The GeneralRelation track aims to evaluate ontology matching systems on their ability to identify a wider range of semantic relations beyond equivalence. The benchmark includes subclass, superclass, overlap, and disjoint relations to reflect real-world ontology alignment scenarios.
This track addresses the growing demand for semantic integration where simple equivalence mappings are insufficient. It provides enriched datasets derived from product classification standards, STROMA/TaSeR test cases, and modified OAEI tracks, especially Anatomy.
Participants can download the complete benchmark dataset via Zenodo:
Three major sources are used:
Evaluation follows the isAmong framework alignment framework, which measures semantic agreement between predicted and reference alignments at the leaf-class level, rewarding partially correct correspondences.
For inquiries or to request evaluation support, please contact the track organizers:
Xianhao Liu (xliu@stibosystems.com),
Sven Hertling (sven.hertling@uni-mannheim.de),
Michael R. Hansen (mire@dtu.com)