LLMs4OL 2025: Large Language Models for Ontology Learning
The 2nd LLMs4OL Challenge @ ISWC 2025
ISWC 2025, Nara, Japan | 2-6 November
Embedded Files
Task D - Non-Taxonomic Relation Extraction
Task D - Non-Taxonomic Relation Extraction
Identify non-taxonomic, semantic relations between types.
Identify non-taxonomic, semantic relations between types.
This task aims to extract non-hierarchical (non-taxonomic) semantic relations between concepts in an ontology. Unlike taxonomy discovery, which deals with is-a relationships, this task focuses on other meaningful associations such as part-whole (part-of), causal (causes), functional (used-for), and associative (related-to) relationships. For example, in a medical ontology, discovering that “Aspirin” treats “Headache” adds valuable relational knowledge that enhances the utility of an ontology.
SubTasks
SubTasks
Given a set of ontological types and relationships, identify (head-type, relation, tail-type) triplets that form a non-taxonomic (none other than is-a) relationship.
- SubTask D1 – SWEET
Extracts non-taxonomic relationships between environmental and geoscience concepts using the SWEET Ontology.
- SubTask D2 – FoodOn
It focuses on identifying relationships in the FoodOn Ontology, such as functional and compositional links like milk used for cheese production or ingredients part of the dish.
- SubTask D3 – GO
This subtask focuses on identifying non-taxonomic relationships between specified types in genomic data.
Datasets
Datasets
The datasets per given SubTasks D1, D2, and D3 are available at the "TaskD-NonTaxonomicRE" directory of the challenge repository here https://github.com/sciknoworg/LLMs4OL-Challenge/tree/main/2025.
Per dataset, three key files were given as follows:
The file with "_re_relations.txt" suffix is a plain text file listing relationships (one per line). This is the list of possible relationships among types.
The file with "_re_types.txt" suffix is a plain text file listing types (one per line). This is the list of types for forming non-taxonomic relations.
A JSON file with "_re_pairs.json" suffix that represents non-taxonomic relationships between head and tail types with their relationships. This file contains ground truth with the following format.
[
{
"ID": "NR_a182a800",
"head": "obsolete histone H4-S1 phosphorylation",
"tail": "histone H4S1 kinase activity",
"relation": "consider"
},
....
]
🧪 During the testing phase, only types and relations will be provided, and participants are expected to return "_re_pairs.json" representing non-taxonomic relationships.
Page updated
Google Sites
Report abuse