Authors:
Vanessa Ribeiro
1
;
Jeanne Louize Emygdio
1
;
Guilherme Paiva
1
;
Bruno Praciano
1
;
Valério Martins
1
;
Edna Canedo
1
;
2
;
Fábio Mendonça
1
;
Rafael Timóteo de Sousa Júnior
1
and
Ricardo Puttini
1
Affiliations:
1
National Science and Technology Institute on Cyber Security, Electrical Engineering Department, University of Brasília (UnB), P.O. Box 4466, Brasília DF, Brazil
;
2
Department of Computer Science, University of Brasília (UnB), P.O. Box 4466, Brasília DF, Brazil
Keyword(s):
Natural Language Processing, Artificial Intelligence, Public Administration Agency, Jurisprudence, Antitrust.
Abstract:
Natural Language Processing (NLP) and Machine Learning (ML) resources can be used in Jurisprudence to deal more accurately with the large volume of documents and data in this context to provide speed to the execution of processes and greater accuracy to judicial decisions. This article aims to present applied research with a qualitative approach and exploratory objective, technically characterized as a case study. The research was conducted in a Brazilian federal public administration agency to verify the existence of antitrust practices in the pharmaceutical field and the monitoring of such practices by the institution. To this end, a methodological path was established based on three stages: building the corpus, running the NLP pipeline and consultation of the results in the Jurisprudence Search System (BJ System). In compliance with the objective of the case study, it was possible to identify the performance of the agency around the domain elicited, as well as indications of the e
xistence of antitrust practices, since the 276 documents retrieved from the BJ system relate directly to routine processes executed by the agency, either in the sense of investigation, trial or analysis of the business practices.
(More)