Research And Application Of Automobile Insurance Assess System Based On Case-Based Reasoning

With the increasing of the automobiles' quantity, the traffic accidents are becoming more and more. To assess damaged automobile correctly is a toilsome and complicated job. Now, the automobile insurance assess trades is not very standard in our country. The result of assess is influenced because the professional quality of assess personnel and the standard in the market are not very good. How to help assess personnel assess quickly and exactly through computer system and artificial intelligence technical and how to accumulate and develop the experience of insurance assess, guarantee this trades to have a standard is an important task. This dissertation intends to use the technology of CBR to build a model of automobile insurance assess. This model is applied to the "Automobile Insurance Assess System Based on Case-Based Reasoning". It can make the automobile insurance assess more reasonable and improve the quality and efficiency. The main contents are as follows:(1) The basic theories of CBR have been deeply analyzed and researched. To the features of automobile insurance assess, compare the features of CBR with some other artificial intelligence technical such as expert system, neural network, machine learning and etc. The application in automobile insurance assess system is discussed.(2) The case expression method is combined frame expression with expression based on feature. And constructing the automobile insurance assess database with the method of classified tactics based on it.(3) The index tactics based on multi-layered is used when searching for the cases. Building an automobile insurance assess' multi-layered index tree. During mating the cases, it uses nearest neighbor approach with weight to judge similarity match between cases.(4) Using the technology of CBR to establish a model of automobile insurance assess system. A prototype system is designed and implemented, which has completed the predestinate function. The test to the system proved the reasonability and the validity of the design.