Estimating Tourist Arrivals by User Generated Content Volume in Periods of Extraordinary Demand Fluctuations

Abstract

In extraordinary situations, like the Covid-19 pandemic, irregular demand fluctuations can hardly be predicted by traditional forecasting approaches. Even the current extent of decline of demand is typically unknown since tourism statistics are only available with a time delay. This study presents an approach to benefit from user generated content (UGC) in form of online reviews from TripAdvisor as input to estimate current tourism demand in near real-time. The approach builds on an additive time series component model and linear regression to estimate tourist arrivals. Results indicate that the proposed approach outperforms a traditional seasonal naïve forecasting approach when applied to a period of extraordinary demand fluctuations caused by a crisis, like Covid-19. The approach further enables a real-time monitoring of tourism demand and the benchmarking of tourism business in times of extraordinary demand fluctuations.