Metric truth as a basis for fuzzy linguistic reasoning

The Fuzzy Metric-Truth Reasoning Approach to Decision Making in Soft Computing Milieux

“…In any case, however, we have to obey two basic metarules: First, the statement is true iff Dist(P, Q) = O. Second, the statement is false iff Dist(P, Q) is maximal (Niskanen, 1993). In this respect the FMT approach resembles systems based on similarity measures such as used in ( Ruspini, 1991;Schwartz, 1991).…”

“…Within the FMT approach the crucial idea is that we encounter the concept of truth by applying the fuzzified theory of Truth/ikeness (Niskanen, 1993). The original theory (see Niiniluoto, 1987), which is bivalent by nature, assumes that the truth value of a statement depends on the distance between this statement and its true counterpart, and the true counterpart corresponds to a fact of the actual world.…”

“…Hence, in (Niskanen, 1993) an alternative linguistic user interface, quasi-natural language' (QNL), was suggested in which expressions of the type presented in Table II, infer alia, were accepted. In general, the structure of QNL is based on the SOT and the foregoing standpoints.…”

“…We will focus on linguistic, logical and semantic aspects. A more thorough examination can be found in (Niskanen, 1993).…”

“…This aim, which is an essential constituent of Lotfi Zadeh's goal, "computing with words", means that we attempt to combine certain techniques of symbolic approach, such as formal language construction and utilization of reasoning models, with the numeric approach of fuzzy systems and psychologic aspects of natural language. As (Niskanen, 1993) considered the logico-methodological aspects at a general level, below we focus on the applicability of this approach, particularly in the case of the fuzzy if-then clauses, because they are widely used within control and decision making. A forthcoming paper will consider empiric aspects in the light of certain typical data examples (Niskanen, 1996).…”