Anais EnAJUS 2021

ISSN 2674-8401

Why Natural Language Processing is indispensable to entrench Brazil’s precedents system

Autoria: Guilherme Ramos de Morais


Sessão Presencial 02 - 26/10/2021, Das 12h10 às 14h00 (Horário de Brasília) Das 16h10 às 18h00(Horário de Portugal)
Mediação: Lara Patrício Tavares (Centro de Administração e Políticas Públicas, Instituto Superior de Ciências Sociais e Políticas, Universidade de Lisboa)


Since 1963, Brazil has tried to approximate its civil law system to a system based on precedents, especially due to the fact that it has an enormous litigation rate, with more than seventy-seven million lawsuits in progress. However, in order to consolidate a system of precedents, it is indispensable that tried cases, first of all, may be searchable and able to be found. Stated these preliminary considerations, this draft exposes some manners in which artificial intelligence, by natural language processing, can enhance digital research tools and, consequently, how it can contribute to social benefits such as legal certainty, isonomy, celerity, reduction of Judiciary costs, and due process of law. In order to prove these arguments, this draft was written based on a review of pertinent literature, and is structured as follows: Section 1 introduces the theme with some preliminary considerations; Section 2 explains Brazil’s case law system’s evolution throughout the last sixty years; Section 3 exposes how natural language processing can significantly contribute in this novel system; Subsection 3.1 presents some natural language processing techniques applicable to retrieve information about case law; Section 4 provides critiques regarding the use of artificial intelligence in law; and Section 5 presents the final considerations.


Civil law; Common Law; Natural Language Processing; Legal Certainty; Isonomy.
PDF Todos os trabalhos desta edição