OntoLearner 1.5.0

OntoLearner: A Modular Python Library for Ontology Learning with LLMs.

OntoLearner: A Modular Python Library for Ontology Learning with LLMs

OntoLearner is a modular and extensible architecture designed to support ontology learning and reuse. The conceptual and functional architecture of OntoLearner is shown as following. The framework comprises three core components—Ontologizers, Learning Tasks, and Learner Models—structured to enable reusable and customizable ontology engineering workflows.

🧪 Installation

OntoLearner is available on PyPI and you can install using pip:

pip install ontolearner

Next, verify the installation:

import ontolearner

print(ontolearner.__version__)

Please refer to Installation page for further options.

🔗 Essential Resources

Resource	Info
📚 OntoLearner Documentation	OntoLearner's extensive documentation website.
🤗 Datasets on Hugging Face	Access curated, machine-readable ontologies.
🚀 Quickstart	Get started quickly with OntoLearner’s main features and workflow.
🕸️ Learning Tasks	Explore supported ontology learning tasks like LLMs4OL Paradigm tasks and Text2Onto.
🧠 Learner Models	Browse and configure various learner models, including LLMs, Retrieval, or RAG approaches.
📚 Ontologies Documentations	Review benchmark ontologies and datasets used for evaluation and training.
🧩 How to work with Ontologizer?	Learn how to modularize and preprocess ontologies using the Ontologizer module.
🤗 Ontology Metrics Dashboard	Benchmark ontologies with their metrics and complexity scores.

🚀 Quick Tour

Get started with OntoLearner in just a few lines of code. This guide demonstrates how to initialize ontologies, load datasets, and train an LLM-assisted learner for ontology engineering tasks.

Basic Usage - Automatic Download from Hugging Face:

from ontolearner import Wine

# 1. Initialize an ontologizer from OntoLearner
ontology = Wine()

# 2. Load the ontology automatically from HuggingFace
ontology.load()

# 3. Extract the learning task dataset
data = ontology.extract()

To see the ontology metadata you can print the ontology:

print(ontology)

Now, explore 150+ ready-to-use ontologies or read on how to work with ontologizers.

Learner Models:

from ontolearner import AutoRetrieverLearner, AgrO, train_test_split, evaluation_report

# 1. Programmatic import of an ontology
ontology = AgrO()
ontology.load()

# 2. Load tasks datasets
ontological_data = ontology.extract()

# 3. Split into train and test sets
train_data, test_data = train_test_split(ontological_data, test_size=0.2, random_state=42)

# 4. Initialize Learner
task = 'non-taxonomic-re'
ret_learner = AutoRetrieverLearner(top_k=5)
ret_learner.load(model_id='sentence-transformers/all-MiniLM-L6-v2')

# 5. Fit the model to training data and then predict over the test data
ret_learner.fit(train_data, task=task)
predicts = ret_learner.predict(test_data, task=task)

# 6. Evaluation
truth = ret_learner.tasks_ground_truth_former(data=test_data, task=task)
metrics = evaluation_report(y_true=truth, y_pred=predicts, task=task)
print(metrics)

Other learners:

• LLM-Based Learner
• RAG-Based Learner

LearnerPipeline: The OntoLearner also offers a streamlined LearnerPipeline class that simplifies the entire process of initializing, training, predicting, and evaluating a RAG setup into a single call.

# Import core components from the OntoLearner library
from ontolearner import LearnerPipeline, AgrO, train_test_split

# Load the AgrO ontology, which includes structured agricultural knowledge
ontology = AgrO()
ontology.load()  # Load ontology data (e.g., entities, relations, metadata)

# Extract relation instances from the ontology and split them into training and test sets
train_data, test_data = train_test_split(
    ontology.extract(),      # Extract annotated (head, tail, relation) triples
    test_size=0.2,           # 20% for evaluation
    random_state=42          # Ensures reproducible splits
)

# Initialize the learning pipeline using a dense retriever
pipeline = LearnerPipeline(
    retriever_id='sentence-transformers/all-MiniLM-L6-v2',  # Hugging Face model ID for retrieval
    batch_size=10,       # Number of samples to process per batch (if batching is enabled internally)
    top_k=5              # Retrieve top-5 most relevant support instance per query
)

# Run the pipeline on the training and test data
# The pipeline performs: fit() → predict() → evaluate() in sequence
outputs = pipeline(
    train_data=train_data,
    test_data=test_data,
    evaluate=True,           # If True, computes precision, recall, and F1-score
    task='non-taxonomic-re'  # Specifies that we are doing non-taxonomic relation prediction
)

# Print the evaluation metrics (precision, recall, F1)
print("Metrics:", outputs['metrics'])

# Print the total elapsed time for training and evaluation
print("Elapsed time:", outputs['elapsed_time'])

# Print the full output dictionary (includes predictions)
print(outputs)

⭐ Contribution

We welcome contributions to enhance OntoLearner 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 find this repository helpful or use OntoLearner in your work or research, feel free to cite our publication:

@inproceedings{babaei2023llms4ol,
  title={LLMs4OL: Large language models for ontology learning},
  author={Babaei Giglou, Hamed and D’Souza, Jennifer and Auer, S{\"o}ren},
  booktitle={International Semantic Web Conference},
  pages={408--427},
  year={2023},
  organization={Springer}
}

or:

@software{babaei_giglou_2025_15399783,
  author       = {Babaei Giglou, Hamed and D'Souza, Jennifer and Aioanei, Andrei and Mihindukulasooriya, Nandana and Auer, Sören},
  title        = {OntoLearner: A Modular Python Library for Ontology Learning with LLMs},
  month        = may,
  year         = 2025,
  publisher    = {Zenodo},
  version      = {v1.3.0},
  doi          = {10.5281/zenodo.15399783},
  url          = {https://doi.org/10.5281/zenodo.15399783},
}

---

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