Tourism demand forecasting using tourist-generated online review data

Abstract

This study aims to forecast international tourist arrivals to Hong Kong from seven English-speaking countries. A new direction in tourism demand modeling and forecasting is presented by incorporating tourist-generated online review data related to tourist attractions, hotels, and shopping markets into the destination forecasting system. The main empirical findings indicate that tourism demand forecasting based on tourists’ online review data can substantially improve the forecasting performance of tourism demand models; specifically, mixed data sampling (MIDAS) models outperformed competing models when high-frequency online review data were included in traditional time-series models.