论文部分内容阅读
利用关联规则挖掘算法及Clementine挖掘工具定量分析了火灾统计中起火场所、火灾原因和起火时间(月份、时间区间)之间的关联性。在采用关联规则表征参数的基础上,引入规则平均直接财产损失(人员死亡)比和规则总直接财产损失(人员死亡)比来表示不同规则损失的大小。以北京市2000—2006年火灾统计数据为例,对其火灾发生和人员死亡数据库进行挖掘,得到起火场所、火灾原因及起火时间(月份、时间区间)之间的1-项集、2-项集和3项集关联规则。不同起火场所、时间区间、月份和火灾原因下得出的火灾发生频率和损失分析结论可为消防管理部门有针对性地采取消防监督管理措施和有效提高消防执勤战备提供指导。
Association rules mining algorithm and Clementine mining tools were used to quantitatively analyze the correlation between the fire place, the cause of fire and the fire time (month, time interval) in the fire statistics. Based on the use of association rules to characterize the parameters, the ratio of direct average loss of property (death) to the total direct loss of property (death of persons) is introduced to indicate the size of different rules. Taking the fire statistics of Beijing from 2000 to 2006 as an example, the database of fire occurrences and deaths was excavated, and the 1-itemsets, 2-items of the fire place, the cause of the fire and the time of fire (month and time) Set and 3 set association rules. The analysis of the frequency of fire occurrence and loss under different fire places, time intervals, months and fire causes can provide guidance for the fire control department to take targeted fire supervision and management measures and effectively improve the fire fighting duty readiness.