Bridging Relational Databases to Context-Aware Services

The provision of context-aware services in terms of the anytime/anywhere/anymedia paradigm requires the consideration of certain context properties in terms of time, location, device, or simply the user’s preferences. The specific characteristics of context, comprising among others large amounts of data, frequent and concurrent updates and heterogeneous sources often in terms of XML schemata, demand for an efficient management of context data. Relational databases seem to be a promising key technology for the efficient storage, retrieval, and processing of XML-based context data at the server-side. This is not least since relational databases allow to seamlessly combine context data with existing legacy data, hereby constituting a major advantage compared with native XML databases or object-oriented databases. The focus of this paper is twofold. First, requirements for managing context data are elaborated from a database perspective. Second, based on this, a generic approach called X-Ray, mediating between existing XML-based context and content stored in relational databases, is presented. With this approach, maintainability and changeability of context data is enhanced, since the mapping knowledge is not hard-coded but rather reified within a meta schema. The meta schema allows to automatically compose context data out of the relational database when requested and decompose them when they have to