PSAM 16 - Paper Overview

PSAM 16 Conference Paper Overview

Welcome to the PSAM 16 Conference paper and speaker overview page.

Lead Author: Plínio Ramos Co-author(s): July B. Macedo - julybias@gmail.com Caio B. S. Maior - caio.maior@ufpe.br Márcio C. Moura - marcio.cmoura@ufpe.br Isis D. Lins - isis.lins@ufpe.br

Combining BERT with numerical features to classify injury leave based on accident description
Workplace safety is a major concern in many industries. In this context, accident investigation reports provide useful knowledge to support companies to propose preventive and mitigative measures. However, the information presented in accident reports databases is normally large, complex, also filled out with redundant data. Thus, a complete human review of the entire database is arduous, considering numerous reports produced by a company. Therefore, natural language processing (NLP)-based techniques are suitable for analyzing a massive amount of textual information. In this paper, we adopted NLP techniques to determine whether or not an injury leave would be expected. The methodology was applied on 648 accident reports collected from an actual hydroelectric power company and focused on the accident agent categories. We employ Bidirectional Encoder Representations from Transformers (BERT), a state-of-art natural language processing method, to tackle the aforementioned problem. The text representations provided by BERT model were combined with numerical and binary features extracted from the accident reports. These combined features are input to an MLP that predicts the occurrence of the accident leave for a given accident. Indeed, accident investigation reports to provide a useful knowledge to support decisions in the safety context.

Combining BERT with numerical features to classify injury leave based on accident description

Workplace safety is a major concern in many industries. In this context, accident investigation reports provide useful knowledge to support companies to propose preventive and mitigative measures. However, the information presented in accident reports databases is normally large, complex, also filled out with redundant data. Thus, a complete human review of the entire database is arduous, considering numerous reports produced by a company. Therefore, natural language processing (NLP)-based techniques are suitable for analyzing a massive amount of textual information. In this paper, we adopted NLP techniques to determine whether or not an injury leave would be expected. The methodology was applied on 648 accident reports collected from an actual hydroelectric power company and focused on the accident agent categories. We employ Bidirectional Encoder Representations from Transformers (BERT), a state-of-art natural language processing method, to tackle the aforementioned problem. The text representations provided by BERT model were combined with numerical and binary features extracted from the accident reports. These combined features are input to an MLP that predicts the occurrence of the accident leave for a given accident. Indeed, accident investigation reports to provide a useful knowledge to support decisions in the safety context.

PSAM 16 Conference Paper Overview

Welcome to the PSAM 16 Conference paper and speaker overview page.

Paper PL113 Preview

Author and Presentation Info

Download paper PL113.

Download the presentation PowerPoint file.