Transfer learning for biomedical named entity recognition with BioBert

Anthi Symeonidou, Viachaslau Sazonau, Paul Groth

Research output: Contribution to journalConference articlepeer-review

4 Scopus citations

Abstract

We apply a transfer learning approach to biomedical named entity recognition and compare it with traditional approaches (dictionary, CRF, BiLTSM). Specifically, we build models for adverse drug reaction recognition on three datasets. We tune a pre-trained transformer model, BioBERT, on these datasets and observe the absolute F1-score improvements of 6.93, 10.46 and 13.31. This shows that, with a relatively small amount of annotated data, transfer learning can help in specialized information extraction tasks.

Original languageEnglish
JournalCEUR Workshop Proceedings
Volume2451
StatePublished - 2019
Externally publishedYes
Event15th International Conference on Semantic Systems, SEMPDS 2019 - Karlsruhe, Germany
Duration: Sep 9 2019Sep 12 2019

Keywords

  • Adverse drug reaction
  • BIO tagging
  • BioBERT
  • Drug safety
  • Named entity recognition
  • Text mining
  • Transfer learning

Fingerprint

Dive into the research topics of 'Transfer learning for biomedical named entity recognition with BioBert'. Together they form a unique fingerprint.

Cite this