Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 10, 2025Journal of Computer Science and Informatics Engineering (CoSIE)

Comparative Performance of IndoBERT and IndoLEM Baseline Models for Post-Disaster Health Information Extraction from Indonesian Online News

View Full Paper
Ask AI
Bookmark
Share

Authors

NINalar IstiqomahFNFanny Novika

Discussion

Loading...

Member takes

Overview

Performance evaluation shows IndoBERT extracted health impacts better than baseline models, indicating its utility in disaster monitoring.

Key Points

  • IndoBERT achieves 90.00% accuracy in extracting health impacts from Indonesian news, outperforming baseline models.
  • The F1-score for IndoBERT is 88.26%, significantly higher than mBERT at 72.93% and XLM-R at 76.44%.
  • Fine-tuning using 1,137 disaster-related articles improved IndoBERT's performance in Named Entity Recognition tasks.
  • Findings indicate that floods are linked to diarrhea and skin diseases, while volcanic eruptions relate to respiratory infections.

Cite This Study

Istiqomah et al. (2025) studied this question.

synapsesocial.com/papers/68c1aad354b1d3bfb60e38cdhttps://doi.org/10.55537/cosie.v4i3.1174
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Sentimen Analysis Social Media for Disaster using Naïve Bayes and IndoBERT2024
  2. 2Indonesian disaster named entity recognition from multi source information using bidirectional LSTM (BiLSTM)2024 · 6 citations
  3. 3EVALUATION OF INDOBERT AND ROBERTA: PERFORMANCE OF INDONESIAN LANGUAGE TRANSFORMER MODELS IN SENTIMENT CLASSIFICATION2025 · 2 citations
  4. 4COMPARATIVE PERFORMANCE OF TRANSFORMER AND LSTM MODELS FOR INDONESIAN INFORMATION RETRIEVAL WITH INDOBERT2025
  5. 5Implementation of Text Mining for Evaluating the Relevance Between News Headlines and Content on a Web-Based Platform2025