A unifying framework for situation identification methodologies

Situation identification methodologies in pervasive computing aim to abstract low level context data into more meaningful high level contexts for use by context-aware application developers and users. Many situation identification techniques have been developed and successfully applied to limited scenarios. It is not possible to apply a single technique to a wide range of diverse applications. An ideal solution would allow the combination and and interchanging of techniques as appropriate. We propose a unifying framework for existing situation identification methodologies that will automatically select the best techniques for an application.