Concept Lattice Simplification Based On Rich Concept Set And Their Application

As the core data structure of form concept analysis, concept lattice already has been widely used in many fields. With the appearance of great deal of data, the information integrity of concept latice need much more time and space. So, concept lattice simplification became an importance aspect in the research of the form concept analysis.The main research of concept lattice simplification fixed attention on two aspects at present ,one is context reducing ,the other is improve on the lattice construction algorithm. Such as context clarified and standard reducing, better lattice construction algorithm and parallel algorithm used in concept lattices construction and so on. In fact, we can waive less important nodes to simplify concept lattice model.To reduce the size of concept lattice model, this paper proposes the approximate self-similarity of concept lattices to measure the relativity of every two attribute blocks in the concept lattice. At last, putting forward a method based on rich concept set to simplify concept lattice that distinguish diffierrent blocks is earier, and carrying out the algorithm of take out rich concept set from concept lattice. Interrelated experimentations are given at the same time.The main idea of concept lattice modle simplification based on rich concept set is that uses as less concept node as possible to show more information and keep the structure. So, one can get more information from the rich concept set by less jump when we navigate by the characteristics of concept lattice from the top to the end.The main contents are as follows:(1) Summarize the existing methods that simiplify the concept lattices model, such as formal context reduction, pruning algorithm and parallel algorithm used in concept lattices building process, concept cluster and congruence map and so on.(2) Proposing the approximate self-similarity of concept lattices to measure the relativity of every two attribute blocks in the concept lattice. Result shows that the approximate self-similarity of concept lattice is smaller , distinguish diffierrent blocks from concept lattice is easier.(3) Putting forward a method based on rich concept set to simplify concept lattice that distinguishs diffierrent blocks easily.