Applying Cluster Analysis for the Investigation of Travel Behavior and User Profiles

Abstract

Urbanization leads to a surge in demand for transportation and infrastructure improvements. In this context, understanding and optimizing travel behavior are crucial for effective transportation planning. This research investigates travel behavior patterns and user profiles in the realm of urban mobility. The study adopts an approach utilizing real-world data from an activity-based dataset collected through a survey. The methodological framework is characterized by a multi-step process which includes data preprocessing, cleaning, and aggregation, as well as principal component analysis and k-means cluster analysis with inertia evaluation for an optimal number of clusters. The cluster analysis unveils seven distinct clusters. Stability lovers are elderly people who prefer public transport, happiness seekers are attraction-driven car users, weekend shoppers, park goers, and sports practitioners rely on their cars for their activities, too. Furthermore, inflexible travelers value the service quality and "routine enthusiasts" stick to travel routines. Notably, bicycle usage prevails among stability lovers and routine enthusiasts, while shared transportation gets little attention in any of the clusters. By recognizing the adaptability of this methodology to specific city contexts, current research provides a way to understand travel behavior thus offering valuable insights for informed transportation policy planners.