Banca de QUALIFICAÇÃO: FERNANDA GUEDES QUEIROZ DE LIRA

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : FERNANDA GUEDES QUEIROZ DE LIRA
DATE: 14/05/2025
TIME: 10:00
LOCAL: Ambiente Virtual ( https://meet.google.com/iiz-njbf-hsm?hs=122&authuser=2 )
TITLE:

ARTIFICIAL INTELLIGENCE AND ACCESS TO JUSTICE: AN ANALYSIS OF THE ENFORCEMENT OF JUDGMENTS IN COLLECTIVE ACTIONS AT THE FEDERAL REGIONAL COURT OF THE 5TH REGION

KEY WORDS:

Artificial Intelligence; Access to Justice; Reasonable Duration of Proceedings; Judgment Enforcement; Collective Actions.

PAGES: 120
BIG AREA: Ciências Sociais Aplicadas
AREA: Direito
SUMMARY:

Artificial intelligence (AI)-based solutions have become increasingly vital within the Brazilian Judiciary, particularly in light of the persistent issue of procedural congestion, which poses significant challenges to the realization of fundamental principles such as legal certainty, access to justice, and the reasonable duration of proceedings. Within this context, the present research focuses on the enforcement phase of collective actions before the Federal Regional Court of the 5th Region (TRF5). The selection of collective actions as the object of study is justified by their increasing procedural complexity, the high number of affected beneficiaries, and their strategic role in fostering procedural economy, ensuring equality of treatment among parties, and enhancing democratic participation in public administration.The study adopts an inductive methodological approach, proceeding from empirical observation of specific judicial cases to broader theoretical generalizations. This inductive method is integrated into a broader empirical strategy of descriptive and exploratory nature, combining both qualitative and quantitative analyses. The research methodology encompasses a comprehensive literature review, documentary analysis, and the practical implementation of a machine learning-based tool.The general objective of the study is to develop and validate a machine learning model, based on the Positive-Unlabeled (PU) Learning technique, capable of identifying patterns in appellate court decisions and assisting in the automated triage of enforcement proceedings related to collective actions. For this purpose, 3,000 judicial decisions were manually labeled, using data obtained through the public version of the TRF5’s Júlia system API, which provides access exclusively to second-instance rulings. These data enabled the training and evaluation of the proposed classifier. The results demonstrate the technical feasibility of automating the triage of enforcement proceedings and suggest significant improvements in terms of procedural efficiency, predictability, and case management quality. The study concludes that the application of artificial intelligence in this context not only contributes to the acceleration of judicial workflows involving collective actions but also reinforces key principles of access to justice and judicial effectiveness. Moreover, it offers practical insights for institutional and normative improvements within the TRF5 and highlights promising directions for future research on the integration of technological solutions in the judiciary.

COMMITTEE MEMBERS:
Presidente - 2353000 - ELIAS JACOB DE MENEZES NETO
Interno - 1254860 - FABRICIO GERMANO ALVES
Interno - 1051231 - LUCIANO ATHAYDE CHAVES
Interno - 1358062 - MARCO BRUNO MIRANDA CLEMENTINO