Using ontologies and relatedness metrics for semantic document analysis on the Web

Abstract

The problem of defining new methodologies for Knowledge Representation has a great influence on information technology and, from a general point of view, on cognitive sciences. In the last years some new approach strictly related to the above matter has been proposed and some of them are based on ontologies to reduce conceptual or terminological mess and to have a common view of the same information. We here propose an ontological model to represent information and we implement it in a content based information system for scoring documents w.r.t. a given topic an ad hoc metric; we use the Web as our context of interest. © 2008 Springer-Verlag Berlin Heidelberg.