Information Filtering and Retrieving of Context-Aware Applications within the MoBe Framework

Due to the appearance and widespread diffusion of new mobile devices (PDAs, smartphones etc.), the traditional notion of computing is quickly fading away, giving birth to new paradigms and to new non-trivial problems. Hence, the scientific community is searching for models, technologies, and architectures in order to suitably describe and guide the implementation of this new computing scenario. It is clear that the notion of context (whether physical or virtual or a mixture of both) plays a fundamental role, since it influences the computational capabilities of the devices that are in it. The present work directly addresses this problem proposing MoBe, a novel architecture for sending, in push mode, mobile applications (that we call MoBeLets) to the mobile devices (cellular phones, smartphones, PDAs, etc.) on the basis of the current context the user is in. The context is determined by both an ad-hoc MoBe infrastructure and data from sensors on the mobile device (or in its surroundings). To avoid a MoBeLet overload, our architecture exploits context-aware information retrieval and filtering techniques, which are briefly discussed.