Ontowebber: model-driven ontology-based web site management
Yuhui Jin, Stefan Decker, Gio C. M. Wiederhold · 2001
Abstract. Building data-intensive Web sites and especially Web portals is a costly task, which requires considerable effort for data integration and maintenance and usually does not result in many reusable components. This is mainly because most of the design is hard-coded in static or dynamic Web pages. In this paper we integrate three different approaches to create a comprehensive solution to Web site and Web portal creation dubbed OntoWebber. OntoWebber integrates (1) the explicit modeling of different aspects of Web sites, (2) the use of ontologies as the foundation for Web portal design and (3) semi-structured data technology for data integration and Web site modeling. The resulting system and methodology supports the creation of reusable specifications of Web sites. OntoWebber is the basis for creating the Semantic Web Community Portal as part of the OntoAgents project, which will help the Semantic Web research community (distributed on the Web) to exchange and share knowledge conveniently and effectively. Building data-intensive Web sites is a high-effort task, which usually does not result in many reusable components, mainly because nowadays most of the design is hard-coded in HTML and executable code like CGI scripts, Active Server Pages (ASP) or Java Server Pages (JSP). Web portals is a special kind of data-intensive Web sites, which presents a large collection of information related to specific topics and are often organized by hierarchical directories. Examples of Web Portals are Yahoo!, and company portals, which present available resources inside and outside the company to their employees to facilitate cooperation. Knowledge management and dynamic personalization are key features of these Web portals, which make the management of these portals even more demanding than that of an ordinary Web site. Building and maintaining a portal requires considerable effort for data integration and maintenance, since quite often available information (e.g. inside a large corporation) is heterogeneous, distributed and constantly changing. Therefore it is highly desirable to automate the data integration and maintenance tasks as much as possible.