Visualizing tourist behaviour in Melaka using HDBSCAN and multiple linear regression

This research offers a visual solution for patternizing Melaka's tourist attractions based on Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) and Multiple Linear Regression (MLR). Conventional techniques for patternizing tourists' behavior, e.g., surveys, are time-consuming with a limited sample size. To address this, the current research utilizes geo-tagged photo data and past tourist arrival data to patternize travel behavior and forecast tourist arrival. HDBSCAN clusters tourist attractions effectively, revealing areas of density, and MLR accurately predicts tourist visits according to influence factors. The system is realized as a web application, offering interactive heatmaps and trend analysis using visualization methods like line charts, bar charts, pie charts, and scatter plots. The results of the usability test show high satisfaction, which reflects the system's potential for supporting tourism agencies and businesses in making decisions. This combination of machine learning and data visualization enables more effective, data-driven management of tourism in Melaka.