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

Generalizable Natural Language Processing Framework for Migraine Reporting from Social Media.

Yuting Guo1, Swati Rajwal2, Sahithi Lakamana1

  • 1Department of Biomedical Informatics, Emory University, Atlanta, GA.

AMIA Joint Summits on Translational Science Proceedings. AMIA Joint Summits on Translational Science
|June 23, 2023
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

Self-Reported Tianeptine Experiences on Reddit: Natural Language Processing-Assisted Qualitative Study.

JMIR infodemiology·2026
Same author

Courthouses and Craniotomies: Medical Malpractice Indemnity Payments and Litigation Verdicts in Skull Base and Open Cerebrovascular Neurosurgical Practice in the United States throughout the 21st Century.

Journal of neurological surgery. Part B, Skull base·2026
Same author

Extracting patient reported cannabis use and reasons for use from electronic health records: a benchmarking study of large language models.

medRxiv : the preprint server for health sciences·2026
Same author

Monitoring novel psychoactive substance trends on social media: Analysis of discussions and dashboard implementation.

Drug and alcohol dependence·2026
Same author

A cannabis use Reddit dataset for aspect-based sentiment analysis.

Data in brief·2026
Same author

Acetylcholinesterase inhibitory activity of phthalimide derivatives as anti-alzheimer agents: QSAR, ARKA, Hybrid ARKA-RASAR, virtual screening, molecular docking and ADMET studies.

Molecular diversity·2026

Social media platforms like Twitter and Reddit contain valuable self-reported migraine information. Researchers developed a system to identify these posts, enabling new ways to study migraine management and patient experiences.

Area of Science:

  • Neurology
  • Computational Linguistics
  • Public Health

Background:

  • Migraine is a common and debilitating neurological condition.
  • Existing research on real-world migraine management is often limited to traditional sources.
  • Social media offers a vast, largely untapped resource for understanding patient experiences.

Purpose of the Study:

  • To confirm the availability of self-reported migraine data on social media.
  • To create an automated system for detecting migraine-related posts.
  • To evaluate social media's utility for migraine research.

Main Methods:

  • Manual annotation of 5750 Twitter and 302 Reddit posts.
  • Development and evaluation of supervised machine learning models for text classification.

Related Experiment Videos

  • Analysis of detected social media posts for migraine therapies and sentiments.

    • Main Results:

    • A highly accurate text classification system was developed (F1 score 0.90 on Twitter, 0.93 on Reddit).
    • Substantial migraine-related discussions, including therapies and sentiments, were identified on social media.
    • The study demonstrates the feasibility of using social media for migraine research.

    Conclusions:

    • Social media platforms are rich sources of self-reported migraine information.
    • Automated analysis of social media data can provide insights into migraine management.
    • This approach lays the groundwork for large-scale social media-based migraine studies.