenge. The theory of concept lattices,proposed by Wille.R in 1982,is all efficient tool for knowledge discovery, knowledge representation. The concept lattice is the core data structure of formal concept analysis directly described by the extension and intension of a concept, by partial order suggests that the link between the concepts and use of information Hasse diagram for data visualization. Fuzzy concept lattice is an extension of the classical concept lattice can describe the reality of uncertain information, has a broader range of applications. Formal Concept Analysis premise application is to construct concept lattice, and with the increasing size of data and the complexity of fuzzy concept lattice, making how to efficiently construct fuzzy concept lattice become a new research topic. In view of this situation, this paper analyzed the questions of fuzzy concept lattice construction.Firstly, describe the issue of formal concept analysis, summarize and elaborate on the existing theory and methods for classic concept lattice and fuzzy concept lattice. Give the problems of current fuzzy concept lattice construction, thus made the study of this paper.This paper presents a fuzzy concept lattice parallel construct algorithm Para Fu La(Parallel Fuzzy Lattice), to make up for lack of serial fuzzy concept lattice construction algorithm, to better meet the needs of fuzzy concept lattices applications. The algorithm is based on existing fuzzy concept lattice construction algorithm, considering the fuzzy concept lattice structure characteristics and construction tasks, first define the fuzzy concept hierarchy and the same layer node concept, this method is not only a clear description of the nature of vague concepts, to better reflect the structural characteristics of fuzzy concept lattices. Then analyzed the same layer fuzzy concept lattice nodes, know that the same layer structure fuzzy concepts’ constructed tasks independently. According to the number of computing nodes involved in the task assignment to construct fuzzy concept lattice, parallel computing and integrates the results of each task. Finally, in order to improve the efficiency of search fuzzy concept lattice, the mapping function is introduced to simplify the traversal of the search space. The proposed method stratifies the fuzzy concept nodes,make parent node constructed tasks independently, and parallel construct fuzzy concept lattice, using the advantages of parallel computing to improve construction efficiency. By algorithm analysis and experimental results, it is proved that the algorithm has good performance in the case of large data set‘s construct task.