Research on Ontology Model of the Urban Critical Infrastructure System Interdependence

Shiying Wang, Xiang‐Yang Li, Yu Guang-huaProceedings of the 2018 3rd International Conference on Politics, Economics and Law (ICPEL 2018) 2018 · DOI

Abstract

The paper uses ontology method to describe complex interdependency in urban Critical Infrastructure System (CIS). Through the literature review, existing research results have not been excavated the essential causes and influencing factors of the urban CIS interdependency formation. Because the complex network interdependencies in the urban CIS has deep connotation, and the complexity increases with the urban CIS scale. In the research of urban CIS interdependency, attributes such as interdependent nodes and interdependent edges should be taken seriously also, and the data and information contained in them should be refined. The ontology theory can meet above requirements. The paper establishes an interdependency ontology model to generally express the urban CIS interdependency.

Introduction

Modern cities rely on Critical Infrastructure Systems (CIS)—such as power, water, and gas—that are increasingly intertwined. A failure in one system, like a power outage, often triggers "cascade reactions," stopping water pumps, disrupting gas transmission, and halting transit. These interdependencies are often physical, geographic, or logical.

Current research often fails to capture the granular "essential causes" of these interdependencies. Data is frequently siloed or lacks the attribute-level detail needed for accurate vulnerability assessment. This paper proposes a two-layer ontology framework to map these complex relationships, moving beyond simple network models to a formal "eABC" (extended Abstraction, Actuality, Temporality) ontology. This provides a standardized, machine-readable language for emergency managers to better predict and respond to disasters.

Methodology

The researchers utilize Ontology modeling, specifically using the Web Ontology Language (OWL) and the Protégé modeling tool. Their approach is structured into three main components:

1. The Interdependence Element Set ($DO$)

A formal set of eight elements is defined to describe any connection between infrastructures: timestamp ($T$), nodes ($V$), attributes ($A$), interdependence style ($R$), level ($H$), and rules ($X$).

2. Top-level Ontology (UCISOntology_Top)

Based on an extended version of the "ABC" model (Abstraction, Actuality, Temporality), the authors add a "Situation" concept. This layer represents objective existence independent of specific events, covering concepts like Actuality (physical objects), Temporality (disaster levels), and Situation (weather/geography).

3. Application Ontology (UCISOntology_Applied)

This layer refines top-level concepts for specific field knowledge, defined as a quaternion group: {Hazard, Body, Environment, Interdependency}. To handle diverse data types (precise numbers, intervals, or text), the authors introduce three additional elements: Class, Attribute, and Value.

The proposed ontology model illustrating the relationships between Critical Infrastructure, Situation, and Hazards.

Case Study: The 2014 Dalian Oil Pipeline Explosion

The framework was validated using a qualitative case study of the June 30, 2014, Dalian oil pipeline explosion. A drilling error led to an oil leak that flowed into a municipal sewage pipe network, causing fires and explosions in the Dalian Jinzhou development zone. The authors mapped the entities (oil pipeline, sewage network), hazards (leakage, fire), and environment (underground water pipelines) into their ontology to demonstrate how the model captures the "mesh" of interdependencies.

Key Insights and Results

  • Beyond Nodes and Edges: Traditional network theory often treats connections as simple lines. This paper argues that "interdependent edges" have deep connotations and specific attributes (like distance or physical overlap) that must be refined to understand risk.
  • Standardization for Interoperability: By creating a formal specification of shared conceptualization, the model helps solve the "information island" problem, allowing different departments (e.g., Power vs. Water) to use the same "dictionary" during a crisis.
  • Binary Transformation: To manage complexity, the authors demonstrate how multivariate interdependencies can be transformed into binary relationships within the ontology, making the data easier for automated processing.

Limitations and Future Work

While the paper provides a robust structural framework, it is primarily conceptual. Future work is needed to:

  • Demonstrate Predictive Power: The model has been applied to past events but requires testing for real-time predictive accuracy.
  • Automated Reasoning: Developing specific reasoning engines or algorithms that use this ontology to suggest emergency interventions.
  • Scalability: Determining if this manual ontological mapping can be automated or scaled to mega-city levels without becoming computationally prohibitive.

Summary updated

Abstract

The paper uses ontology method to describe complex interdependency in urban Critical Infrastructure System (CIS). Through the literature review, existing research results have not been excavated the essential causes and influencing factors of the urban CIS interdependency formation. Because the complex network interdependencies in the urban CIS has deep connotation, and the complexity increases with the urban CIS scale. In the research of urban CIS interdependency, attributes such as interdependent nodes and interdependent edges should be taken seriously also, and the data and info rmation contained in them should be refined. The ontology theory can meet above requirements. The paper establishes an interdependency ontology model to generally express the urban CIS interdependency.
BibTeX
@inproceedings{wang2018research,
  author = {Shiying Wang and Xiang‐Yang Li and YU Guang-hua},
  title = {{Research on Ontology Model of the Urban Critical Infrastructure System Interdependence}},
  booktitle = {Proceedings of the 2018 3rd International Conference on Politics, Economics and Law (ICPEL 2018)},
  year = {2018},
  doi = {10.2991/icpel-18.2018.115},
}
Figure