Assessing Port Safety Incidents through Causality Networks and Simulation Modeling

초록

Port operations involve heavy machinery and hazardous environments, resulting in frequent safety incidents that must be systematically reported and managed. This study aims to convert unstructured incident reports into a quantitative framework for analyzing how safety risks propagate through port operations. Causal factors are extracted from incident narratives using natural language processing and represented as directed causality networks. A Monte Carlo simulation model is subsequently constructed to quantify both the marginal and cascading effects of individual causal factors. The proposed approach introduces a novel integration of text？based causal inference and simulation？based risk modeling, enabling a more rigorous assessment of port safety. Empirical results demonstrate the approach’s effectiveness in identifying high？impact causal factors and informing targeted safety interventions.

키워드