Banca de DEFESA: Guilherme Ramos de Morais

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : Guilherme Ramos de Morais
DATE: 22/08/2022
TIME: 10:00
LOCAL: https://teams.microsoft.com/l/meetup-join/19%3ameeting_MGNjYTAxYWQtNmMxYy00YjM1LTgxYz QtOGJkYjY2N2Vj
TITLE:

"Presumptions and guidelines to implement natural language processing in jurisprudential information retrieval: an adoption of AI4SG".

KEY WORDS:

Civil procedural law. Precedents system. Artificial intelligence. AI4SG. Natural language processing. Information retrieval.

PAGES: 152
BIG AREA: Ciências Sociais Aplicadas
AREA: Direito
SUMMARY:

Brazil's precedent system can only operate properly if a previous judicial decision is able to be found. It happens that, given the volume of procedural information in Brazil, it is humanly impossible to catalog relevant information from all judicial decisions, if not by means of automated information extraction systems. And it is precisely at this point that the two problems analyzed in this paper are inserted: first of all, is the Brazilian system of precedents, in theory, capable of generating positive social impacts or not? If it does generate positive impacts, how can it be made operational when faced with the amount of information generated daily in the Brazilian Judiciary? In order to answer these questions, a literature review was conducted to, firstly, analyze the theoretical assumptions of Brazilian’s precedents system and the relevance of artificial intelligence for its effectiveness; then, analyze the ethical principles that should guide the development of artificial intelligence - theory of Artificial Intelligence for Social Goods with adaptations to the law; and, finally, how natural language processing with artificial intelligence can be developed in an efficient and ethical way in retrieving of jurisprudential information, which was synthesized in three theoretical-pragmatic guidelines.

BANKING MEMBERS:
Externo à Instituição - DIERLE JOSÉ COELHO NUNES - PUC MINAS
Interna - 2729067 - DANIELA MARQUES DE MORAES
Presidente - 2250715 - FABIANO HARTMANN PEIXOTO
Interna - 1150035 - FERNANDA DE CARVALHO LAGE