Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Improving Translational Accuracy02:07

Improving Translational Accuracy

2.5K
2.5K

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Generative AI Chatbot for Diabetes Management: Formative 2-Part Qualitative Study Using DTalksBot Involving Patients and Clinicians.

JMIR formative research·2025
Same author

Identification of pathogenic <i>MEFV</i> variants in Korean patients with familial Mediterranean fever via whole-genome sequencing: a case report.

Journal of rheumatic diseases·2025
Same author

Type 1 interferon signature and allograft inflammatory factor-1 contribute to refractoriness to TNF inhibition in ankylosing spondylitis.

Nature communications·2025
Same author

Application of a Growth-Rate Model to Enhance Subgroup Identification in Heterogeneous Clinical Courses of the Idiopathic Inflammatory Myopathy-Associated Interstitial Lung Disease and Its Prognostic Implication.

International journal of rheumatic diseases·2025
Same author

Clinical Factors Associated With Pneumocystis Pneumonia Despite Its Primary Prophylaxis: When to Stop Prophylaxis?

Arthritis & rheumatology (Hoboken, N.J.)·2025
Same author

Serum Biomarkers of Pulmonary Damage and Risk for Progression of Rheumatoid Arthritis-Associated Interstitial Lung Disease.

The Journal of rheumatology·2025
Same journal

Multimodule Human-Artificial Intelligence Collaboration Pipeline for Large Language Model-Assisted Thematic Analysis Across Digital Health Interview Studies: Comparative Evaluation Study.

JMIR medical informatics·2026
Same journal

Graph Network Feature Space Fusion for Predicting Irregularly Sampled Medical Time-Series Data: Deep Learning Model Development and Validation Study.

JMIR medical informatics·2026
Same journal

Intrasystem Repeatability of S-Detect for Breast Ultrasound Classification With Identical Static Images: Single-Center Retrospective Repeatability Study.

JMIR medical informatics·2026
Same journal

Clinician Perspectives on Ambient AI Scribes in the Intensive Care Unit: Qualitative Interview Study.

JMIR medical informatics·2026
Same journal

IdeaDistiller-AI Support for Idea Synthesis in Concept Mapping: Algorithm Development and Validation Study.

JMIR medical informatics·2026
Same journal

Pregnancy-Related Clinical Codes in Unlikely Populations in Primary Care.

JMIR medical informatics·2026
See all related articles

Related Experiment Video

Updated: Jun 9, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

506

Multifaceted Natural Language Processing Task-Based Evaluation of Bidirectional Encoder Representations From

Kyungmo Kim1, Seongkeun Park2, Jeongwon Min1

  • 1Interdisciplinary Program for Bioengineering, Seoul National University, Seoul, Republic of Korea.

JMIR Medical Informatics
|October 30, 2024
PubMed
Summary
This summary is machine-generated.

This study compared Bidirectional Encoder Representations from Transformers (BERT) models for clinical notes in Korean and English. Multilingual BERT (M-BERT) showed superior performance in reading comprehension and knowledge inference tasks.

Keywords:
NLPlarge language modelsnatural language inferencenatural language processingreading comprehensiontransformer

More Related Videos

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
05:48

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception

Published on: August 9, 2024

1.4K
Examining Bilingual Language Control Using the Stroop Task
05:31

Examining Bilingual Language Control Using the Stroop Task

Published on: February 26, 2020

14.6K

Related Experiment Videos

Last Updated: Jun 9, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

506
Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
05:48

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception

Published on: August 9, 2024

1.4K
Examining Bilingual Language Control Using the Stroop Task
05:31

Examining Bilingual Language Control Using the Stroop Task

Published on: February 26, 2020

14.6K

Area of Science:

  • Natural Language Processing
  • Clinical Informatics
  • Artificial Intelligence

Background:

  • Bidirectional Encoder Representations from Transformers (BERT) models are increasingly used in clinical applications like patient classification.
  • Current research often overlooks thorough clinical context comprehension assessments and comparative studies on BERT models in non-English medical documents.
  • The applicability of English-trained BERT models to non-English clinical contexts remains unconfirmed.

Purpose of the Study:

  • To evaluate the contextual understanding abilities of various BERT models applied to mixed Korean and English clinical notes.
  • To identify the most effective BERT model for non-English clinical notes.

Main Methods:

  • Pretrained BERT-base, BioBERT (BERT for Biomedical Text Mining), KoBERT (Korean BERT), and Multilingual BERT (M-BERT) using data from 164,460 patients.
  • Compared model performances across 7 fine-tuning tasks to assess contextual comprehension.

Main Results:

  • BERT-base and BioBERT excelled in classification tasks (highest F1-score 89.32), demonstrating effectiveness even with limited Korean tokens.
  • Multilingual BERT (M-BERT) outperformed in reading comprehension (F1-score 93.77) and knowledge inference (hit@10 score 95.41).
  • Model performance varied by task and token usage, with fewer unknown tokens generally yielding better results.

Conclusions:

  • Various BERT models show effectiveness in multilingual clinical domains.
  • Findings provide a reference for selecting appropriate BERT models in clinical and language-based applications.
  • Further research can explore optimal BERT model selection for specific non-English clinical tasks.