Validation of an approach for finding good anchor nodes in ontologies in the semantic web

When matching different semantically-tagged web service interfaces in the semantic web, there is often a need to match one ontology to another ontology on the fly. In many cases knowing up front which are the best nodes to start the mapping process (these are called anchor nodes) can make the ontology mapping much easier. Traditionally, all anchor node selections have been performed by humans. However, this is not practical for on-the-fly situations. In previous research, we suggested a method for determining anchor nodes automatically, and performed a pilot study. In this paper we further that previous research with an expanded study. This study finds that anchor nodes can be automatically detected using certain metrics.