Analogy relations and inference

An approach to analogical reasoning is introduced which is based upon the so-called resemblance relations i.e. fuzzy binary relations which are reflexive, transitive, and monotonic with respect to the set inclusion. The analogical reasoning approach is given a complete characterization of resemblance relations using the representation theorems for fuzzy binary transitive relations. The analysis of some examples shows the relationship between this approach to analogical reasoning and fuzzy inference modeled through the compositional rule of inference.<
>