HOME       POLICIES       SUBMISSION        PEOPLE        ARCHIVES         CONFERENCES        CONTACT

ABSTRACT


The Brazil’s Supreme Court deal with more than a hundred thousand lawsuits. Thats is a huge number per each one of its eleven ministers. Each one of this suitcase must be separated by theme (“Tema de Repercussão Geral”) before its final analysis. In order to reduce the time consumed in this workflow, we present an automation of identifying five important suitcase pieces for the themes classification task. In this paper we present results of specific convolutional neural network model.


KEYWORDS

Natural Language Processing, Convolutional Neural Networks, Document Classification.

ARCHIVES
To return to the Volume/Number webpage, click here.
THE INTERNATIONAL JOURNAL OF FORENSIC COMPUTER SCIENCE - IJoFCS

Volume 14, Number 1, pages 8-15, DOI: 10.5769/J201901001 or http://dx.doi.org/10.5769/J201901001


Lawsuit Documents Classification using a CNN for Brazil’s Supreme Court


By  Nilton Correia da Silva, Fabricio Ataides Braz, Teófilo Emídio de Campos, Davi Benevides Gusmão, Felipe Borges Chaves, Davi Alves Bezerra, Gabriel Gomes Ziegler, Pedro Henrique Gonçalves Inazawa, and Victor Hugo Dias Coelho



To download this paper, click here.