by Tho et al.
This article describes algorithms to automatically generate fuzzy ontologies. The authors identify concept by combining fuzzy logic with FCA (=formal concept analysis) which defines formal contexts to represent relationships between objects and attributes in a domain. Formal concepts are generated from the identified context.
Clustering the concepts yields taxonomic relations, a reasoning technique adds instances and another algorithm is used to integrate additional attributes. The method yields an ontology $$O=(C, A, R=(R_T, R_N), X)$$ with the concepts $$C$$, attributes $$A$$, taxonomic ($$R_T$$) and non-taxnomic ($$R_N$$) relations and axioms $$X$$.
The authors apply the F1 measure and the average uninterpolated precision (a typical measure for evaluating hierarchical constructs) to evaluate their approach.