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 Experiment Videos

A method for named entity normalization in biomedical articles: application to diseases and plants.

Hyejin Cho1, Wonjun Choi1, Hyunju Lee2

  • 1School of Electrical Engineering and Computer Science, Gwangju Institute of Science and Technology, 123 Chemdangwagi-ro, Buk-gu, Gwangju, Republic of Korea.

BMC Bioinformatics
|October 15, 2017
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

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

Sort by
Same author

A pilot randomized controlled trial of mentalization-based treatment and structured clinical management for outpatients with borderline personality disorder in South Korea.

Borderline personality disorder and emotion dysregulation·2026
Same author

M1UK Lineage <i>Streptococcus pyogenes</i> Identified in Korean Adolescents with Toxic Shock Syndrome.

Infection & chemotherapy·2026
Same author

Multimodal Validation of an sEMG-Based Visual Biofeedback System for Deep Abdominal Muscle Activation in Healthy Adults: A Randomized Controlled Proof-of-Concept Trial.

Healthcare (Basel, Switzerland)·2026
Same author

Adverse childhood experiences and well-being in middle-aged men in South Korea: mediating role of andropause symptoms.

Health and quality of life outcomes·2026
Same author

Clinical and genetic association of comorbid provisional PMDD and ADHD in female patients with bipolar disorder.

Journal of affective disorders·2026
Same author

Epidemiology and clinical characteristics of invasive group A streptococcal infection in the Republic of Korea, 2015-2024: a nationwide multicenter study.

The Lancet regional health. Western Pacific·2026

This study introduces a novel approach for normalizing biological entities using word embeddings, improving accuracy beyond traditional dictionaries. The method enhances the extraction of biological knowledge from research articles.

Area of Science:

  • Biomedical Informatics
  • Natural Language Processing

Background:

  • Named Entity Recognition (NER) is crucial for extracting biological knowledge from texts.
  • Entity normalization maps identified names to standard concepts, but dictionaries are often incomplete.
  • Neural network algorithms show promise for handling large volumes of biomedical data.

Purpose of the Study:

  • To develop an improved approach for normalizing biological entities, including diseases and plants.
  • To leverage word embeddings for representing semantic spaces in biological entity normalization.
  • To enhance the accuracy of entity normalization beyond the limitations of existing dictionaries.

Main Methods:

  • Utilized word embeddings to create semantic representations of biological entities.
  • Trained models using data from the National Center for Biotechnology Information (NCBI) disease corpus and manually constructed plant corpus, combined with unlabeled PubMed abstracts.
Keywords:
Disease namesEntity name normalizationNamed entity recognitionNeural networksPlant namesText mining

Related Experiment Videos

  • Incorporated both labeled training data and large unlabeled datasets for robust word representations.

    • Main Results:

    • Achieved F-scores of 0.808 for disease normalization and 0.690 for plant normalization.
    • Demonstrated superior performance compared to methods using only training data or only unlabeled data.
    • Outperformed the best system in the BioCreative V disease normalization task when dictionaries were limited.

    Conclusions:

    • The proposed word embedding approach provides robust performance for normalizing diverse biological entities.
    • The method effectively improves normalization accuracy, especially when domain-specific dictionaries are not comprehensive.
    • The developed model and plant corpus are publicly available for further research and application.