Potential trend discovery for highway drivers on spatio‐temporal data

Inter-city transportation plays an important role in modern cities, and has accumulated massive spatio-temporal data from various sensors by IoT (Internet of things) technologies. Travel characteristics and future trends of highway behind data are valuable for traffic guidance and personalized service. As a routine domain analysis, trend discovery for highway drivers faces challenges in processing efficiency and predictive accuracy. Insufficient profiles of those drivers are available directly, sensible executive latency on huge data is hard to guarantee, and inadequate features among spatio-temporal correlations hinder the analytical accuracy. In this paper, a travel-characteristic based method is proposed to discover the potential trend of payment identity for highway drivers. Considering time, space, subjective preference and objective property, travel characteristics are modeled on monthly data from highway toll stations, through which predictive errors can be reduced by gradient boosting classification. With real-world data of one Chinese provincial highway network, extensive experiments and case studies show that our method has second-level executive latency with more than 85% F1-score for trend discovery.