Context recognition from incomplete situation with uncertainty management

Context is a key factor for providing intelligent service in ubiquitous computing environment. The usefulness of context is based on its correctness. As the context plays a crucial role in service performance, it is necessary to guarantee the reliability of context. But, it is almost impossible to scan the full aspects of the environment for the limited computing capability and unlimited modality of reality. In this paper, we propose a method of recognizing context from incomplete situation with uncertainty management. In order to make the supposition for the missing data, probabilistic inference method is employed and to manage the uncertainty of information, fuzzy logic is employed.