LLMs4OL 2026: Large Language Models for Ontology Learning
The 3rd LLMs4OL Challenge @ ISWC 2026
ISWC 2026, Bari, Italy | 25-29 October
Embedded Files
Reuse Task
Reuse Task
Ontology Extension
Definition: Given a partial ontology, add new concepts and relations extracted from text.
Motivation: In practice, ontologies are rarely created from scratch, domains evolve over time, and knowledge grows incrementally. This task tests whether systems can adapt and expand an ontology, not just generate one from raw text.
Objective
Objective
Given an existing ontology, the participants must extend it using the given new terms and types with accompanying raw text. Unlike the End-to-End OL task, this task starts from a partially built ontology and focuses on incremental knowledge addition rather than constructing a complete ontology from scratch.
What must participants build?
What must participants build?
A system that, starting from unstructured text, extends a primitive ontology includes:
Term Typings (mappend instances to classes)
Taxonomic Discovery (is-a/subclass)
Non-taxonomic relations
All stages must be automatically derived from text and given a primitive ontology via an integrated pipeline.
A toy example
A toy example
Provided raw text in input:
A humidity sensor monitors humidity in the house. The smart humidifier receives readings from the humidity sensor and adjusts the water flow. A mobile app allows the user to control the humidifier remotely.
Provided terms:
user, mobile app, humidity level
Provided types:
application, person, measurement, humidity sensor
Existing ontology:
(thermostat, instance-of, device)
(temperature sensor, instance-of, sensor)
(sensor, is-a, device)
(system, is-a, device)
➡️ 1. Term Typing
humidity level ---> measurementuser ---> personmobile app ---> application
➡️ 2. Taxonomic Discovery
(application, is-a, device)
➡️ 3. Non-Taxonomic RE
None
The final output would be as follows:
(humidity sensor, instance-of, measurement)
(user, instance-of, person)
(mobile app, instance-of, application)
(application, is-a, device)
Dataset
Dataset
More information on train/test datasets (formatting) will be available soon! Stay tuned!
Evaluation Metrics
Evaluation Metrics
Standard Metrics: Precision, Recall, F1
Standard Task-specific Metrics: Precision, Recall, F1 scores, for Term Typing, Taxonomy Discovery, and Non-Taxonomic RE
Page updated
Google Sites
Report abuse