A1 Refereed original research article in a scientific journal

Improving dictionary-based named entity recognition with deep learning




AuthorsNastou, Katerina; Koutrouli, Mikaela; Pyysalo, Sampo; Jensen, Lars Jyhl

PublisherOxford University Press

Publication year2024

JournalBioinformatics

Journal name in sourceBioinformatics

Volume40

Issue2 Supplement

First page ii45

Last pageii52

eISSN1367-4811

DOIhttps://doi.org/10.1093/bioinformatics/btae402

Web address https://doi.org/10.1093/bioinformatics/btae402

Self-archived copy’s web addresshttps://research.utu.fi/converis/portal/detail/Publication/457882809


Abstract

Motivation: Dictionary-based named entity recognition (NER) allows terms to be detected in a corpus and normalized to biomedical databases and ontologies. However, adaptation to different entity types requires new high-quality dictionaries and associated lists of blocked names for each type. The latter are so far created by identifying cases that cause many false positives through manual inspection of individual names, a process that scales poorly.

Results: In this work, we aim to improve block list s by automatically identifying names to block, based on the context in which they appear. By comparing results of three well-established biomedical NER methods, we generated a dataset of over 12.5 million text spans where the methods agree on the boundaries and type of entity tagged. These were used to generate positive and negative examples of contexts for four entity types (genes, diseases, species, and chemicals), which were used to train a Transformer-based model (BioBERT) to perform entity type classification. Application of the best model (F1-score = 96.7%) allowed us to generate a list of problematic names that should be blocked. Introducing this into our system doubled the size of the previous list of corpus-wide blocked names. In addition, we generated a document-specific list that allows ambiguous names to be blocked in specific documents. These changes boosted text mining precision by ∼5.5% on average, and over 8.5% for chemical and 7.5% for gene names, positively affecting several biological databases utilizing this NER system, like the STRING database, with only a minor drop in recall (0.6%).


Downloadable publication

This is an electronic reprint of the original article.
This reprint may differ from the original in pagination and typographic detail. Please cite the original version.




Funding information in the publication
This work was supported by Novo Nordisk Foundation [grant number NNF14CC0001] and from the Academy of Finland [grant number 332844]. K.N. has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Sklodowska-Curie [grant number 101023676]. M.K. has received funding from Novo Nordisk Foundation [grant number NNF20SA0035590]. This paper was published as part of a supplement financially supported by ECCB2024.


Last updated on 2025-27-01 at 19:36