With the acceleration of urbanization, cascading disaster risks (CDR) as a typical risk mode have become the main threat to cities. After experiencing several cascading disasters, such as typhoon Lekima, how to clarify the basic features of CDR and achieve risk modelling has turned to be increasingly significant for building resilient city. However, the complexity of CDR brings about the difficulty to quickly map such risk mode depending entirely on expertise. Therefore, this paper attempts to provide a CDROntology system built by concepts and relations, and make full use of the historical cases to drive the modelling of the target CDR with case‐based reasoning. Firstly, we describe the basic structure and content of CDR and give a three‐level CDROntology system with the explanation of modelling primitives. Then, taking CDROntology system as the basis, a case‐driven selection process is proposed to provide the modelling source for the target CDR. Furthermore, set covering and manual correction methods are adopted to model the evolutionary risk chain and the specific risk scenario of the target case. Finally, a case study is given to illustrate the use of CDROntology system and case‐driven method for building a predictive risk model in typhoon‐triggered cascading disasters.

中文翻译：

---

改善级联灾害的应急准备：案例驱动的风险本体建模

随着城市化进程的加快，级联灾害风险（CDR）作为一种典型的风险模式已经成为城市的主要威胁。在经历了几次级联灾难（例如台风莱基玛）之后，如何弄清CDR的基本特征并实现风险建模对于建设具有韧性的城市变得越来越重要。但是，CDR的复杂性带来了难以完全依赖专业知识来快速绘制此类风险模式的困难。因此，本文试图提供一种由概念和关系构建的CDROntology系统，并充分利用历史案例，以基于案例的推理来驱动目标CDR的建模。首先，我们描述了CDR的基本结构和内容，并给出了三​​级CDROntology系统以及建模原语的解释。然后，以CDROntology系统为基础，提出了一个案例驱动的选择过程，为目标CDR提供建模源。此外，采用覆盖和手动更正方法来建模演化风险链和目标案例的特定风险方案。最后，通过案例研究说明了CDROntology系统和案例驱动方法在台风引发的级联灾害中建立预测风险模型的应用。