Natural Language Processing (NLP)-based quantitative risk assessment in Workplace Health and Safety

Traditional safety strategies have been exhausted, necessitating evidence-based safety innovations. One such solution is leveraging digital technology to analyze historical injury data. However, a significant amount of construction site data remains untapped. This project aims to develop tools using Machine Learning (ML), Text Mining, and Large Language Models (LLMs) to analyze injury and accident reports, focusing on improving health and safety performance in the construction sector. Specifically, the project will implement LLM to analyze injury/fatality reports, establish causal links between accident precursors, and develop a quantitative risk-assessment model. This model will quantify the impact of precursors on injury type and severity, enabling timely extraction of critical information for designing efficient injury prevention strategies. We are seeking a Ph.D. candidate with a passion for solving complex global problems and the persistence to work in a multicultural, collaborative academic environment. While a basic understanding of OHS issues is welcome, a background in ML and LLM technology will be advantageous for the candidate's selection.