Characteristics analysis of intercontinental sea accidents using weighted association rule mining: Evidence from the Mediterranean Sea and Black Sea

• Published in: Ocean engineering 2023-11, Vol.287, p.115839, Article 115839
• Main Authors: Huang, Daozheng, Liang, Tiantian, Hu, Shenping, Loughney, Sean, Wang, Jin
• Format: Article
• Language: English
• Subjects: Accidents prevention; Association rules; Data mining; Marine traffic; Weighted association rule mining; Weighted transaction data
• ISSN: 0029-8018; 1873-5258
• DOI: 10.1016/j.oceaneng.2023.115839

With the continuous growth of international trade and the number of ships, the pressure on marine safety is increasing. To strengthen the efficiency and pertinence of accident prevention and control measures, in-depth research on the characteristics of marine traffic accidents is of great significance. This study uses a Weighted Association Rule Mining (WARM) approach to investigate the association between marine traffic accidents characteristics. The Mediterranean Sea and Black Sea play a crucial role in global maritime transportation. The marine traffic accidents in the Mediterranean Sea and Black Sea region from 2006 to 2020 are collected from the Lloyd's List Intelligence (LLI) database. Then, WARM is applied to obtain the characteristics of marine traffic accidents. The findings show that Flag of Convenience (FOC) vessels in the Mediterranean have a lower accident rate than non-FOC vessels do. Moreover, the results reveal a strong relationship between accidents and the age, gross tonnage, and ship type. Older general cargo ships with a gross tonnage between 500 and 3000 tons are more prone to accidents. This research provides insights for authorities to develop targeted measures for preventing specific types of accidents and enhancing marine safety. •In this study, the main objective was to investigate maritime accident characteristics and improve navigation safety.•It is fully considered that in the association rule approach, different accidents may have their own weights.•The weighted association rule approach is employed to mine the causal characteristics of ship traffic accidents.•The results of the study reveal a strong relationship between the age, gross tonnage, and ship type of the accidents.