An ontology-based approach to management of model knowledge

基於本體論的模型知識的管理

Student thesis: Doctoral Thesis

View graph of relations

Author(s)

  • Ou LIU

Related Research Unit(s)

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date15 Feb 2006

Abstract

Decision models are important resources in Decision Support Systems (DSSs). Several methods have been proposed to facilitate the management of model knowledge. However, they can hardly support the sharing and reuse of the model knowledge in a collaborative decision-making environment. Also, current methods do not consider vague and imprecise information which is commonly found in model management of DSSs. There is a need for a formal and semantically rich specification of model knowledge supporting vague information, which is useful to enhance the sharability and reusability of model knowledge. The emergences of new Internet technologies such as Web services and XML-based documentation, and the evolution of artificial intelligence such as ontology engineering, also create new opportunities in model management. In this thesis, an ontology-based approach to the management of model knowledge is proposed, where decision ontology is designed to represent the model knowledge. A fuzzy description logic is defined for formal and semantically rich specification of decision ontology. Then an XML-based language (DOML: Decision Ontology Markup Language) is designed to represent the decision ontology and make it interchangeable over the Internet. A Web service-based implementation architecture is also proposed for implementation of a practicable model management system. The architecture makes it possible for any program to connect and access the models in a standard way. The applicability of the proposed methods and theories is illustrated by a practical model management application in Mainland China. Keywords: Model Knowledge, Model Management, Ontology, Fuzzy Description Logic, XML, Web Services

    Research areas

  • Decision support systems, Decision making