A context awareness and prediction support system for efficient management of U-City

As sensor-related technology advances a variety of sensor data, interaction, information sharing and interchange among relevant systems have become more active. We also need a system that can either prevent crimes by forecasting context and coping with crime results as well as context management of city for high level of safety for city life with efficiency. In this dissertation a context awareness and prediction system are presented for more efficient and advanced management of u-City based on an ontology modeling for utilization of huge amount of information involved. Inference rules and facts are presented and generated so they can be effectively applied to contexts of u-City. The mechanism realizes higher accuracy of context prediction of events which can be predictable by existing history information. Especially, the system can be applied distinctively on each function of u-City to the ontology models with information transferred from sensors to the legacy system. The results of the proposed system can be used as practically useful references on customized context awareness, inference, mining and prediction that can support a efficient responding methods according to u-City modeling, definition, and inference rule of complicated context information.