OntoAligner: A Comprehensive Modular and Robust Python Toolkit for Ontology Alignment

OntoAligner is a Python library designed to simplify ontology alignment and matching for researchers, practitioners, and developers. With a modular architecture and robust features, OntoAligner provides powerful tools to bridge ontologies effectively.

🧪 Installation

You can install OntoAligner from PyPI using pip:

bash pip install ontoaligner

Alternatively, to get the latest version directly from the source, use the following commands:

bash git clone git@github.com:sciknoworg/OntoAligner.git pip install ./ontoaligner

📚 Documentation

Comprehensive documentation for OntoAligner, including detailed guides and examples, is available at ontoaligner.readthedocs.io. Below are some key tutorials with links to both the documentation and the corresponding example codes.

| Example | Tutorial | Script | |:-------------------------------|:--------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------:| | Lightweight | 📚 Fuzzy Matching | 📝 Code | | Retrieval | 📚 Retrieval Aligner | 📝 Code | | Large Language Models | 📚 LLM Aligner | 📝 Code | | Retrieval Augmented Generation | 📚 RAG Aligner | 📝 Code| | FewShot | 📚 FewShot-RAG Aligner | 📝 Code | In-Context Vectors Learning | 📚 In-Context Vectors RAG | 📝 Code | Knowledge Graph Embedding | 📚 KGE Aligner | 📝 Code | eCommerce | 📚 Product Alignment in eCommerce | 📝 Code

🚀 Quick Tour

Below is an example of using Retrieval-Augmented Generation (RAG) step-by-step approach for ontology matching:

```python from ontoaligner.ontology import MaterialInformationMatOntoOMDataset from ontoaligner.utils import metrics, xmlify from ontoaligner.aligner import MistralLLMBERTRetrieverRAG from ontoaligner.encoder import ConceptParentRAGEncoder from ontoaligner.postprocess import raghybridpostprocessor

Step 1: Initialize the dataset object for MaterialInformation MatOnto dataset

task = MaterialInformationMatOntoOMDataset() print("Test Task:", task)

Step 2: Load source and target ontologies along with reference matchings

dataset = task.collect( sourceontologypath="assets/MI-MatOnto/miontology.xml", targetontologypath="assets/MI-MatOnto/matontoontology.xml", referencematchingpath="assets/MI-MatOnto/matchings.xml" )

Step 3: Encode the source and target ontologies

encodermodel = ConceptParentRAGEncoder() encodedontology = encoder_model(source=dataset['source'], target=dataset['target'])

Step 4: Define configuration for retriever and LLM

retrieverconfig = {"device": 'cuda', "topk": 5,} llmconfig = {"device": "cuda", "maxlength": 300, "maxnewtokens": 10, "batch_size": 15}

Step 5: Initialize Generate predictions using RAG-based ontology matcher

model = MistralLLMBERTRetrieverRAG(retrieverconfig=retrieverconfig, llmconfig=llmconfig) model.load(llmpath = "mistralai/Mistral-7B-v0.3", irpath="all-MiniLM-L6-v2") predicts = model.generate(inputdata=encodedontology)

Step 6: Apply hybrid postprocessing

hybridmatchings, hybridconfigs = raghybridpostprocessor(predicts=predicts, irscorethreshold=0.1, llmconfidenceth=0.8)

evaluation = metrics.evaluationreport(predicts=hybridmatchings, references=dataset['reference']) print("Hybrid Matching Evaluation Report:", evaluation)

Step 7: Convert matchings to XML format and save the XML representation

xmlstr = xmlify.xmlalignmentgenerator(matchings=hybridmatchings) open("matchings.xml", "w", encoding="utf-8").write(xml_str) ```

Ontology alignment pipeline using RAG method:

```python import ontoaligner

pipeline = ontoaligner.OntoAlignerPipeline( taskclass=ontoaligner.ontology.MouseHumanOMDataset, sourceontologypath="assets/MI-MatOnto/miontology.xml", targetontologypath="assets/MI-MatOnto/matontoontology.xml", referencematching_path="assets/MI-MatOnto/matchings.xml", )

matchings, evaluation = pipeline( method="rag", encodermodel=ontoaligner.encoder.ConceptRAGEncoder(), modelclass=ontoaligner.aligner.MistralLLMBERTRetrieverRAG, postprocessor=ontoaligner.postprocess.raghybridpostprocessor, llmpath='mistralai/Mistral-7B-v0.3', retrieverpath='all-MiniLM-L6-v2', llmthreshold=0.5, irragthreshold=0.7, topk=5, maxlength=512, maxnewtokens=10, device='cuda', batchsize=32, return_matching=True, evaluate=True )

print("Matching Evaluation Report:", evaluation) ```

⭐ Contribution

We welcome contributions to enhance OntoAligner and make it even better! Please review our contribution guidelines in CONTRIBUTING.md before getting started. You are also welcome to assist with the ongoing maintenance by referring to MAINTENANCE.md. Your support is greatly appreciated.

If you encounter any issues or have questions, please submit them in the GitHub issues tracker.

💡 Acknowledgements

If you use OntoAligner in your work or research, please cite the following preprint:

bibtex @inproceedings{babaei2025ontoaligner, title={OntoAligner: A Comprehensive Modular and Robust Python Toolkit for Ontology Alignment}, author={Babaei Giglou, Hamed and D’Souza, Jennifer and Karras, Oliver and Auer, S{\"o}ren}, booktitle={European Semantic Web Conference}, pages={174--191}, year={2025}, organization={Springer} }

This software is archived in Zenodo under the DOI and is licensed under .