Exploring satisfaction with air-HSR intermodal services: A Bayesian network analysis

Air and high-speed rail (HSR) intermodal service (AHIS) breaks through the barriers of aviation and HSR, which builds a modern integrated transportation system. However, this system also poses a challenge to operators to provide satisfactory travel services for passengers. This paper aims to identify the service indicators that influence travelers' overall satisfaction with AHIS and the relationships between them based on research data acquired from a passenger behavior survey at Shijiazhuang Zhengding International Airport (SJW) in 2019. First, a Bayesian network (BN) is constructed by integrating the greedy thick thinning (GTT) algorithm with expert knowledge. Then, sensitivity analysis and overall satisfaction prediction are conducted to determine the correlation and influence effect between service indicators and overall satisfaction. The research findings are as follows: (1) Compared to a binary logit model, the Bayesian network shows high fitting and prediction accuracies. (2) Transfer time is negatively correlated with satisfaction, for AHIS with the same total travel time, travelers tend to choose services with less transfer time since this choice increases their satisfaction. Interestingly, passengers are more tolerant of the travel time of airline than HSR. (3) Service indicators such as real-time information, arrival punctuality and ticket price have the highest sensitivity values for overall satisfaction. The results can provide useful suggestions for AHIS operators.