Comparison of concept lattice reduction based on discernbility matrixes

Abstract

Attribute reduction is an important issue in formal concept analysis, and its efficient reduction algorithm is accordingly important. For two existing lattice reduction methods based on different discernbility matrixes, we give their algorithms and compare their runtime. The experiments show the great difference about these two methods, which reveals that it will be more efficient from the viewpoint of parent-child relation when study the lattice reduction of formal contexts and other relative problems.