Maintenance Notice

Due to necessary scheduled maintenance, the JMIR Publications website will be unavailable from Wednesday, July 01, 2020 at 8:00 PM to 10:00 PM EST. We apologize in advance for any inconvenience this may cause you.

Who will be affected?

Accepted for/Published in: Journal of Medical Internet Research

Date Submitted: Jan 3, 2025
Date Accepted: Aug 21, 2025

The final, peer-reviewed published version of this preprint can be found here:

Public Perspectives on Palliative and Hospice Care: Social Media Content Analysis Using Topic Modeling and Multiclass Sentiment Analysis

Kim A, Woo K

Public Perspectives on Palliative and Hospice Care: Social Media Content Analysis Using Topic Modeling and Multiclass Sentiment Analysis

J Med Internet Res 2025;27:e70836

DOI: 10.2196/70836

PMID: 40957008

PMCID: 12485256

Analyzing Public Perspectives on Palliative and Hospice Care: Topic Modeling and Multi-class Sentiment Analysis in Social Media

  • Aeri Kim; 
  • Kyungmi Woo

ABSTRACT

Background:

Palliative care enhances dignity and quality of life for patients with serious illnesses by managing distressing symptoms and supporting families. However, inadequate awareness and misconceptions often hinder patients and their families from accessing these services. Understanding of public perspectives on palliative and hospice care can facilitate the development of targeted educational resources and awareness campaigns. As social media becomes an important source of health information, analyzing such publicly available online sources can yield valuable insights into perceptions of palliative and hospice care.

Objective:

This study analyzed public perspectives posted on a popular social media platform in South Korea, to understand perceptions, challenges, needs, and sentiments related to palliative and hospice care.

Methods:

Data were collected from Naver Knowledge iN, a popular online public forum in South Korea, encompassing 34,501 texts posted between 2002 and 2024. After applying inclusion and exclusion criteria, 9147 relevant perspectives were analyzed. Contextualized topic modeling was used to identify themes, and the optimal model was selected based on coherence, diversity scores, and expert feedback. In addition, multi-class sentiment analysis using a fine-tuned KoBERT model classified sentiments into six categories. The multi-class sentiment model’s performance was evaluated using accuracy, precision, recall, and F1-score.

Results:

Social media discussions on palliative and hospice care have increased steadily over time. Topic modeling identified nine themes, with “ethical and legal concerns” and “medical care in hospitals” peaking in recent years, suggesting growing public interest in these areas. “Emotional and psychological support” emerged as the predominant theme, reflecting a significant need for psychosocial assistance among patients and their families. Sentiment analysis revealed that “sadness,” “anxiety,” and “neutral” were common emotions across many topics. Notably, themes such as “emotional and psychological support,” “disease treatment outcomes and prognosis,” “medical care in hospitals,” “financial issues,” and “symptom management” were predominantly associated with “sadness,” while “administrative and volunteer services,” “ethical and legal concerns,” and “nutrition management” were more closely linked with “anxiety.”

Conclusions:

This study highlights public concerns about palliative and hospice care in South Korea, including ethical dilemmas, caregiving burden, and emotional distress. Findings underscore the need for communication strategies that address informational, emotional, and psychological needs. Culturally sensitive, interactive communication tools, such as AI-powered chatbots and public education campaigns, may help dispel misconceptions and promote timely, informed decisions about palliative and hospice care.


 Citation

Please cite as:

Kim A, Woo K

Public Perspectives on Palliative and Hospice Care: Social Media Content Analysis Using Topic Modeling and Multiclass Sentiment Analysis

J Med Internet Res 2025;27:e70836

DOI: 10.2196/70836

PMID: 40957008

PMCID: 12485256

Download PDF


Request queued. Please wait while the file is being generated. It may take some time.

© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.