Effectiveness of Natural Language Processing Based Machine Learning in Analyzing Incident Narratives at a Mine

Rajive Ganguli,Preston Miller,Rambabu Pothina

doi:10.3390/min11070776

Abstract

To achieve the goal of preventing serious injuries and fatalities, it is important for a mine site to analyze site specific mine safety data. The advances in natural language processing (NLP) create an opportunity to develop machine learning (ML) tools to automate analysis of mine health and safety management systems (HSMS) data without requiring experts at every mine site. As a demonstration, nine random forest (RF) models were developed to classify narratives from the Mine Safety and Health Administration (MSHA) database into nine accident types. MSHA accident categories are quite descriptive and are, thus, a proxy for high level understanding of the incidents. A single model developed to classify narratives into a single category was more effective than a single model that classified narratives into different categories. The developed models were then applied to narratives taken from a mine HSMS (non-MSHA), to classify them into MSHA accident categories. About two thirds of the non-MSHA narratives were automatically classified by the RF models. The automatically classified narratives were then evaluated manually. The evaluation showed an accuracy of 96% for automated classifications. The near perfect classification of non-MSHA narratives by MSHA based machine learning models demonstrates that NLP can be a powerful tool to analyze HSMS data.

Highlights

Workers’ health and safety is of utmost priority for the sustainability of any industry
According to the recent estimates published by the International Labour Organization (ILO), 2.78 million workers die from occupational accidents and diseases worldwide [1]
That is further strengthened by the low false positive rates for the distinct categories, i.e., when a particular model for a distinct category claims that a narrative belongs to that category, the classification is most likely valid

Summary

Introduction

Workers’ health and safety is of utmost priority for the sustainability of any industry. 374 million workers suffer from non-fatal accidents, and lost work days represent approximately 4% of the world’s gross domestic product [2,3]. It is, not surprising that researchers are constantly investigating factors that impact safety [4,5], or finding innovations and technology to improve safety [6,7]. As to the U.S mining industry, for years 2016–2019, the National Institute for Occupational Safety and Health (NIOSH), a division of the US Centers for Disease Control and Prevention (CDC) reports 105 fatal accidents and 15,803 non-fatal lost-time injuries [8]. With the advances in natural language processing (NLP), there is an opportunity to create NLP-based tools to process and analyze such textual data without requiring human experts at the mine site

Objectives

Methods

Discussion

Conclusion