基于粗糙集的應(yīng)急案例中概率規(guī)則挖掘方法

摘要： 應(yīng)急案例作為描述突發(fā)事件發(fā)生、發(fā)展及應(yīng)對(duì)過(guò)程的文本,蘊(yùn)含了潛在的規(guī)律與寶貴的經(jīng)驗(yàn)。為了挖掘應(yīng)急案例中各要素間潛在的關(guān)聯(lián)關(guān)系,構(gòu)建出基于粗糙集的應(yīng)急案例中概率規(guī)則挖掘方法。首先,構(gòu)建出應(yīng)急案例知識(shí)五元組,描述應(yīng)急案例共性特征,并將諸多應(yīng)急案例信息組織成一張應(yīng)急案例決策表;然后,應(yīng)用遺傳算法對(duì)應(yīng)急案例決策表進(jìn)行屬性約簡(jiǎn),進(jìn)而獲取概率規(guī)則;最后,以大興安嶺林區(qū)50起重特大火災(zāi)案例為例,闡述方法的具體執(zhí)行過(guò)程,并通過(guò)兩組測(cè)試實(shí)驗(yàn)證明了方法的可行性和有效性。該方法描述了應(yīng)急案例的共性本體特征,具有較高的可重用性,有利于為決策者采取應(yīng)急管理措施提供決策支持。
As texts describing the process of occurrence,development and response of emergencies,emergency cases contain underlying laws and valuable experience. In order to mine the potential relationship among the elements in the emergency cases,the probability rule mining method based on rough set is developed. First,we build a knowledge quintuple to describe the common features of emergency cases,and the information of emergency cases are organized into a decision table; then we use genetic algorithms to reduce the attributes of the decision table of emergency cases,and thus the probability rules can be acquired. Finally,taking the case of 50 cases of heavy fire in Daxinganling forest area as an example,the concrete implementation process of the method is expounded,and two groups of testing experiments prove the feasibility and validity of the method. This method describes the common ontology characteristics of emergency cases,having high reusability,and it is helpful to provide decision support for decision-makers to take emergency management measures.