This repository contains replications of models described in the paper VICTOR: a dataset for Brazilian legal documents classification
This repository contains implementations of models to perform the document classification task described in the VICTOR: a dataset for Brazilian legal documents classification paper. In addition, this repository contains variations of the paper’s models.
The source code from the paper’s authors is available at https://github.com/peluz/VICTOR-dataset
The deep learning models in this repository were implemented in PyTorch (the paper’s authors utilized Keras).
The Victor dataset contais documents from Brazilian Federal Supreme Court (aka STF) lawsuits. The acess to it can be required at http://www.gpam.unb.br/datasets/victor