BioRED track lasigeBioTM submission: Relation extraction using Domain Ontologies with BioRED

The present repository contains the data used for our participation at the BioCreative VIII Track 1 Biored in both sub-tasks.

The paper describing our participation (team lasigeBioTM) will be available here.

Data availability

Tools

• Named-Entity Recognition: HunFlair
• Relation Extraction: K-RET

Knowledge Bases

NER

• ChemicalEntity
• GeneOrGeneProduct
• OrganismTaxon
• DiseaseOrPhenotypicFeature
• CellLine

RE

• Human Phenotype Ontology (HPO)
• Human Disease Ontology (DO)
• Chemical Entities of Biological Interest (ChEBI)
• NCBITaxon Ontology
• Gene Ontology (GO)

Sub-task 1 Results

Official Runs

RUN 1	P	R	F
Entity pair:	0.1942	0.9914	0.3248
Entity pair+Relation type:	0.0435	0.2218	0.0727
Entity pair+Novelty:	0.0770	0.3963	0.1290
Entity pair+Relation type+Novelty:	0.0177	0.0908	0.0296

Unofficial Runs

		RUN 1				RUN 2				RUN 3
	P	R	F		P	R	F		P	R	F
Entity pair:	0.1942	0.9914	0.3248		0.1942	0.9914	0.3248		0.1942	0.9914	0.3248
Entity pair+Relation type:	0.0825	0.4214	0.1381		0.0801	0.4091	0.1340		0.0435	0.2218	0.0727
Entity pair+Novelty:	0.0770	0.3963	0.1290		0.0770	0.3963	0.1290		0.0770	0.3963	0.1290
Entity pair+Relation type+Novelty:	0.0329	0.1695	0.0552		0.0309	0.1592	0.0518		0.0177	0.0908	0.0296

Sub-task 2 Results

Official Runs

RUN 1	P	R	F
NER:	0.0003	0.0001	0.0001
Normalization:	0.4477	0.0710	0.1226
Entity pair:	0.0563	0.0219	0.0315
Entity pair+Relation type:	0.0128	0.0050	0.0072
Entity pair+Novelty:	0.0187	0.0073	0.0105
Entity pair+Relation type+Novelty:	0.0047	0.0018	0.0026

Unofficial Runs

		RUN 1				RUN 2
	P	R	F		P	R	F
NER:	0.7703	0.6411	0.6998		0.7703	0.6411	0.6998
Normalization:	0.4212	0.2280	0.2958		0.4212	0.2280	0.2958
Entity pair:	0.0427	0.0633	0.0510		0.0427	0.0633	0.0510
Entity pair+Relation type:	0.0110	0.0163	0.0131		0.0084	0.0125	0.0100
Entity pair+Novelty:	0.0207	0.0307	0.0247		0.0207	0.0307	0.0247
Entity pair+Relation type+Novelty:	0.0058	0.0086	0.0069		0.0040	0.0060	0.0048