中文摘要：

利用因子分析法对日本2005年统计数据进行了分析。从13个常用火灾统计指标中提取出火灾后果总量、火灾发生概率、平均死亡人数及平均经济损失4个公因子，并采用多元线性回归方法对若干社会经济因素与4个公因子建立多元线性模型。根据模型的判定系数和各自变量的偏相关系数，分析了若干社会经济因素与4个火灾风险公因子的相关性并对重要度进行排序。结果表明，4个公因子意义明确且相互独立，可全面表征城市火灾风险程度。基于分析数据，火灾后果总量公因子与人口总量呈显著线性相关；火灾发生率公因子主要受人员结构、人口流动性及居住条件3个方面影响，与人口总量、经济水平、教育水平之间不存在线性相关性；平均死亡人数公因子随消防训练提高而显著降低；平均经济损失公因子与选用的任何一个社会经济因素之间都没有线性相关性。

英文摘要：

The present paper is aimed to find some factors which can comprehensivcly characterize the urban fire risk and then analysis their relationship with social-economic factors. Four common factors for urban fire risk characterization, which including fire total damage, deaths on avcragc, loss on average and fire probability, have been derived from 13 frequently-used fire statistical indicators by means of factor analysis. These common factors are found to be meaningful and independent from each other. Furthermore. the relationships between the common factors and a number of socio-economic factors are analyzed by the method of multiple linear regressions. Based on the data analysis, it can he concluded that the factor of fire total damage shows a significant linear correlation with the population. The factor of fire probability is mainly influenced by three factors including the proportion of children population, the population mobility and the living conditions. In addition, there is no linear correlation between the population, the eeonotnic level and the educational level. The factor of deaths on average would be reduced according to the increase of fire training level and there is no linear correlation between the factor of loss on average and any socio-economic factors considered in this paper.