Research and implementation of the context-aware middleware for controlling home appliances

Smart homes integrated with sensors, actuators, wireless networks and context-aware middleware will soon become part of our daily life. The paper describes a context-aware middleware providing an automatic home appliance control based on a user´s preference. The context-aware middleware utilizes 6 basic data for learning and predicting the user´s preferences in multimedia content: the pulse; body temperature; facial expression; room temperature; the time; the location. The test results show that the pattern of an individual´s preferences can be effectively evaluated and predicted by adopting the proposed context model.