March 3, 2026Open Access

Adapting a relation extraction pipeline for the BioCreAtIvE II task

Key Points

The pipeline achieved competitive outcomes in relation extraction tasks relevant to the biomedical domain, enhancing data handling.
Key evidence indicates that the adapted system performed comparably against established state-of-the-art information extraction mechanisms.
Development involved an information extraction pipeline tailored to biomedical contexts, facilitating easier integration with BioCreAtIvE datasets.
Successful adaptation highlights the potential for existing systems to improve efficiency and effectiveness in processing biomedical information.

Abstract

The Second BioCreAtIvE Challenge provided an ideal opportunity to evaluate biomedical nlp techniques. Prior to the Challenge, an information extraction pipeline was developed to extract entities and relations relevant to the biomedical domain, and to normalise the entities to appropriate ontologies. With minimal eﬀort, the pipeline was adapted to work with the BioCreAtIvE data and achieved results that appear competitive with existing state-of-the-art systems.

Connected Papers

Building similarity graph...

Analyzing shared references across papers

Discussion

Authors

Claire; id_orcid 0000-0001-7019-1250 Grover

Barry; id_orcid 0009-0008-1295-8888 Haddow

Ewan Klein

Actions

References and Citations

Connected Papers

Building similarity graph...

Analyzing shared references across papers

Adapting a relation extraction pipeline for the BioCreAtIvE II task

Key Points

Abstract

Citation Network

Connected Papers

Discussion

Authors

Actions

References and Citations

Citation Network

Connected Papers

Discussion

Cite this study