What question did this study set out to answer?

The aim is to leverage Large Language Models to improve the interpretation of safety climate survey data.

March 23, 2026Procedia Computer Science0 citationsOpen Access

Bridging Quantitative Safety Analytics and Computational Linguistics: An LLM-Augmented Data-to-Text Framework

Key Points

The aim is to leverage Large Language Models to improve the interpretation of safety climate survey data.
Analyzed a dataset of safety climate surveys from 2015 to 2024.
Trained Large Language Models (LLMs) on this dataset.
Cooperated with experts in computational linguistics to interpret findings.
LLMs effectively convert complex statistical analysis into understandable insights.
Data-to-text production shows promise in providing clear recommendations for safety improvement.

Abstract

This study explores a dataset of a decade of safety climate surveys from 2015 to 2024 that has been used to train Large Language Models (LLMs) to suggest practical recommendations to enhance safety organisation in this oil and gas company. The Study shows how LLMs can be used as an assistant in converting the complicated statistical analysis into detailed insights that connect the findings with real language, which was performed via cooperating LLMS with a specialist in computational linguistics. The results show the effectiveness of data-to-text production as a translation and interpretation process.

Perguntar à IA

Bookmark

View Full Paper