Banca de DEFESA: JULY BIAS MACEDO

Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.
STUDENT : JULY BIAS MACEDO
DATE: 20/12/2022
TIME: 13:00
LOCAL: Online
TITLE:

DEVELOPMENT OF NATURAL LANGUAGE PROCESSING-BASED SOLUTIONS FOR RISK ANALYSIS: APPLICATIONS TO A HYDROPOWER COMPANY AND AN O&G INDUSTRY

KEY WORDS:

Risk analysis, accident investigation reports, natural language processing, text mining, oil refineries, hydroelectric power company.

PAGES: 128
BIG AREA: Engenharias
AREA: Engenharia de Produção
SUMMARY:

Risk Analysis (RA) is crucial to prevent and mitigate potential risk events; however, there are several challenges related to RA. For instance, accident investigation reports are useful sources of information to support safety professionals to propose measures to prevent or mitigate identified occupational accident root causes. Nevertheless, reports’ low quality and lack of detail may limit their usefulness. Moreover, the quality of Quantitative Risk Analysis (QRA) strongly relies on the identification of all potential hazards with major consequences related to the operation of an industrial system, which is usually performed by multiple experts and consumes a considerable amount of time and effort. Since valuable knowledge about an industrial system is stored in the form of textual data, Natural Language Processing (NLP) techniques can be helpful since it can be applied to extract, organize, and classify information from text. Although several studies contributed to the advance of RA, most studies applying NLP focus primarily on automatically identifying patterns from reactive data, such as accident reports, and do not consider the quality of information contained in these documents. In addition, different forms of text data store relevant knowledge about industrial systems and their respective risks, especially proactive data such as documents resulting from preliminary risk studies, and adoption of these data could support preventive risk studies. The main purpose of this study is to develop NLP-based solutions to different issues faced in RA. Thus, this thesis presents NLP-based methodologies to (i) support the diagnosis of the content quality and the understanding of accident investigation reports and (ii) automatically identify risk features from documents to support the initial stage of QRA. First, we analyzed a dataset of accident investigation reports of a hydroelectric power company through different traditional NLP approaches. The outputs from the models allowed us to identify problems in the accident reports and propose improvements as well as check the positive impact of the proposed changes. Therefore, we showed the importance of the company’s safety culture to keep safety technicians engaged in carefully constructing an accident database. Second, we showed that information contained in past risk studies of an oil refinery can be reused to train models to identify different risk features. The proposed methodology can support risk analysts in the initial stage of QRA to identify and assess potential hazards associated to the operation of an industrial system.

COMMITTEE MEMBERS:
Externo à Instituição - PIERO BARALDI
Interna - 2732514 - ISIS DIDIER LINS
Interno - 2805590 - MARCELO HAZIN ALENCAR
Presidente - 2766188 - MARCIO JOSE DAS CHAGAS MOURA
Externa à Instituição - MARILIA ABILIO RAMOS