Modern Web Information Systems (WIS) need to satisfy a large number of requirements coming from different WIS stakeholders. Modeling WIS by focusing at one design aspect at-a-time helps the implementation of these requirements. During the last years several model-driven methodologies have been proposed to support the WIS design. Strategic modeling is usually the first step in WIS design. It is a very general characterization of WIS which answers questions like: what is the purpose of the WIS?, which are the WIS users?, what functionality is provided by the WIS?, what is the content of the WIS?, what is the layout and atmosphere of the presentations provided by the WIS?, etc. It is only after answering the above questions at a high abstract level that the designer can proceed with the detailed specifications of the WIS. Data integration is one of the most important characteristics of WIS. Some examples of domains in which data integration is present are: public services and bioinformatics. WIS need to support user interfaces that make a lot of data coming from different sources available to the user in a transparent way. The Semantic Web technologies seem to facilitate the data integration problem on the Web by providing the necessary languages to describe the data semantics. Very often the Web user browses pages that he will like to view again at a later time. The present browsing history mechanisms included in Web browsers proved to be insufficient for an adequate retrieval of already seen information. A semantical organization of the previously visited pages can improve the process of retrieving previously seen data. There is an increasing demand to make WIS personalizable so that these systems better deal with the user interests. WIS design methodologies do propose adaptation techniques in order to realize WIS personalization. Despite the fact that some of these adaptation techniques are very similar (or even the same) in different methodologies, the notations to specify WIS personalization aspects are quite different. By defining a reference model for specifying WIS personalization one could improve the reuse of the personalization specifications and also enable a seamless translation between different specific personalization specifications. The above issues are some of the topics that are tackled in the workshop papers. We hope that we did raise the readers’ interest so that they will have a close look at the papers and possibly contribute to the fascinating and challenging area of WIS modeling.
|Plaats van productie||Wollongong, NSW, Australia|
|Uitgeverij||University of Wollongong, School of IT and Computer Science|
|Status||Gepubliceerd - 2005|