Jump to: navigation, search
at AI-Lab
Status: prototype
Last release: (Oct 1 2007)
License: open source
Affiliation: University of the Aegean, ICSE department, AI-Lab

This page describes an ontology mapping approach under development, called AUTOMS (Automated Ontology Mapping through Synthesis) which integrates state-of-the-art and innovative techniques for ontology mapping, requiring no human involvement. The presented approach uses the semantic matching algorithm of the HCONE-merge approach to ontology merging, and extends it in several ways:

1. It integrates a state-of-the-art lexical matching algorithm, named COCLU. The algorithm was originally developed for use in ontology population systems. However we have tested it successfully in mapping ontology concepts and properties.

2. It uses the WordNet 2.0 lexicon. Previous versions of the semantic matching algorithm delivered lower recall percentages due to the fact that many of the ontology concepts did not have a lexical entry in WordNet 1.7.

3. It integrates the structural matching method SMR (structure matching rules) to "catch" missed concept pairs. Some concept pairs do not have a lexical entry in WordNet 2.0 or even if they do have, they do not belong to the same WordNet synset (i.e. they are not synonyms). To discover these mappings and increase recall, additional heuristic rules that check the structure of concepts/properties have been used.

4. It computes the mappings of properties in addition to that of concepts.

5. It improves interoperability with Semantic Web tools. The source ontologies are OWL-DL ontologies, and the mappings are produced in an automatic manner.

6. It reports increased precision and recall: Due to the combination of innovative and state of the art matching algorithms, the performance of the proposed approach is increased.

7. Performs well as far as its execution time is concerned.

AUTOMS has been developed within an internal project of University of the Aegean, Dept. of Information and Communication Systems Eng., AI-Lab.