Modeling International Tourist Arrivals: An NLP Perspective

This paper develops a novel regression framework to estimate international tourist arrivals in 37 countries from the Organization for Economic Co-operation and Development (OECD) countries by combining significant socio-economic-environment features and a natural language processing (NLP) based social media index. The index is developed by fine-tuning a pre-trained BERT model using the reviews of different countries collected from TripAdvisor to generate a tourist feedback score, which is used as an additional feature with the other OECD features for tourism arrival estimation using an adaptive boosting regression technique. The outcomes demonstrate that the proposed framework performs reasonably well than other relevant regression techniques. The findings of this study can be utilized to make decisions that support the growth of sustainable tourism.