Towards Real-Time Context Awareness for Mobile Users: A Declarative Meta-Programming Approach

Abstract

We envision a future with thousands of publicly available context sources acquired via sensors supplying real-time information about mobile users’ current circumstances with information on the Web. Such information can be harnessed for real-time decision making in daily life. Since context sources and information can be combined in myriad ways and be reasoned about in different ways, there is a need for some means to represent such aggregations, to create new aggregations, or to reason with such aggregations. Our basic idea is that the way context is aggregated to infer situations can be encapsulated and modularised in what we call “situation programs.” As we use a declarative programming approach, situation programs are readable, yet executable – a situation program encapsulates rules and queries to context sources, which can be executed to determine if a particular situation is occurring. Situation programs are treated as first-class entities and can be exchanged or loaded to increase the repertoire of situations an application can detect, or to provide alternative ways for an application to detect a situation or reason about situations. We describe our initial prototype Logic Programming for Context-Aware Programming System (LogicCAP-S) based on the language LogicCAP, and discuss extensions towards the mobile environment.