Rapid development of the city makes it faces increasing challenges in controlling fire accidents. In order to improve the effectiveness in urban fire control, the distribution of urban fire risk should be identified accurately. Accordingly, a new method for urban fire risk assessment is proposed in this study. The method is based on the points of interest (POIs) data. The original POIs need to be sorted and reclassified because they are commonly repeatable and unreasonably classified. After reclassification, the weight of each category of the POIs in terms of fire risk is determined based on the historical fire accident data and experts' judgment. Then, the kernel density analysis (KDA) is conducted on all categories of the POIs by using ArcGIS software. After that, the distribution of the fire risk can be obtained and visualized based on the weights and KDA results of all categories of the POIs. Moreover, Zhongyuan District of Zhengzhou city is taken as an example to verify the effectiveness of the proposed method. The results indicate that the high fire risk areas are mainly located in southwest of the target district. Accordingly, specific measures are proposed to mitigate fire risks in the referred region.

城市的快速发展使其在控制火灾事故方面面临越来越大的挑战。为了提高城市消防的有效性，需要准确识别城市火灾风险的分布。因此，本研究提出了一种新的城市火灾风险评估方法。该方法基于兴趣点 (POI) 数据。原始 POI 需要进行排序和重新分类，因为它们通常是可重复和不合理分类的。重新分类后，根据历史火灾事故数据和专家判断，确定各类别兴趣点在火灾风险方面的权重。然后，使用ArcGIS 软件对所有类别的POI 进行核密度分析（KDA）。在那之后，基于所有类别POI的权重和KDA结果，可以获取并可视化火灾风险的分布。并以郑州市中原区为例，验证所提方法的有效性。结果表明，火灾高风险区主要位于目标区的西南部。据此，提出具体措施，以减轻所指地区的火灾风险。