Properties of Concept Lattice Reduction Based on Matrix Factorization

摘要

Formal Concept Analysis is a well known method for the analysis of object-attribute data. However it is difficult to deploy when it comes to large datasets. Several complexity issues of this method can be addressed by reduction using matrix factorization methods. However their behavior within this environment is not easily explainable, but can be partly illustrated using several both general and Formal Concept Analysis-specific measures. We have also introduced a new notion of Primary Concepts which may help in understanding the reduction process.