Revisiting and Versioning in Virtual Special Reports

Abstract

Adaptation/personalization is one of the main issues for web applications and require large repositories. Creating adaptive web applications from these repositories requires to have methods to facilitate web application creation and management and to ensure reuse, sharing and exchange of data through the internet/intranet. Virtual documents deal with these issues. In our framework, we are interested in adaptive virtual documents for author-oriented web applications providing several reading strategies to readers. These applications have the following characteristics: authors have know-how which enables them to choose document contents and to organize them in one or more consistent ways. A reading strategy and the corresponding content are semantically coherent and convey a particular meaning to the readers. Such author’s know-how can be represented at knowledge level and then be used for generating web documents dynamically, for ensuring reader comprehension and for sharing and reuse. Then an adaptive virtual document can be computed on the fly by means of a semantic composition engine using: i) an overall document structure - for instance a narrative structure - representing a reading strategy for which node contents are linked at run time, according to user’s needs for adaptation, ii) an intelligent search engine and semantic metadata relying on semantic web initiative, and iv) a user model. In this paper, we focus on a semantic composition engine enabling us to compute on the fly adaptive/personalized web documents in the ICCARS project. Its main goal is to assist the journalist in building adaptive special reports. In such a framework, adaptation, personalization and reusability are central issues for delivering adaptive special reports.