Automatic generation of semantic fields for resource discovery in the semantic web

In this paper we present and evaluate two approaches for the generation of Semantic Fields, which are used as a tool for resource discovery in the Semantic Web. We mainly concern ourselves with semantic networks that describe their interests and resources by means of ontologies. Semantic Fields are intended to help users to locate these resources by specifying a brief description (also as an ontology). We propose two ways to automatically build Semantic Fields. The first one is used in the KREIOS approach, which is based on the pre-computation of distances between all the ontology pairs. The second one is based on a fast incremental clustering algorithm, which groups together similar ontologies as they are published. These groups constitute a pre-computed set of Semantic Fields.