Taxis play an important role in offering comfortable and flexible service within Singapore's public transport system. Due to the inherent randomness in taxi service system, many taxi companies still rely on the drivers' experience to seek passengers. Today, Singapore's five taxi companies now use some form of wireless and GPS (Global Position System) satellite to track taxis traveling in urban area. GPS-equipped taxis can be viewed as ubiquitous mobile sensors which enable us to collect large amounts of location traces of individuals or objects. In this paper, we first investigate the characteristics of travel behavior of urban population. Next, a taxi business intelligence system is proposed to explore the massive transportation data based on spatial-temporal data mining techniques. Furthermore, various taxi business models are created to make comprehensive analysis on taxi business problems. Finally, the value of the taxi business intelligence system is demonstrated by applying it to some real-world scenarios. Results show that this system can significantly improve the quality of taxi services.