OWL 2 DL ontology merging framework tailored to the production domain

• preprocessing: translations, spell checks, and interactive debugging
• matching: string-similarity-based, terminological, and structural algorithms
• correspondence selection: greedy and optimal regarding the overall similarity score
• postprocessing: link ontology creation and interactive debugging
• quality assessment: calculation of precision, recall, and F-measure

• data/: production vocabulary, helper scripts for downloading ontos, and reference alignments
• docs/: configuration files for various examples, utilities for processing OAEI outputs and reference alignments, graphical abstract
• queries/: queries for information extraction
• src/: sources for creating, loading, preprocessing, matching, and merging the ontologies
• dependency-installer.sh: bash utility for installing dependencies
• cleanup.sh: bash utility for removing temporary and generated files

• Python 3.7
• bash recommended

• on Linux, run the bash script dependency_installer.sh to set up a virtual environment with the packages required
• minimal example: simply run prom.py
• production process example:
  1. download ontologies and preprocess them using the bash script download_ontos.sh in data/
  2. adapt the config file, as a reference cp. the file alt_config.yml in docs/
  3. run prom.py
• for running OAEI benchmarks, cp. the instructions in the utility scripts in docs/

For scientific use, please cite as follows:

@article{ocker2022merging,
  title = {A framework for merging ontologies in the context of smart factories},
  author = {Ocker, Felix and Vogel-Heuser, Birgit and Paredis, Christiaan JJ},
  journal={Computers in Industry},
  volume={135},
  pages={103571},
  year = {2022},
  publisher={Elsevier},
  doi={10.1016/j.compind.2021.103571}
}

GPL v3.0

Felix Ocker - felix.ocker@tum.de
Technical University of Munich - Institute of Automation and Information Systems