Related Experiment Video
Updated: Jun 27, 2025
07:28
A Contusive Model of Unilateral Cervical Spinal Cord Injury Using the Infinite Horizon Impactor
Published on: July 24, 2012
19.6K
Contextualizing injury severity from occupational accident reports using an optimized deep learning prediction model.
Mohamed Zul Fadhli Khairuddin1, Suresh Sankaranarayanan2, Khairunnisa Hasikin3
1Institute of Medical Science Technology, Universiti Kuala Lumpur, Kajang, Selangor, Malaysia.
Peerj. Computer Science
|April 25, 2024
Summary
This study uses deep learning to predict occupational injury severity from narratives. The optimized Bidirectional Long Short-Term Memory (Bi-LSTM) model achieved high accuracy, offering insights for workplace safety.
Area of Science:
- Occupational Safety and Health
- Data Science
- Machine Learning
Background:
- Conventional occupational injury assessment relies on structured data.
- Unstructured injury narratives contain rich, underutilized information.
- A novel deep learning approach is proposed to analyze these narratives.
Purpose of the Study:
- To develop and validate a deep learning model for predicting occupational injury severity.
- To extract actionable insights from unstructured injury narratives.
- To enhance occupational injury assessment and proactive safety measures.
Main Methods:
- Utilized Natural Language Processing (NLP) on US Occupational Safety and Health Administration (OSHA) data (2015-2023).
- Applied Term Frequency-Inverse Document Frequency (TF-IDF) and Global Vector (GloVe) embeddings.
- Developed and optimized a Bidirectional Long Short-Term Memory (Bi-LSTM) model with feature importance analysis.
Main Results:
- Optimized Bi-LSTM model achieved high accuracy: 0.95 for hospitalization, 0.98 for amputation.
- Demonstrated faster model processing times compared to traditional classifiers.
- Feature importance analysis identified keywords linked to injury causes, enhancing interpretability.
Conclusions:
- The optimized Bi-LSTM model provides an effective tool for improving workplace safety.
- Findings support proactive safety measures, contributing to business productivity and sustainability.
- Establishes a foundation for predictive analytics in occupational safety and health.
