Tool support for designing CML based context models in pervasive computing
Johnson Fong, Jadwiga Indulska, Ricky Robinson · International Conference on Pervasive Services · 2010
of context-aware applications is very complex as context information is gathered from a variety of sources that are error prone and differ in the quality of information they provide. The current software engineering approach to building context- aware applications employs appropriate context modelling and reasoning techniques to facilitate the application development process. However, the design of context models is highly challenging, as they need to capture (i) a variety of context information required by the applications (such as location, activities and devices, etc.), (ii) the relationships and dependencies among them, and (iii) their ambiguity, timeliness, and (iv) required Quality of Information. This paper describes a semi- automatic, forward engineering context modelling tool that provides design support for developing context models designed using the Context Modelling Language (CML). It promotes reuse of existing context information and automatically generates code for mapping context models into their runtime representations. The goal is to enable an error resistant context model design process and significantly reduce applications development effort and time.