Fuzzy concept lattices: Examples using the Gene Ontology

Abstract

Much research in the use of concept lattices for knowledge discovery and data mining has occurred in the past several years. Various approaches have also been proposed to create fuzzy formal contexts and to transform these into fuzzy concept lattices. This paper first briefly reviews concept lattices and then presents several approaches to creating fuzzy concept lattices. One of these approaches is demonstrated with bioinformatics data, specifically using gene annotation data files. The evidence code specified with an annotation is translated into a numeric value in (0, 1] and is interpreted as the degree of association between the gene or gene product and the annotating Gene Ontology term. These degrees of association are used to create the fuzzy formal context which can then be used to create a fuzzy concept lattice.