Problem that frequently occurred in taxi company is occupation and fleet distribution. Sometimes many passengers are not served, meanwhile also many unoccupied taxi. The problem occurs due to dispatching taxis is not matched with the taxi demand in different area during different time periods. Nowadays, taxi equipped with GPS and taximeter that continuously produce huge spatio-temporal data. This data can be used to help taxi problem. In this paper, we propose graph based analysis for GPS taxi data and apply graph clustering to discover frequent origin-destination flow. We also build a data visualization tools for information analysis purpose. The experimental results show that taxi demand and frequent origin-destination characteristic is varies over time.
Discovering frequent origin-destination flow from taxi GPS data
2016-10-01
1836223 byte
Conference paper
Electronic Resource
English
Urban traffic zone division method based on urban taxi origin-destination flow clustering
European Patent Office | 2024
|Contextualized Spatial-Temporal Network for Taxi Origin-Destination Demand Prediction
ArXiv | 2019
|