Fuzzy Logic Rules Modeling Similarity-based Strict Equality
Ginés Moreno, Jaime Penabad, Carlos Vázquez
DOI: http://dx.doi.org/10.15439/2014F387
Citation: Proceedings of the 2014 Federated Conference on Computer Science and Information Systems, M. Ganzha, L. Maciaszek, M. Paprzycki (eds). ACSIS, Vol. 2, pages 119–128 (2014)
Abstract. A classical, but even nowadays challenging research topic in declarative programming, consists in the design of powerful notions of <<equality> > , as occurs with the flexible (fuzzy) and efficient (lazy) model that we have recently proposed for hybrid declarative languages amalgamating functional-fuzzy logic features. The crucial idea is that, by extending at a very low cost the notion of <<strict equality> > typically used in lazy functional (HASKELL) and functional-logic (CURRY) languages, and by relaxing it to the more flexible one of similarity-based equality used in modern fuzzy-logic programming languages (such as LIKELOG and BOUSI-PROLOG), similarity relations can be successfully treated while mathematical functions are lazily evaluated at execution time. Now, we are concerned with the socalled <<Multi-Adjoint Logic Programming approach> > , MALP in brief, which can be seen as an enrichment of PROLOG based on weighted rules with a wide range of fuzzy connectives. In this work, we revisit our initial notion of SSE (<<Similarity-based Strict Equality> > ) in order to re-model it at a very high abstraction level by means of a simple set of MALP rules. The resulting technique (which can be tested on-line in dectau.uclm.es/sse) not only simulates, but also surpass in our target framework, the effects obtained in other fuzzy logic languages based on similarity relations (with much more complex/reinforced unification algorithms in the core of their procedural principles), even when the current operational semantics of MALP relies on the simpler, purely syntactic unification method of PROLOG.