An Event-driven Context Model in Elderly Health Monitoring

Elderly health-monitoring systems provide various proactive services according to context. Context plays an essential role in semantic understanding of human activities from sensor data. We present a novel event-driven context model for elderly health-monitoring application, which supports a distributed system with multimodal sensors. The context model is a dynamic hierarchical structure modeling human actions, activities, and environment information in home scenario. Context awareness is achieved by detecting events associated to the context hierarchy. Common knowledge ontology is employed to guide high-level context reasoning. Experiments of this method in elderly health monitoring show the effectiveness of the proposed context model.