Development of online travel Web scraping for tourism statistics in Indonesia

Introduction. This research aims to study a novel approach to producing tourism statistics, especially accommodation statistics, in Indonesia using scraping of online travel agent Websites. Method. Accommodation data (e.g., room availability and price) were gathered from two of the largest online travel agencies in Indonesia. All data were collected automatically from the sites’ URLs listed in the sitemap. Analysis. The data were collected daily from 6 March to 27 July 2019. Datasets from the two Websites were merged. The room occupation rate (ROR) for each province was calculated and compared with the official statistics from Statistics Indonesia. Results. The results show that the online room occupancy rates and official statistics have a similar pattern indicating the use of the Web scraping technique provides valuable information, to measure the room occupation rate with an advantage in terms of cost and collection time. Conclusions. It is feasible to use big data as a proxy of or a complement to official statistics, especially in tourism statistics. By using the Web scraping technique, the indicator that usually requires significant time and cost can be done in real-time and less cost. This new approach would improve the quality of tourism statistics produced by BPS Statistics Indonesia.