Understanding Tourism Behaviour Through Exploratory Data Analysis with Machine Learning on Search Engine Data: Case Study in Bangka Belitung Islands, Indonesia

Abstract

Analyzing tourist behaviour through Google search data offers a dynamic, real-time approach to understanding travel preferences. This study employs Exploratory Data Analysis (EDA) alongside machine learning techniques such as hierarchical clustering, Principal Component Analysis (PCA), Strength Variables Index (SVI), heat map generation, and correlation matrix analysis to explore the key tourism drivers in Bangka Belitung. These drivers are categorized into demand-side factors—including tourist preferences, curiosity, seasonality, and economic conditions—and supply-side factors, such as transportation, accommodation, activities, pricing, culinary tourism, and local attractions. The findings reveal that transportation and accommodation consistently emerge as the most influential drivers in both regions, highlighting the importance of accessibility and lodging availability. Bangka emphasizes culinary experiences and price sensitivity, while Belitung is more influenced by economic conditions and seasonality. Peak tourism periods are identified during Chinese New Year in February, New Year, and mid-year school holidays in June to July. In Belitung, culinary tourism and seasonal activities see increased interest during February and October, while Bangka shows steady interest in beach-related activities and culinary offerings throughout the year. . Misalignment between supply-side factors, such as limited affordable accommodation or transportation options, can impact tourism performance during these periods. These insights offer practical recommendations for local governments, tourism boards, and businesses to refine marketing strategies, enhance tourist experiences, and optimize tourism infrastructure. . Focusing on affordable travel and culinary experiences for Bangka seasonal tourism and economic preferences for Belitung can help maximize tourism potential and drive sustainable growth in the region