The intercontinental liner shipping services transport containers between two continents and they are crucial for the profitability of a global liner shipping company. In the daily operations of an intercontinental liner shipping service, however, container slot bookings from customers can be freely cancelled during a booking period, which causes loss of revenue and low utilization of ship capacity. Though a pain-point of the liner shipping industry, the container slot cancellation problem has not yet been well investigated in the literature. To fill this research gap, this study aims to estimate the probability for the cancellation of container slot booking in the long haul transports of the intercontinental liner shipping service by considering the primary influential factors of cancellation behavior. To achieve the objective, a container slot booking data-driven model is developed by means of a time-to-event modeling technique. To incorporate the effect of booking region on the cancellation probability, we introduce the frailty term in the model to capture the regionality of the container shipping market. Our case study with real slot booking data shows that the developed model performs well in forecasting the loaded containers of the slot booking requests. In addition, we shed light on how the internal factors of slot booking and external factors of shipping market influence the probability of cancellation.

洲际班轮运输服务在两大洲之间运输集装箱，对于全球班轮运输公司的盈利能力至关重要。然而，在洲际班轮运输的日常运营中，客户的集装箱舱位预订在订舱期内可以随意取消，造成收入损失和运力利用率低下。集装箱舱位取消问题虽然是班轮运输业的一个痛点，但尚未在文献中得到很好的研究。为了填补这一研究空白，本研究旨在通过考虑取消行为的主要影响因素来估计洲际班轮运输长途运输中集装箱订舱取消的概率。为了实现这一目标，通过事件时间建模技术开发了集装箱舱位预订数据驱动模型。为了考虑订舱区域对取消概率的影响，我们在模型中引入脆弱项来捕捉集装箱运输市场的区域性。我们对真实舱位预订数据的案例研究表明，开发的模型在预测舱位预订请求的装载集装箱方面表现良好。此外，我们还阐述了订舱的内部因素和航运市场的外部因素如何影响取消概率。