A Declarative Semantics for a Fuzzy Logic Language Managing Similarities and Truth Degrees

Tuning Fuzzy Logic Programs with Symbolic Execution

“…Currently, the system can be used to compile MALP programs to standard Prolog code, draw derivation trees, generate declarative traces, and execute MALP programs, and it is ready for being extended with powerful transformation and optimization techniques [11,12,13]. Our last update described in [14,15], allows the system to cope with similarity relations cohabiting with lattices of truth degrees, since this feature is an interesting topic for being embedded into the new tuning technique in the near future. Another interesting direction for further research, consists in combining our approach with recent fuzzy variants of SAT/SMT techniques.…”

“…Most of these systems implement (extended versions of) the resolution principle introduced by Lee [19], such as Elf-Prolog [10], F-Prolog [20], generalized annotated logic programming [17], Fril [4], MALP [24], R-fuzzy [9], the QLP scheme of [31] and the many-valued logic programming language of [36,34]. There exists also a family of fuzzy languages based on sophisticated unification methods [33] which cope with similarity/proximity relations, as occurs with Likelog [3], SQLP [8], Bousi∼Prolog [16,32] and FASILL [14,15]. Some related approaches based on probabilistic logic programming can be found, e.g., in [29,7].…”

On the Incorporation of Interval-Valued Fuzzy Sets into the Bousi-Prolog System: Declarative Semantics, Implementation and Applications

“…Proximity-based Logic Programming is a framework that provides us with the capability of enriching semantically classical logic programming languages by using Proximity Equations (PEs). A limitation of this approach is that PEs are mostly defined for a specific domain [6,23], being the designer who manually fixes the values of these equations. This fact makes harder to use PLP systems in real applications.…”

“…Inside the former and current frameworks of fuzzy logic programming [5,6,7,8,9,10], we argue that lexical reasoning might be an appropriate way for tackling this challenge, because of this type of knowledge is usually expressed linguistically. However, from a computational point of view, this source of information involves vagueness and uncertainty and, consequently, it must be specifically addressed.…”