Applied Research Of Dog Diseases Diagnostic Expert System Based On Web

With the people's living standards improving, more and more families are beginning to raising dog, accordingly dog diseases rises frequently. A great deal of persons with clinical knowledge and rich experience who give in dog diseases diagnosis and treatment are needed accordingly. But our existing clinical treatment of diseases in animals and the number of personnel level will be very difficult to meet this growing demand, and this contradiction is becoming more and more prominent. It is very difficult to resolve this contradiction simply relying on the professional training of clinical treatment of animal diseases in a short period of time. So, the professional expert systems for dog diseases diagnosis and treatment are good approaches to solve the major problem.Knowledge acquisition is the development of expert systems "bottleneck". The study established basic pattern of knowledge acquisition. Firstly, the relevant raw knowledge of dog diseases was built directly to tables. Secondly, experts defined the value of the relationship among diseases, symptoms and therapy principles according to their experience and identified the weight of related rules and the opening of the required certainty thresholds. Finally, knowledge engineers optimized the knowledge and built knowledge tables. The results showed that this method can be used to acquire time-saving knowledge, knowledge engineers and experts in the field to alleviate the workload, and also can obtain expert knowledge as much as possible.Because the amount of knowledge used in dog diseases treatment is very large, and the relationship among them is quite complex, the production rules is used to represent these knowledge. This method of knowledge representation preserves modular and understandable features of these rules. These rules were saved in Oracle9i database. The powerful search and management capabilities of the database make up for the deficiency of the inefficient reasoning in production rules.In dog diseases diagnostic process, the symptoms and conclusions of uncertainty is prevailing, based on the uncertainty of these knowledge, we improved C-F models,established an uncertainty reasoning model with weighted factors. At the same time, the express form of uncertainty evidence, combination evidence uncertainty algorithm as well as evidence uncertainty transmission algorithm is given. This model may express well dog disease diagnoses the symptom and the conclusion uncertainty analysis and the treating processes in the expert system, reduce the subjective factor to the clinical diagnosis and the treatment conclusion influence.This research uses Jbuilder9, WebLogic and Oracle9i for the integrated development environment, has realized the Web-based expert system of dog diseases diagnosis.