In order to guarantee certain levels of QoS, a Web application needs to adapt itself to different execution contexts. However, because of the lack of coordination support in Web platforms, service providers respond to the challenge by simply providing multiple versions of a Web application, one for each context. We argue this top-down approach is neither efficient nor scalable: developing a context-specific application requires considerable effort and expertise while the ever-changing Internet never stops generating interesting contexts which can be exploited for better deployment. As an alternative, we propose a three-layer, bottom-up approach to building context-aware Web applications. At the bottom layer, we characterize a context-specific Web application with a particular component distribution plan which provides details for composing individual objects. In the middle layer, recursively defined configurations provide a bridge which relates high-level context features to low-level component distribution properties, where a configuration is a combination of configurations and/or component distribution properties. At the top level, a context management system selects desirable configurations according to the execution contexts. © IFIP International Federation for Information Processing 2007.
CITATION STYLE
Chang, P. H., & Agha, G. (2007). Towards context-aware web applications. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4531 LNCS, pp. 239–252). Springer Verlag. https://doi.org/10.1007/978-3-540-72883-2_18
Mendeley helps you to discover research relevant for your work.