Research Of Ontology-based Agricultural Knowledge Modeling And Reasoning

Knowledge modeling is an important research content of knowledge-based systems,following expert systems in knowledge engineering domain. With the proposed ofOntology, the study of ontology-based knowledge modeling is the mainstreamdirection of the development of knowledge engineering. With the characteristics ofagricultural knowledge, related research of ontology-based knowledge modeling andreasoning techniques are carried out in this dissertation. The main research contentsare summarized as follows:(1) An ontological organization frame is proposed for agricultural knowledgeaimed at the problems existed in researches of agriculture knowledge systems andagricultural ontology. Through analysis for different forms of knowledge (descriptive,procedural, etc.) and different application types (search, diagnosis, solution, etc.),with the purpose of knowledge application, various ontological models are builtaccording to the characteristics of knowledge and requirements of application.(2) A semi-automatic agricultural professional ontology construction method isproposed for semi-structured network information resources. It contains four stages:data collection, DB-ontology mapping, generating OWL code and checking andrefining. By this semi-automatic construction method, the development time wasshorten and the ontology’s build progress was greatly improved. While expertassistance ensures the reliability of the resulting ontology and knowledge consistency.It provides a feasible method for ontology learning in the case of semi-structured datasources with comprehensive Web information, clear web page structure, specificsubject.(3) For problem solving knowledge, construction of agricultural task ontology arestudied. An ontology-based agricultural knowledge modeling method is proposed fortask knowledge. Thinking of integrating the PSMs and ontologies, knowledge modelis built through structure analysis in knowledge level and establishing conceptualmodel in ontology level. PSM guides the knowledge organization structure andprovides a way for the reuse of reasoning. Ontologies describes the knowledge withexplicit specification for knowledge reuse. As kernel of knowledge engineeringresearch, PSMs and ontologies provide the backbones for building structured andreusable knowledge models.(4) The ontology reasoning techniques are studied aiming at the professional ontologies. The characteristics of the different ontology inference engine in differentprotégéversions are compared, and ontology reasoning techniques, such as SWRL,SQWRL and SPARQL are analyzed by experimentation. The methods ofSWRL-based pig disease diagnosis and ontology-based semantic retrieval areproposed, possible ontology reasoning and application are described.(5) Aiming at the task ontologies, inference mechanisms, such as combination ofqualitative decision and quantitative model, uncertain reasoning and multi-pathreasoning, are studied in knowledge-based reasoning. Practical applications show thatthe knowledge models established through PSMs and ontologies technology, not onlyconducive to knowledge sharing and reuse, knowledge model’s maintenance andexpansion, but also achieve the reuse of reasoning module own to its inference engineindependent of the knowledge base.(6) The prototype systems for ontology-based knowledge modeling and the Chineseontology building have been developed. Through the development of the tools, theprocesses of agricultural knowledge acquisition and maintenance can be accelerated,at the same time the development of knowledge applications and agricultural ontologyservices can be promoted.