Toshiharu Iwatani

New Design Concepts for the FLINS-Fuzzy Lingual System: Text-based and Fuzzy-centered Architectures

A fuzzy natural language communication system called the Fuzzy Lingual System (FLINS) is currently under development at the Laboratory for International Fuzzy Engineering Research (LIFE). The final goal of the FLINS project is to create a lingual computer, that is, a domain-independent teach, question, and answer (TQA) system. In this paper, we propose two new design concepts, text-based architecture and fuzzy-centered architecture, to realize our goal. Two experimental systems were built to determine the feasibility of those approaches before the development of FLINS. One is a text-based natural language understanding system, called the AB-System. The other is a fuzzy expert system called FOREX, which predicts exchange rate trends according to fuzzy rules and fuzzy data.

Fuzzy natural language communication system-FLINS: concept and conversation examples

The Fuzzy Computing project (FC project) at the Laboratory for International Fuzzy Engineering Research (LIFE) in Japan is developing a natural language communication system called FLINS, which is short for Fuzzy Lingual System. The final goal of the project is to implement a lingual computer that can communicate and learn, both by being taught and on it own, through use of fuzzy natural language. This paper gives an overview of the FC project, especially FLINS. It first briefly presents the key concepts of FLINS, then explains the three-layered fuzzy inference and the inference control based on the three-layered inference that are the special features of FLINS, and finally gives conversation examples.<<ETX>>

Treatment of fuzziness in natural language by fuzzy lingual system-FLINS

The Laboratory for International Fuzzy Engineering Research (LIFE) is developing a natural language communication system called Fuzzy Lingual System (FLINS). This paper briefly describes methods used in FLINS to treat fuzziness in natural language, namely an inference method for natural language propositions involving fuzzy quantifiers, an inference method extended the case-based reasoning by fuzzy theory and a learning method using fuzzy analogy.<<ETX>>

Basic structure of three-layered fuzzy inference in FLINS-fuzzy lingual system

The Laboratory for International Fuzzy Engineering Research (LIFE) in Japan is developing a natural language communication system called FLINS, which is short for fuzzy lingual system. The final goal of the project is to implement a lingual computer that can communicate and learn, both by being taught and on it own, through use of a fuzzy natural language. In this paper, the goal of FLINS is briefly presented. Next, the problems and limitations of conventional systems are summarized, and new design concepts, text-based architecture and fuzzy-centered architecture, to overcome those problems and limitations are proposed. Finally, the structure of the three-layered fuzzy inference mechanism, which is one of the most important features of FLINS, is described.<<ETX>>

Three-layered fuzzy inference and self-wondering mechanism as natural language processing engine of FLINS

We are developing a natural language communication system called FLINS, which is short for Fuzzy Lingual System. The final goal of the project is to implement a lingual computer that can communicate and learn, both by being taught and on it own, through use of a fuzzy natural language. New design concepts, text-based architecture and fuzzy-centered architecture, to overcome the problems and limitations of the conventional systems are proposed. Next the problem solving by the three-layered fuzzy inference, and the meta-inference by the combination of the three-layered fuzzy inference and the self-wondering mechanism are described as the natural language processing engine of FLINS.<<ETX>>

Fuzzy Analogy Based Reasoning and Classification of Fuzzy Analogies

Conventional research on analogical reasoning (AR, for short) theory has not yet addressed the management of fuzzy matching between two different predicates, though human beings essentially utilize fuzzy matching when they infer analogically. On the other hand, fuzzy logic has been successfully applied to deductive and inductive reasoning to make them more flexible. Although the goal of both fuzzy logic and AR is to achieve more flexible human-like reasoning, there have been few applications of fuzzy logic to analogical reasoning. In this paper, fuzzy-analogy based reasoning (F-ABR), an extension of ABR, is proposed. In ABR, an analogy represents clear partial agreements between two knowledge areas, each of which is described as a set of predicates. In F-ABR, a knowledge area is a set of fuzzy predicates and a fuzzy analogy means fuzzy partial agreements between two knowledge areas. Using fuzzy logic, F-ABR can infer in a way that is more flexible and human-like than conventional ABR. This paper discusses three topics: first, a fuzzy matching method between two fuzzy predicate symbols is described. A fuzzy analogy contains a set of pairs composed of a predicate symbol and a similarity degree, and three methods for calculating the similarity degree are also described. Second, methods for classifying and ordering fuzzy analogies, based mainly on similarity degrees, are introduced. These methods are necessary when selecting a single fuzzy analogy to use in the reasoning process. Finally, the features of each type of fuzzy analogy are analyzed in order to show that many kinds of flexible reasoning can be achieved by selecting a fuzzy analogy.

An acquisition method of implicit knowledge using fuzzy analogy

A method to acquire implicit knowledge among several given cases using fuzzy analogy, called fuzzy analogical learning (FAL), is briefly described. This method is proposed as a learning feature in a natural language communication system, a fuzzy lingual system, which aims to emulate the observational features of human learning and problem solvings based on experiences. Cognitive models of learning and problem solvings for the further study of FAL and a prototype implementation for demonstrating the validity of the method are described as well as a formal description of the computational model.<<ETX>>

An Inference Method for Fuzzy Quantified Natural Language Propositions Based on New Interpretation of Truth Qualification

In this paper, we propose a method that affects inference results leading to a new interpretation of a truth qualification by adding a weight attribute to truth qualified fuzzy sets. With this method, we can obtain different inference results depending on the truth qualifiers by transforming a statement with fuzzy quantified and truth qualified natural language propositions. We applied our method to two examples transforming a fuzzy predicate of the natural language propositions and showed an effectiveness of the method.

FLINS-fuzzy natural language communication system

Demonstrates a natural language communication system called FLINS (Fuzzy LINgual System). The system handles fuzziness in natural language, and it can teach, question and answer by natural language. In the demonstration, we show how the system works, especially how the inference engine works through example dialogs in an ordinary life using the Sun workstation, a speech recognition device (PE200) and a speech synthesizing device (DECtalk).<<ETX>>

An inference method for fuzzy quantified natural language propositions based on new interpretation of truth qualification

In this paper, we propose a method that affects inference results leading to a new interpretation of a truth qualification by adding a weight attribute to truth qualified fuzzy sets. With this method, we can obtain different inference results depending on the truth qualifiers by transforming a statement with fuzzy quantified and truth qualified natural language propositions. We applied our method to two examples transforming a fuzzy predicate of the natural language propositions and showed an effectiveness of the method.