Study On The Horse Disease Remote Diagnosis Expert System Based On The Method Of CBR-RBR Integrated Reasoning

With the continuous development of the national economy,the speed of the progress of China's horse industry is also accelerating.But at the same time,it faces with new opportunities and challenges.In recent years,the specialized breeding and cultivation in horse research,demonstration work had played a certain effect on the development of the horse industry.But at the same time,to make progress in the production and breeding of horses,the occurrence of disease is still restricting the sustainable development of the horse industry.At present,there are many problems in horse breeding,such as horse production and breeding more distributed in the grassroots,disease diagnosis infrastructure conditions are relatively backward,domain experts are scarce,the grassroots horse veterinarian with rich clinical experience can not meet the needs of the development of the industry in terms of quantity and quality.In view of the horse disease occurred frequently,difficult diagnosis of horse disease,low diagnostic accuracy,this study aimed to construct horse disease diagnostic and management information system with expert-level,to reduce the misdiagnosis rate and missed diagnosis rate of clinical diagnosis and treatment of horse disease,to improve the service quality of grassroots horse veterinarian.The horse remote diagnosis and management information system were designed in accordance with the N-tier architecture and object-oriented technology,the development and construction were based on.NET framework technology,using tools such as Asp.net and SQL Server 2008 database.This study investigated the user requirements of the grassroots horse veterinary by visiting 7 demonstration areas in the Xinjiang Uygur Autonomous region,analyzed and a summary of horse disease knowledge characteristics and laws.Knowledge acquisition was under the collaboration of domain experts.By means of artificial knowledge acquisition,the expert knowledge and professional knowledge of horse disease were summarized and analyzed.The system used the Object-Attribute-Value(O-A-V)ternary knowledge representation to improve the production rules of knowledge representation and framework knowledge representation.Building the knowledge and database was based on the relational database.In this study,the rule knowledge base,case knowledge base,treatment database,case database,prevention database,multimedia database and working database of diagnosis and management system were respectively constructed.On the basis of analyzing the thinking mode and diagnostic method of domain expert,some improvements were made to the traditional rules of reasoning.Finally,the system adopted the rule reasoning which was based on fuzzy rule promotion theory as the core reasoning method,integrated with case based reasoning.The integrated reasoning mechanism fully combined the advantages of the two reasoning theories and improved the accuracy of the system diagnosis.The final implementation of the system functions as follows:(1)The horse disease remote diagnosis expert system and horse disease remote management information system had been successfully developed,which can diagnose 118 kinds of horse diseases and realized the management function of relevant information.(2)Horse disease remote diagnosis expert system,in view of the rule confidence coefficient is fixed in traditional rule based reasoning,the confidence coefficient multi-valued logic was applied to evaluate the uncertainty of knowledge.This study adopted the rule reasoning which was based on fuzzy rule promotion theory as the core reasoning method,which was integrated with case based reasoning.CBR as a supplement to the improved RBR and the guarantee of the result verification.(3)Horse disease remote diagnosis expert system realized the intelligent diagnosis of horse disease.The system can provide 19 disease symptom information group,more than 500 pictures and video multimedia information,and 118 kinds of auxiliary diagnosis of horse disease.According to the clinical symptom information provided by the users,the diagnosis results were delivered to the users through the diagnosis reasoning process.During the operation,users can also view the typical symptoms of pictures,video and other multimedia,the system can provide decision support for clinical diagnosis and treatment.(4)Horse disease remote management information system can record the specific incidence of sick horses,and statistical analysis of the incidence of disease.It provided a detailed reference for the disease diagnosis and prevention and control of grass-roots horse veterinary.It achieved the management function of immune,de-worming and disinfection management functions.According to the corresponding work procedures,it achieved the reminder function,and the occurrence of the disease was avoided due to the lack of the function of the immune system of the body.(5)Horse disease remote management information system also provided the typical medical records and disease prevention knowledge learning function for users.The system not only displayed the information in the form of text information,but also provided a wealth of learning resources,such as the typical disease symptoms picture and video multimedia materials.In addition,the system provided the remote teaching platform,to achieve users and experts network platform information exchange.(6)The horse disease remote diagnosis and management information system had been applied to the demonstration area,and had been verified by grouping.The evaluation results showed that the system had a good performance at the level of disease diagnosis,user friendliness,the clarity of questions description and completeness of the disease knowledge base.The system had as strong practicality and operability,and can provide disease-assisted diagnosis and treatment and management services for the grass-roots horse veterinarians.