DynDiff: A Tool for Comparing Versions of Large Ontologies

Authors

Benavides S.D., Cardoso S.D., Silveira M.D., Pruski C.

Reference

CEUR Workshop Proceedings, vol. 3573, 2022

Description

Ontologies have been widely adopted to represent domain knowledge. The dynamic nature of knowledge requires frequent changes in the ontologies to keep them up-to-date. Understanding and managing these changes and their impact on other artefacts become important for the semantic web community, due to the growing volume of data annotated with ontologies and the limited documentation describing their changes. In this paper, we present a method to automatically detect and classify the changes between different versions of ontologies. We also built an ontology of changes (DynDiffOnto) that we use to classify the changes and provide a context that makes them more comprehensible for machines and humans. We evaluate the algorithm with different ontologies from the biomedical domain (i.e. ICD9-CM, MeSH, NCIt, SNOMED-CT, GO, IOBC, and CIDO) and we compare the results with COnto-Diff. We observed that for small ontologies, COnto-Diff computes the Diff faster, but the opposite is observed for larger ontologies. The higher granularity of DynDiffOnto requires more rules to compute the Diff. It can partially justify the lower performance for small ontologies, but DynDiff provides a richer documentation for end-users.