Building Customizable Context-aware Systems

Abstract

Context-aware systems provide proactive services for users based on environment contexts. However, most of the existing frameworks were introduced for special applications or technology without flexible support for different domain applications. This paper proposes context-aware systems, and addresses the important issues of system customization, evolutionary context rules, interaction mechanism and context-usage patterns. Customization allows effectively configuring and managing its functional modules to fit domain applications with minimum configuration. Context rules are constantly extracted and evolved by accommodating new data to evolutionary mining. Interaction mechanism focuses on the mutual effects between systems and context world, making system response more immediately and accurately. Context-usage patterns explicitly define the typical context-usage ways in the systems. Therefore, the system proposed is adaptive to apply different context domains with scalability and reusability to provide more intelligent, personalized and proactive services to users.