Application-oriented Context Modeling and Reasoning in Pervasive Computing

Context-awareness is considered as a key problem in designing more adaptive applications in pervasive computing community. Context modeling and reasoning, which deal with high-level abstraction and inference of pervasive contextual information, are important research areas of context-awareness computing. Efforts have been put into these two areas and several prototype systems have been proposed. However, the huge amounts of contextual information in pervasive environment make existing systems inefficient, even useless. In this paper, we propose enhanced application-oriented context modeling and reasoning (EACMR) system to deal with this problem. In EACMR system, we develop a context ontology (ACMRONT) to model the contexts in the pervasive computing. Different from previous researches, a context filter is designed to classify contexts by their importance. To promote the performance of the system, the reasoner in EACMR only deals with the application-related contextual information rather than all the available contextual information. Experiments about EACMR system demonstrate its higher performance than those of previous systems