Research On Association Rules Mining In Emergency Case Scenarios And Inference

The outbreak of emergency is characterized by occurring suddenly,complexity and uncertainty of evolution,causing huge damage,that it is difficult to accurately grasp the evolution trend of emergency and make scientific decision in short time,complexity environment and huge pressure situation,which leads to the traditional “predict-response” mode out of control,and the “scenario-response” mode has become an basic paradigm to response emergencies effectively.At present,emergency occurs frequently,a large number of historical emergency cases is accumulated,these emergency case scenarios recorded the whole process of occurrence,development and evolution of emergencies,and there are large amounts of evolution law and rule knowledge in the emergency case scenarios,these rules are the key basis for the decision maker to make quick and scientific decisions.But the real time information of emergency acquired are uncertain and fuzzy,while according to those rules,it's difficult to obtain an accurate trend output of the emergency using conventional matching reasoning method.Therefore,how to dig out the association rules between emergency case scenarios and take advantage of effective reasoning method to realize emergency scenario evolution reasoning is the scientific problem of improving the efficiency and effectiveness of emergency decision-making,which need to be solved.To improve the understanding of evolution law in the emergency case scenarios,find out a large number of rule knowledge and effectively solve the problem that the real time information are uncertain and fuzzy in the reasoning so that to better meet the needs of emergency “scenario-response” model,this paper proposes an method of association rules mining in emergency case scenarios and inference.Firstly,the association rules mining in emergency case scenarios is studied,from the perspective of knowledge management and system theory,representing the emergency case scenarios uniformly with knowledge element,it divides the emergency cases into several scenarios with a representation of knowledge element according to the state change of the hazard bearing body,and constructs the scenario sequence matrix with “antecedent scenario-result scenario” form,and transforms the scenario sequence matrix into candidate input and output matrix of association rules,introduces the theory of concept lattice,creates the corresponding decision formal context,and come to extracts the implicit association rules in these emergency case scenarios,combined with the expert experiences and domain knowledge to modify and supplement the rules.Secondly,the reasoning process of the emergency scenario evolution based on belief rule-base is studied.it calculates the weights of the concept elements of association rules based on the statistic method,and transforms the association rules into confidential rules.Taking effective method to transform the emergency real-time scenario information into confidence form,using the evidential reasoning algorithm to reason and fusion the belief rules,and realizing the reasoning of emergency scenario evolution.Finally,the feasibility and effectiveness of the proposed method is verified by the case of forest fires.This paper puts forward an method of association rules mining in emergency case scenarios and inference based on belief rule-base,it divides the emergency cases into several scenarios,identifies and explores the implied rule knowledge between emergency scenarios from the attribute micro level,which improves the application value and reusability of historical emergency cases,establishes the knowledge base for realizing the reasoning of emergency scenario evolution.It takes advantage of the evidential reasoning algorithm,effectively overcomes the shortcomings of the uncertainty and fuzzy emergency scenario real-time information,makes the reasoning result more accurate,greatly enhances the predictability of the emergency evolution trend,thus it can provide important guidance and support for the decision maker to make a quick response and to develop a scientific emergency decision-making program.