Distributed Algorithm For Constructing Concept Lattice Study

Concept lattice,with the characteristics of integrality, accuracy and simplicityetc, is a mathematical model which reflects the hierarchy between concepts, and a powerful tool for data analysis and data mining.With existing of many distributed and heterogeneous data sets,and constructing complexity of concept lattice, distributed constructing of concept lattice is an effective way to reduce the time and space complexity of the concept lattice oriented massive data. In this paper, distributed constructiing algorithms of concept lattice are studied by using grid as a computing platform. The main research works can be summarized as follows:Firstly, a distributed constructing method of concept lattice is presented based on grid.Using the grid as a distributed computing platform, a distributed constructing algorithm is presented by taking a scheduling strategy of multiple distribution suitable for the scale of concept lattice constructing. In the end,experimental results validate the algorithm is correct and efficient by taking the star spectra data as the formal context.Secondly,a distributed constructing method of concept lattice is presentd based on pruning.Using a pruning technology to eliminate redundant information which appears in the concept lattice's distributed incremental construction process,a distributed constructing algorithm of concept lattice is prenseted,so that the comparing time of inserted concepts is reduced,and the constructing efficiency of concept lattice is effectively improved.The experiment results validate the correctness and validity of the algorithm by using the celestial spectrum data as the formal context.