A Navigational and Structural Approach for Extracting Contents from Web Portals

In a semantic Web portal, contents are described and organized based on domain ontologies, and are usually extracted from traditional portals. However, with the increasing amount of information generated each day on the Web, updating semantic portals still represents a major challenge, since this task lacks mechanisms to extract and integrate information dynamically. This paper proposes a strategy to help promoting the interoperability between portals. It consists on the extraction of contents from different Web sites on a specific domain, aiming at the instantiation of a domain ontology, and then use it to update and/or populate a semantic portal. This is carried out through the analysis of the navigational and structural characteristics of traditional portals endowed with some semantic potentiality. In order to evaluate this strategy, a tool named NECOW was implemented. NECOW performance was compared to the Google advanced search mode, and showed promising results.