Research On Design Of Multi-Disciplinary Ontology System

Ontology is originally a philosophical term. Since 1990's, some workshops on ontology have been held in the international computer community. Ontology has become one of the main technologies of knowledge representation, knowledge management, knowledge sharing and knowledge reuse by now. It is also becoming a common pivotal subject among the research fields such as NLP, Web IR, database management, knowledge management, heterogeneous data integration, digital libraries, GIS, semantic Web, etc.The main objective of ontology research is knowledge sharing and knowledge reuse. Firstly, ontology analysis can help us to understand the knowledge structure of a domain, identify concepts of the domain, and identify relations among these concepts. Secondly, outcome of ontology analysis for one domain can be reused in other domains, so repeated work can be avoided. Lastly, terms and concepts of an ontology with explicit meanings and definitions make knowledge sharing possible.Neverthelesss, constructing a multi-disciplinary ontology system is an arduous work. This dissertation addresses some methods of it. We call the collection of the methods we adopted in this dissertation MDOS (Multi-Disciplinary Ontology System). MDOS includes methods of multi-disciplinary ontology design, multi-disciplinary ontology analysis, multi-disciplinary knowledge acquisition and multi-disciplinary knowledge representation. It addresses problems of three fields:1) Domain conceptualization: How to identify categories of a domain? How to identify relations among these categories? What kind of properties (or axioms) do these relations own? How to identify slots of a category? How to name these slots? What kind of properties (or axioms) do these slots own?2) Ontology representation: How to represent an ontology? How to make the representation formalized, natural, and effective?3) Ontology analysis: How to analyze and find existing problems (including problems of inconsistentness, redundancy, imperfection, etc.) of categories, individuals, slots and axioms?Main contributions of this dissertation are as follows:1) Based on the analysis of many ontology languages and amount of multi-disciplinary knowledge, a set of principles and a method of ontology construction are put forward. The method includes three main steps: domain analysis, ontology and knowledge representation, and ontology and knowledge analysis. The method has been applied in constructing a large knowledge base covering 19 domains.