Ontology-supported Information Retrieval

By moving towards the semantic Web ontologies as a way to describe a domain of interest will gain importance. Already in today's Web with its overwhelming amount of HTML Web pages ontologies can assist users in their whole information retrieval process. In this article, the author first explains how to support users in describing their information needs by using ontologies. Second, it is shown how to retrieve high quality data sources in a particular subject domain. For demonstration purposes, the examples in the article are based on a user searching for hotels. Third, the author also describes how search agents are able to query and analyze the content of the retrieved pages with the aid of ontologies. In contrast to the "wrapper" technique which is commonly used by most of the available applications to extract data of a known Web site, ontologies offer more flexibility, adaptability and robustness. Fourth, it is shown how the results of different sources are merged and presented as valuable data to the user. Users' benefits are more accurate query definitions in a given domain, delegation of time-consuming search tasks to search agents and comparable search results of higher quality