complex-evaluate 0.0.1

Package to evaluate complex alignments.

Complex Evaluate

A Python library for evaluating complex ontology alignments in EDOAL (Expressive and Declarative Ontology Alignment Language) format adapting precision, recall, and f-measure metrics to the complex matching case.

Requirements

• Python >= 3.9
• NumPy
• SciPy

📦 Installation

pip install complex_evaluate

📖 Usage

Basic Example

from complex_evaluate.evaluate import evaluate_edoal

# Compare two alignment files
precision, recall, f_measure = evaluate_edoal(
    'predicted_alignment.edoal',
    'reference_alignment.edoal'
)

print(f"Precision: {precision:.3f}")
print(f"Recall: {recall:.3f}")
print(f"F-measure: {f_measure:.3f}")

Comparing from strings

from complex_evaluate.evaluate import evaluate_edoal_string

predicted = '''<?xml version="1.0" encoding="UTF-8"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" 
         xmlns="http://knowledgeweb.semanticweb.org/heterogeneity/alignment#">
  <Alignment>
    <map>
      <Cell>
        <entity1>
          <Class rdf:about="http://example.org#ClassA" />
        </entity1>
        <entity2>
          <Class rdf:about="http://example.org#ClassB" />
        </entity2>
      </Cell>
    </map>
  </Alignment>
</rdf:RDF>'''

reference = predicted  # Use same for identity test

p, r, f = evaluate_edoal_string(predicted, reference)
print(f"F-measure: {f}")  # Should be 1.0 for identical alignments

📊 Use Cases

This metric was used in the evaluation of OAEI 2025 in the Complex Matching track https://oaei.ontologymatching.org/2025/results/complex/index.html.

Also, this library is particularly useful for:

• Ontology Alignment Evaluation: Benchmarking alignment approaches on complex matching tasks.
• LLM reasoning training: The metric can enable the training of LLMs to reason about complex alignments, by providing a verifiable reward signal based on the score of the predicted alignment against a reference alignment.

🤝 Contributing

Contributions are welcome! Some areas for improvement:

• Additional similarity metrics.
• Performance optimizations.
• Support for other alignment formats.
• Extended documentation and examples.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

📚 Citation

If you use this library in your research, please cite it as follows:

@inproceedings{DBLP:conf/esws/SousaLS25,
  author       = {Guilherme Henrique Santos Sousa and
                  Rinaldo Lima and
                  C{\'{a}}ssia Trojahn dos Santos},
  title        = {On Evaluation Metrics for Complex Matching Based on Reference Alignments},
  booktitle    = {{ESWC} {(1)}},
  series       = {Lecture Notes in Computer Science},
  volume       = {15718},
  pages        = {77--93},
  publisher    = {Springer},
  year         = {2025}
}

---

Built with ❤️ for the Semantic Web and Ontology Matching community.