Recomendação de documentos no domínio jurídico para o projeto Querido Diário

This work proposes a recommendation system for official diaries content available in “Querido Diário”.

For tokenization of the input data, we use the tokenizer of the trained agent in BERTimbau (BERT agent trained in Portuguese).

We use BERTimbau's embeddings to feed a KNN classifier, which is used in the semantic content recommendation task.

Name		Name	Last commit message	Last commit date
Latest commit History 12 Commits
1st_test_Querido_Diario.ipynb		1st_test_Querido_Diario.ipynb
2nd_test_Querido_Diario+knn(n=1).ipynb		2nd_test_Querido_Diario+knn(n=1).ipynb
3rd_test_Querido_Diario+knn(n=1) (1).ipynb		3rd_test_Querido_Diario+knn(n=1) (1).ipynb
Final_Project_NLP Recommendation of documents in the legal domain.pdf		Final_Project_NLP Recommendation of documents in the legal domain.pdf
LICENSE		LICENSE
README.md		README.md
readD.csv		readD.csv
saveFiles.ipynb		saveFiles.ipynb
tags.xlsx		tags.xlsx

Provide feedback