OntoLearner: A Modular Python Library for Ontology Learning with Large Language Models

2026-07-02 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Expert, quick

Summary

OntoLearner is a new modular Python library designed to unify ontology learning (OL) research by providing a shared infrastructure for systematic evaluation and ontology access. It addresses the fragmented progress in OL, which has historically lacked standardized tools across methods, domains, and evaluation practices. The framework offers 180 machine-readable ontologies across 22 domains and pipeline-ready datasets with train/dev/test splits for three core OL tasks: term typing, taxonomy discovery, and non-taxonomic relation extraction. A large-scale empirical study using OntoLearner evaluated 22 retrieval models and 12 large language models (LLMs), revealing that OL's central challenge lies in ontological complexity rather than model size or architectural sophistication. This indicates a structural mismatch between model knowledge encoding and ontology organization, suggesting effective OL is achievable through cross-domain, multi-task benchmarking. OntoLearner is open-source under an MIT license.

Key takeaway

For AI Scientists and NLP Engineers developing ontology learning solutions, you should prioritize addressing ontological complexity rather than solely focusing on larger or more sophisticated LLMs. The OntoLearner framework provides a critical resource for systematic evaluation, offering 180 ontologies and datasets for term typing, taxonomy discovery, and non-taxonomic relation extraction. Utilize this open-source library to benchmark models against diverse ontological structures, guiding your research towards more effective and robust OL systems.

Key insights

Ontology learning's primary bottleneck is ontological complexity, not LLM capability or size.

Principles

• OL progress needs shared infrastructure.
• Complexity, not model size, limits OL.
• Standardized benchmarking is key.

Method

OntoLearner unifies ontology access, LLM-driven learning pipelines, and standardized benchmarking for term typing, taxonomy discovery, and non-taxonomic relation extraction.

In practice

• Access 180 machine-readable ontologies.
• Utilize pipeline-ready OL datasets.
• Evaluate LLMs across OL tasks.

Topics

• Ontology Learning
• Large Language Models
• Knowledge Graphs
• Benchmarking
• Python Library
• Taxonomy Discovery

Code references

• sciknoworg/OntoLearner

Best for: AI Scientist, NLP Engineer, Research Scientist

Related on AIssential

• BLINKG: A Benchmark for LLM-Integrated Knowledge Graph Generation — BLINKG provides a standardized benchmark to evaluate Large Language Models' effectiveness in mapping data schemes to ontology concepts for Knowledge Graph generation.
• Ontology, Part I — Conflicting definitions and terminology create significant barriers to understanding and implementing ontologies in AI.
• A Pythonic Functional Approach for Semantic Data Harmonisation in the ILIAD Project — A Pythonic functional approach simplifies semantic data harmonisation by abstracting complex Semantic Web standards into familiar Python functions.
• Language is the Bridge — Shared language and explicit linguistic scaffolding are essential for making LLMs and ontologies actionable and trustworthy.
• Knowledge Graph Re-engineering Along the Ontological Continuum (extended version) — The "ontological continuum" offers a principled framework to describe and transform diverse knowledge graph modeling practices.
• The Ontology Pipeline™, Refresh — Effective semantic knowledge infrastructure requires a structured pipeline, human expertise, and robust governance, not just AI-generated outputs.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.