Advanced Computing with Words: Status and Challenges

Abstract

In this chapter, we focus on the status of Advanced Computing with Words (ACWW) and the challenges that it may encounter in the future. First, we elaborate on the notion of Computing with Words (CWW) and its various subareas. Then we present some non-engineering ACWW problems and connect them to more realistic engineering problems, after which we provide a roadmap for solving ACWW problems, and show how the Generalized Extension Principle (GEP) can be used to formulate their solutions. We also propose a syllogistic approach to solving ACWW problems that also uses the Extension Principle but in a different way. Finally, we discuss present and future challenges to ACWW, i.e. we explain what challenges ACWW encounters given the current trend of data abundance and widespread use of Internet for dealing with questions posed in natural languages. Introduction: Computing with Words and Advanced Computing with WordsComputing with Words (CWW or CW) [11,29,34,66,69,70,75,91,104,109,142,144,147,154] was probably conceived as the main area of application of fuzzy logic [132] when the field was originated. It is a methodology of computation whose objects are words rather than numbers, although those words are linked to numbers and classical calculations that can be carried out by computing machinery, via membership functions of fuzzy sets associated with words. To do that, a process called precisiation of meaning [152] is essential, which, in a nutshell, is determining membership functions of the fuzzy sets that model the words. Zadeh distinguishes two subareas for CWW:1. Basic Computing with Words (BCWW), which mainly deals with descriptions of complex systems in terms of fuzzy IF-THEN rules [134]. Soft constraints are generally possibilistic, i.e. they describe (physical) attributes like speed, temperature,

A Novel Fuzzy Modifier Interpolation Rule for Computing With Words

“…Edge case is applied when (2) is satisfied by Interval Type-2 Fuzzy set A. Eq (6) depicts the application of modifier to Interval Type-2 Fuzzy sets and Eq (7) gives the membership of the output edge for Interval Type-2 Fuzzy set when l th rule is fired is given:…”

“…CW engine consists of Rule Base, Fuzzy Inference Engine, and Output Generator. The IF-THEN rules in the CW engine are specified using natural language, which is modeled using either Type -1 or Type -2 Fuzzy sets (Interval Type-2 and General Type-2) [7].…”

On the choice of similarity measures for type-2 fuzzy sets

“…In advanced CW, similarity is often used to choose the most appropriate linguistic approximation for the output of the CW system. This is achieved by finding which word (modelled as a fuzzy set) is most similar to the fuzzy output [28]. Similarity measures are also used to group similar words [40], for example, to ensure that each word has a distinct meaning.…”

Computing with words: Are we ready to "talk" to our cars?

“…Fuzzy logic can be used to interpret such commands and hence the term "Computing With Words". Few researchers tried to build such models for CWW, for example, Mendel [2], focused on the status of Advanced Computing with Words (ACWW) and the challenges that it may encounter in the future, Khorasani et al [3], presented an enhanced inference engine toolkit for supporting computing with words. This digest paper simply propose to use linguistic terms definition as proposed by Kouatli [4] and termed as "Genetic Fuzzimetric Technique" (GFT) where definitions of "closest" and "extreme left" can be represented.…”