Community mobility during Covid-19 pandemic and tourism performance: Data mining approach

The movement control policy imposed by the government worldwide has changed the community mobility during the Covid-19 pandemic. In addition, the limitation of flights and the visiting policy has nearly stopped visitors to tourist destinations. This study contends that the community mobility change in a region during a pandemic relates to the region's tourism-related performance before the pandemic. Data mining approach with CRISP-DM as a framework and Knime Analytics Platform as a tool are used to analyze data on 34 Indonesian provinces. The study aims (1) to present the nature of community mobility fluctuation at the tourism-related area, (2) to group provinces based on the similarity in mobility fluctuation and tourism-related performance, and (3) to characterize provinces across the tourism-related performance. Data are collected from Google's community mobility covering mobility change in retail-and-recreation areas, parks, and transit-and-station as a time series for all provinces. In addition, tourism-related indicators are collected from the Indonesian statistics agency covering length-of-stay and occupancy rates for starred and non-starred hotels. Among three tourism-related areas, transit-and-station experience the highest mobility fluctuation in a decreasing direction. The main finding shows that six provinces with higher visitor length-of-stay and hotel occupancy rates experience greater mobility change. Bali, Yogyakarta, and Jakarta are well-known as domestic and international tourist destinations; North Sulawesi with Bunaken National Marine Park, West Papua with Raja Ampat, and Riau Islands are also popular tourist spots. The result implies that those regions may suffer a higher impact on tourism. This study contributes to the application of data mining to reveal information on publicly available socio-economic indicators.