MODEL ANALITIK PREDIKTIF UNTUK KEDATANGAN WISATAWAN INTERNASIONAL DI INDONESIA DI TENGAH PANDEMI COVID-19 MENGGUNAKAN DATA INTERNET MULTISUMBER

• Main Author: Thalia Andariesta, Dinda
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/64186
• Institution: Institut Teknologi Bandung

The COVID-19 pandemic made a devastating impact on the tourism industry. Under uncertain circumstances, generating an accurate tourism demand prediction is critical to reinforcing foresight capabilities that may assist the government in better understanding tourism demand recovery and formulating appropriate policy. The rapid expansion of the Internet has drastically increased the popularity of online tourism platforms. It encourages the evolving availability of Internet tourism data, which has potential for the emerging tourism analytics domain. Correspondingly, this study presents machine learning models for predicting international tourist arrivals in Indonesia during the COVID-19 pandemic using multisource Internet data. This study analyzes tourist arrivals data from Indonesia Statistical Bureau and utilizes multisource Internet data from TripAdvisor travel forum and Google search engine. The data were analyzed from January 2017 to January 2022. This study employs four predictor variables: temporal factors (i.e., month and year), TripAdvisor (i.e., post, reply, and sentiment), Google Trends (i.e., twelve search volume indexes), and previous tourist arrivals. This study tests four combinations of predictors and develops the prediction model using three machine learning methods: artificial neural network, support vector regression, and random forest. The out-of-sample prediction is examined based on root mean square error, mean absolute percentage error, and mean absolute error. The findings of this study show the positive impact of combining multisource Internet data to improve prediction accuracy. The results also indicate that the superiority of the multisource Internet data is consistent across different machine learning methods. Finally, this study contributes in the following respects. First, this study pioneers the practice of a multisource Internet data approach in predicting tourist arrivals amid the unprecedented COVID-19 pandemic. Second, the use of multisource Internet data to improve prediction performance is validated with real empirical data, and the proposed model can be used to generate short-term tourism demand forecasts. Third, this is one of the few studies to provide perspectives on the current dynamics of Indonesia's tourism demand.