Research On Intelligent Decision Support System In The Group Incidents By Case-based Reasoning

With the development of the economic and social at high speed, Group incidents occurfrequently in our country. The field is not only diversified, but also the forms of the reasons arevaried.Besides, Participants will also be on the rise. Several of large scale and complex groupincidents have impact on social stability and development. Therefore, it is necessary toestablish a practical intelligent decision support system to deal with group incidents, whichuses modern scientific method to make the emergency decision as soon as possible, in order toreduce the danger of social and economic.First of all, as a starting point, it is analyzed the concept of Group accidents from thedifferent perspectives, and deducted basic characteristics features by a large amount of groupincidents. Further analyzing the features and classification to attain qualitative and quantitativeanalysis about properties, forms, severity and scale.Secondly, the content point is to study the key technology of the group accidents based oncase-based reasoning, including case representation, case retrieval, case revise and case retain,of which mainly case representation and case retrieval. Through the analysis of thecharacteristics about large amounts of group incidents, it is applied the f Object-orientedknowledge representation to describe the group accidents information. Based on thecharacteristics of the case retrieval, it is proposed two points：One is that feature attributes isdivided into the positive and fuzzy, of which the numerical type, the numerical interval type,enumeration type are to as positive feature attribute. And the text type feature attribute belongsto as fuzzy feature attribute. The other is that the paper applies Bayesian theory to modifyweights, which get experts consistency judgment matrix through the AHP, to realize theorganic combination of the subjective and objective weight, then to improve the accuracy ofthe knowledge acquisition.Finally, Based on the above theory, it is to design the intelligent decision support systemdesign on the group accidents. On the analysis of the general structure sets up user management subsystem, case-based reason subsystem and case base managementsubsystems, and its database table. In the process of case-based reason, inputting the relevantinformation of target case, and selecting the default attributes, calculating the weights. Thenwith the software interactive platform, calculating the similarity. At the end, to decide thesimilar case according to the similarity threshold by the user, or make plans based on domainknowledge.This research shows that it’s built intelligent decision support system based on theCase-based reasoning applied in group incidents, which it helps to make rapid response toemergency decision, and also inspire half structural and unstructured problems.