Hierarchical relationships analysis: A novel data-driven framework to assess the influential factors of marine accidents

Feng, Y, Wang, X ORCID: 0000-0002-7469-6237, Cao, Y ORCID: 0009-0000-2246-1367, Zhou, J, Wang, J ORCID: 0000-0003-4646-9106 and Yang, Z ORCID: 0000-0003-1385-493X (2026) Hierarchical relationships analysis: A novel data-driven framework to assess the influential factors of marine accidents. Reliability Engineering & System Safety. ISSN 0951-8320

Preview

Text Hierarchical relationships analysis -A novel data-driven framework to assess the influential factors of marine accidents.pdf - Published Version Available under License Creative Commons Attribution. Download (2MB) | Preview

Abstract

Marine accidents are associated with multiple risk influential factors (RIFs), whose interrelationships complicate risk analysis. Existing approaches rarely connect associations among factor states with data-driven influence analysis and weighted hierarchical interpretation. To address above gaps, this study proposes a combined hierarchical analysis framework integrating Combined Association Rule Mining (CARM), data-driven Decision-Making Trial and Evaluation Laboratory (DEMATEL), and a maximum mean de-entropy-based Weighted Total Adversarial Interpretive Structural Model (MMDE-WTAISM). CARM aggregates associations among RIF states into directed factor-level relationships using an expectation-based combined-conviction metric. DEMATEL quantifies direct and indirect network influence with a data-driven transform method. MMDE-WTAISM selects an adaptive threshold and incorporates DEMATEL relationship strengths into weighted indicators and complementary UP-type and DOWN-type hierarchies. The analysis generated 826 factor-level rules, retained 120 direct relationships, and identified 258 reachable relationships, two feedback clusters, and four computational levels. In the full sample, ship certificates, seafarer certificates, manning, regulation, and education and training had the five highest weighted driving-power values. A 1,000-replicate bootstrap of the complete analytical pipeline supported the sampling stability of the principal results. The framework provides a reproducible route from associations among factor states to weighted bidirectional hierarchies without expert scoring, supporting targeted maritime safety governance and operational risk control.

Item Type:	Article
Uncontrolled Keywords:	40 Engineering; 49 Mathematical Sciences; 35 Commerce, Management, Tourism and Services; 01 Mathematical Sciences; 09 Engineering; 15 Commerce, Management, Tourism and Services; Strategic, Defence & Security Studies; 35 Commerce, management, tourism and services; 40 Engineering; 49 Mathematical sciences
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) V Naval Science > V Naval Science (General)
Divisions:	Engineering and Built Environment
Publisher:	Elsevier
Date of acceptance:	21 August 2026
Date of first compliant Open Access:	1 September 2026
Date Deposited:	01 Sep 2026 09:33
Last Modified:	01 Sep 2026 09:33
DOI or ID number:	10.1016/j.ress.2026.113356
URI:	https://researchonline.ljmu.ac.uk/id/eprint/29229

View Item