Liya Ding

Design and Development of Knowware System

This article describes the design and development of knowware system to support automatic construction of knowledge-based systems in various application domains. It first introduces the research background and intelligent components of knowware system, and then discusses a possible development.

Mining multilevel association rules with dynamic concept hierarchy

Association rule mining has attracted wide attention in both research and application areas recently. The mining of multilevel association rules is one of the important branches of it. In most of the studies, multilevel rules will be mined through repeated mining from databases or mining the rules at each individually levels, it affects the efficiency, integrality and accuracy. In this paper, a novel method is proposed to improve this situation by analyzing the rules mined from primitive concept level to obtain multilevel rules. The proposed method also supports dynamic concept hierarchies.

Automatic Construction of Knowledge-Based System using Knowware System

This article first gives a brief introduction to the research background of Knowware System (KWS), and then describes automatic construction of knowledge-based system using KWS. The KWS offers a set of intelligent components together with the function of automatic construction of knowledge-based system, and supports application developer to generate his/her desired hybrid intelligent system without the necessity of being familiar with AI techniques.

Truth Value Flow Inference in Hybrid KBS Constructed by KWS

The Knowware System (KWS) has been proposed as an intelligent tool to support the development of knowledge-based system (KBS). It offers a set of intelligent components as basic processing units for the user to model and develop a customized hybrid intelligent system more easily and conveniently without the necessity of being familiar with AI techniques. The framework of hierarchical knowledge representation and inference has been defined for the knowledge based processing in KBS constructed by KWS. As a continued work, this article discusses the confidence transfer among different types of intelligent components by further extending the truth value flow inference.

Inference in Knowware System

Knowware System (KWS) has been proposed as a framework of intelligent tool for modeling and development of knowledge-based system. It is to support application developers in constructing customized hybrid intelligent system without the necessity of being familiar with relevant AI techniques. As a continued work of previous studies of KWS, this article discusses the development of KWS inference engine from the aspects of truth value flow inference, confidence handling, protocol between intelligent components in knowledge hierarchy, control of execution, and feedback handling, and provides key algorithms.

Application of hybrid logic in inference of Knowware System

Handling of uncertainty has been an important topic discussed in the community of intelligent techniques and soft computing. It is necessary to represent human knowledge and modeling its uncertainty when developing an intelligent system. There are various types of uncertainty in the real world, and randomness and fuzziness are of two basic kinds. How to handle these two kinds of uncertainties appearing simultaneously in a system is a main task of intelligent system development. The Knowware System (KWS) has been developed as an intelligent tool for modeling and development of knowledge-based system (KBS). It is to support application developers in constructing customized hybrid intelligent system without the necessity of being familiar with relevant AI techniques. Modeling and processing uncertainty of different types has its significance for further enhancement of KWS mechanism to better support real applications. In this paper, we will present the modeling of KBS with possible uncertainty and the propose handling hybrid uncertainty in inference of KWS, which includes randomness and fuzziness, based on the hybrid logic and chance theory.

Handling Knowledge Imperfection in Hybrid Logic Inference

When transforming information into knowledge, various forms of imperfection can be introduced. Most imperfections typically result through randomization and fuzzification. The Interval-Valued Confidence (IVC) technique has been introduced to represent the fuzzy truth for facts and knowledge in hybrid knowledge-based systems. An extended format used to represent hybrid imperfection in reasoning with hybrid logic is proposed. The credibility of resulting fuzzy proposition transforms probability into an uncertainty measure on truth. This article modifies existing effort to Extend the Interval-Valued Confidence (EIVC) technique to handle uncertain fuzzy truth of fuzzy proposition. It also considers the fuzzy likelihood of the random proposition for reasoning with hybrid logic in a unified format.

Auto-Categorization of HS Code Using Background Net Approach☆

Abstract The Harmonized System of tariff nomenclature created by the Brussels-based World Customs Organization is widely applied to standardize traded products with Code, Description, Unit of Quantity, and Duty for Classification, to cope with the rapidly increasing international merchandise trade. As part of the function desired by trading system for Singapore Customs, an auto-categorization system is expected to accurately classify products into HS codes based on the text description of the goods declaration so to increase the overall usability of the trading system. Background Nets approach has been adopted as the key technique for the development of classification engine in the system. Experimental results indicate the potential of this approach in text categorization with ill-defined vocabularies and complex semantics.

An Interval-Valued Confidence for Inference in Hybrid Knowledge-Based Systems

Knowledge Based System (KBS) is a problem solving approach that makes use of human knowledge in decision strategies. Modeling and representing imperfect human knowledge associated with uncertainty is an important task in KBS development. There are various types of uncertainty, and randomness and fuzziness are among the most important. Handling hybrid uncertainty in one KBS is critical to support real world applications. Knowware System (KWS) is an intelligent tool designed to support application developers in constructing customized hybrid KBS without requiring developers being familiar with relevant intelligent techniques. It is essential for KWS to construct corresponding inference structure in resulting KBS and process the inference with hybrid uncertainty. To fulfill this requirement the extended Truth Value Flow Inference (TVFI) and Interval-Valued Confidence (IVC) have been defined and developed as ambedded mechanisms of KWS, and the hybrid logic has been adopted for the framework of handling hybrid uncertainty.

Extended Interval-valued Confidence for Inference of Knowware System Using Hybrid Logic

Abstract An important task in developing an intelligent system is to model and represent human knowledge and its uncertainty. There are various types of uncertainty, and randomness and fuzziness are among the most important. Handling these two types of uncertainty appearing simultaneously in a system can be critical to support real world applications. We have developed the Knowware System (KWS) as an intelligent tool to support application developers in constructing customized hybrid knowledge-based systems (KBSs) without requiring developers being familiar with relevant intelligent techniques. The interval-valued confidence (IVC) has been introduced to represent fuzzy truth of facts and knowledge in hybrid KBS constructed by the KWS, and the hybrid logic has been adopted for an extended rule-based reasoning in the KWS. As part of our continued work, in this article, we further define an extended interval-valued confidence (EIVC) to handle both fuzzy truth and randomness of facts and knowledge in the KWS inference under the hybrid logic, by representing probability as an uncertainty measure on fuzzy truth.