A Data Management Platform for Taxi Trajectory-based Tourist Behavior Analysis

Taxis are one of the transportation types that tourists usually choose when visiting unfamiliar cities. Using GPS devices installed in taxis is an efficient way to collect a large amount of movement data of individual tourists. The main problem of understanding tourist movement behavior is how to extract tourist trajectories from raw taxi GPS data, which consist of vehicle ID, geo-coordinates, times-tamp, direction, speed, and hired status etc. This study proposes a data management platform and a data pipeline process that can be scaled to support a large volume of taxi trajectory data and provide the data integration module enriched with a tourist trajectory knowledge base. To construct such knowledge base, we reuse and extend two existing ontologies: Semantic Trajectory Ontology (STO) and Mobility Behavior Ontology (MBO). The results of this study can be used to analyze tourist movement behaviors and hence enabling governments and tourism businesses to gain insight understanding of tourist movements and activities. It can also support a prediction of tourism trends in the future.