Optimizing Requests for the Smart Data Server

The growing complexity of the Web information flow makes it necessary to improve Client/Server communication. In order to efficiently maintain heterogeneous and fast changing data types and different kinds of data sources, a Smart Data Server (SDS) can be used as a general framework for distributed functionality. In this paper, we present three general methods of advanced Client/Server communication: Prediction, Minimization and Request Optimization, implemented and in practical use as modules of the SDS. Even though the overall aim of these features is the improvement of the information flow, the processing is quite different: Prediction aims at making precalculations or pre-fetching of very probable user requests in the future, Minimization leads to a reduction of transferred data while considering system requirements of the client (e.g. display resolution), and Request Optimization is a usage-dependent data request method. We will show that only a skilful combination of these methods can guarantee success in improving network communication.