Abstract:
Computer technology has gradually become part of our daily life. This leads to a concept called ubiquitous computing or pervasive computing where computers are used around us seamlessly and, without notice, users may not realize that they are using computing systems. Within pervasive computing community, context-aware applications are gaining more interests since these applications can adapt their behavior based on surrounding context and environment. By definition, contexts refer to any information that characterizes the situation of an entity of interest. An entity can be a person, place, object, or application. Context information can be either (1) retrieved or sensed by hardware or software, (2) explicitly provided by the users, or (3) derived from other context information, e.g. by means of computation or of composition of other contexts into a composite context. This research proposes a concept of Process-Based Composite Contexts in which a composite context can be composed from other contexts in the form of a process. To develop such a context with Model Driven Architecture, PIM-, PSM-, and code-level metamodels as well as transformation rules between the meta-models are defined. The composite context can then be ported to different platforms using only one PIM model. Also, a development of the composite context for two platforms - Windows Mobile and Android - using ATL transformation language is presented