Evaluation of the anchorage area selection by using association rule mining

Abstract

Anchorage area selection is conducting by ship masters by considering various factors. Some of these factors are distance from anchorage area to berth, water depth, density in the anchorage area, meteorological condition, Vessel Traffic Service (VTS) suggestion etc. Within this context, to analyze the anchorage area selection behavior under different conditions, Automatic Identification System (AIS) data for the year of 2023–2024 and wind data are used.

The research is conducted in the Nemrut Bay and Çandarlı Bay, on the western side of the Türkiye, where several industrial ports and terminals are located. There are eight anchorage areas in the area and these eight-anchorage areas are used by vessels carrying dangerous cargo, vessels carrying non-dangerous cargo and the vessels that will approach the breaking yards. The aim of this study is to reveal the hidden and meaningful relationships between anchorage area selection, vessel related and weather-related factors by using association rule mining (ARM) and Cramer’s V analysis. While the ARM results show the strong and frequent association between variables, Cramer’s V quantifies the strength of association between these categorical variables. These two approaches highlight the anchorage area selection patterns of ship masters and explain relationships among variables.

Studies about anchorage areas are mainly focused on the anchorage location selection. Novelty of this study is to show the anchorage area selection patterns of ship masters under various conditions with ARM and Cramer’s V analysis. Results of this study show the relationships among variables during the anchorage area selection process.