Research On Ontology-based Diagnosis Expert System Of Rice Diseases And Pests

Ontology, which is regarded as a sharing concept model of expressing knowledge, has been paid much more attention since it was proposed. Ontology has been an important modeling tool in the field of knowledge engineering, knowledge management, information retrieval, semantic web and AI. On the basis of the feature and character of rice disease and pest, an expert ontology knowledge base is built using Protege. An ontology reasoning technology is used to diagnose and reason disease and pest. Ontology-based expert system for rice disease and pest diagnosis is researched, and the prototype system is developed. The test showed that disease and pest can be diagnosed accurately. Moreover, the control methods is also be provided.This article mainly focuses on three key techniques in intelligent diagnosis system of rice disease and pest, respectively conceptual modeling of rice disease and pest, uncertain diagnosis reasoning of rice disease and pest, and the design of prototype system. The main contents of research can be as following:1. A rice disease and pest-oriented methods of concepts modeling based on Ontology Web Language (OWL) are researched. According to the occurrence features and knowledge traits of rice disease and pest, the research content of domain knowledge is confirmed. An organization representation of rice disease and pest based on ontology is proposed and realized using OWL. Meanwhile, an ontology model of rice disease and pest knowledge has been built.2. A rice disease and pest-oriented method of uncertain diagnosis reasoning based on ontology is proposed. Through analyzing the characteristics from the domain of rice disease and pest, the knowledge of ontology base is retrieved and matched. A method of ontology-based reasoning and similarity calculation are used to realize disease and pest diagnosis reasoning, and the algorithm of weight coefficient is applied to improve diagnostic accuracy of the system.3. An intelligent diagnosis prototype system of rice disease and pest based on ontology is developed on MyEclipse platform. JSP and Jena technology are adopted to realize rice disease and pest diagnosis reasoning system. Pest diagnosis is mainly based on morphology trait, harm position and harm symptom. And disease diagnosis is mainly based on lesion shape, lesion color and harm symptom. The diagnosis interface and control interface are showed friendly, and designed clearly and used easily.The research is combined with key project subjects of the eleventh five-year national science and technology pillar program:the research and application of double cropping rice expert system in hilly red soil (2009BADC4B02). The intelligent diagnosis system of rice disease and pest based on ontology is developed, which is integrated some functions, such as diagnosis, control and query. The creation of system has a good application value for the diagnosis and control of rice disease and pest in the south hilly red soil region.