The continuous development and complexity of many modern cities offer many research challenges for urban scientists searching for a better understanding of mobility patterns that happen in space and time. Today, very large trajectory datasets are often publicly generated thanks to the availability of many positioning sensors and location-based services. However, the successful integration of mobility data still requires the development of conceptual and database frameworks that will support appropriate data representation and manipulation capabilities. The research presented in this paper introduces a conceptual modeling and database management approach for representing and analyzing human trajectories in urban spaces. The model considers the spatial, temporal and semantic dimensions in order to take into account the full range of properties that emerge from mobility patterns. Several object data types and data manipulation constructs are developed and experimented on top of an urban dataset testbed currently available in the city of Beijing. The interest of the approach is twofold: first, it clearly appears that very large mobility datasets can be integrated in current extensible GIS; second, significant patterns can be derived at the database manipulation level using some specifically developed query functions.