Along with more and more people and cars appear in cities, the stricter requirements, like traffic efficiency, safety, service and cleanness cannot be met easily by using the existing urban traffic plan and operation management methods, which are normally supported by the separated and partial Intelligent Transportation System (ITS). In this paper, big data platform for urban public transportation is proposed, and its architecture and subsystems are developed. Then, Guangzhou's typical APP services for urban public transportation are described in detail, including: real-time bus service, taxi inquiry and call service, mobile payment, and traffic information service, etc.