Research On Concept Lattice Based Association Rules Mining

Association rule mining is a branch in data mining which is researched earlier and also dynamic up to now. The paper analyzes on the background above. Association rule mining is to mine knowledges from a large number of data that which is valuable in describe the natural relations among items in database. For the rules extract in database need to scan database repeatedly, it is crucial to improve the algorithm's efficiency. This paper analyzes association rule mining based on concept lattice, The main contents include as follows : association rules mining analysis,the basic theory of concept lattice,the analysis and improvement to Bordat algorithm and the analysis in how to raise the quality of association rule that extract in concept lattice. In the end of this paper, It demonstrates the benefit of the new algorithm in extend and effective performance through the experiment of the improved algorithm on different data sets; besides, it describes a method to obtain effective association rule from concept lattice.