The Application Of Case Based Reasoning (CBR) On Expert System Of Vegetable

Agricultural expert system is the technology in which expert system is applied in agriculture. Expert system is a branch of artificial intelligence, whose aim is to make the computer have the role of human expert in various fields. Expert system is an intelligent system. Since there is much knowledge and experience in the expert system, it can solve the problem in some field as the human expert. In fact, expert system is a program, which have the capacity of human expert and can finish and simulate the human expert. Futhermore,expert system can connect practical knowledge with experience knowledge.In recent years, the development of the expert system have made agricultural expert system applied widely in our country's agricultural production. Agricultural expert system has been used to guide agricultural production. it has proved that expert system have significant effects for the agriculture sustainable development. As far as the knowledge fuzziness and the lack of vegetable expert system are concerned, the author in this article designed an CBR-based expert system about vegetable pest and insect. Because of the combination of artificial intelligence, expert system, knowledge engineering and information technology, the system have the features of advance and practice.CBR method is a new reasoning mode which has emerged in recent years an is different form traditional RBR. For CBR, existing and old case or experience is used to solve new problem and evaluate solution, explain exception or understand new situations. The reason that CBR is becoming a hot spot is that traditional RBR has some disadvantages.Compared with CBR, the RBR have no memory and can't deal with exception, which cause the low reasoning efficiency and weak performance. There are two parts in CBR. one is the cases, the other is inference engine. the reasoning process can be summarized four steps: query,reuse, adaption and storage.This reasoning mechanism is nearly the same as domain expert. As for the bottleneck of knowledge acquisition existing in traditional vegetable expert system, CBR is applied in vegetable expert system. By CBR, vegetable expert system is reconstructed and a new definition of diagnosis of vegetable disease is provided pest . Finally, algorithm and methods are introduced when the CBR technology is used in vegetable expert system.In this article, vegetable expert system about disease and pest. For agriculture knowledge, there exist .dispersion ,diversity and uncertainty. In CBR system, knowledge representation is based on cases and cases acquisition is more convenient, so the bottleneck of knowledge acquisition is solved. CBR is reuse of past solving result and not the reasoning from the beginning. CBR can guide how to avoid failure by its experience of success and failure in the past. CBR have the capacity to learn constantly, which can make CBR adapt to the solving of future problem. Because of the advantage of CBR and the feature of agriculture itself,CBR technology is used in the expert system.The vegetable expert system is based on java platform and windows 2000 operating system above. The system main includes land system, disease designed ,disease information input ,case index. etc.This paper important introducesis the Algorithm of the case reasons logically process that Use JAVA language into realization in chapter 4.At first build up a type which useds for computing case likeness a degree,That type of structure function hasing two parameter is respectively the object which has already had a new case and an old case in database.That object can accurately describe a vegetable plant diseases and insect pests case and an other type, for example used for the attribute value which describes the case characteristic attribute value, used for the type which describes a case characteristic vector set etc.. For the type of calculation smilar degree, it's important function is that can compute the smilar degree of two objects.While compute similar degree, we will have visit for Cloud View of database.Use to take value and visit and reason that it correspond to the characteristic vector form, in order to get the same characteristic attribute in two of alike value, then compute an end the alike value, at last two according to the characteristic has attribute value data within form, the case alike object will return to alike value, the system saves it, then to compute the case database in the Cloud View of descend a record(next case).So after ending the whole individual case example database form interview, we have to all cases in the case databases with new the alike value of case, again compute highest of three , Therefore we complete the process of index.CBR provides a new way for diagnosis of vegetable disease and pest. The use of which not only help peasant combat disease and pest but also help agriculture expert to diagnose complex diseases by the new conclusion reasoned from unknown cases . Obviously, vegetable expert system based on CBR is of great significance.Although some research has done on the disease and pest knowledge management based on the CBR, there are some problems to solve. Firstly, the key feature and weight eigenvalue in the disease and pest cases was determined by domain expert, which should be determine by other methods to distinguish the full dependence on the domain expert. Secondly, conditional attribute feature is not defined precisely, which may cause dimensions disaster of inference matching. Thirdly, since the maintenance of case base mainly depend on domain expert, there may exist case redundancy and case contradiction, which will affect the reasoning efficiency and accuracy. Finally, there are much data and knowledge that are hidden in the case base, which should be mined. Undoubtedly, this field has wide research prospect.