MINERJUS: solução de apoio à classificação processual com uso de Inteligência Artificial
The electronic court case is a reality in Brazil, where 70% of new cases in all judiciary are virtual. Making proper use of this reality and improving it is paramount to meet the demand for approximately 25 million new processes per year. This project proposes an improvement in the speed and correct...
| Autor principal: | Sousa, Rogério Nogueira de |
|---|---|
| Grau: | Dissertação |
| Idioma: | pt_BR |
| Publicado em: |
Universidade Federal do Tocantins
2019
|
| Assuntos: | |
| Acesso em linha: |
http://hdl.handle.net/11612/1446 |
| Resumo: |
|---|
|
The electronic court case is a reality in Brazil, where 70% of new cases in all judiciary are virtual. Making proper use of this reality and improving it is paramount to meet the demand for approximately 25 million new processes per year. This project proposes an improvement in the speed and correctness of the electronic process classification through the use of Artificial Intelligence. Thus, it is conjectured to assist legal operators responsible for the registration of the initial petition document (creation of the process), as well as those responsible for its analysis, by means of automatic and assertive suggestion on the subject matter of the process, providing greater agility of processing and quality to the information contained in the Brazilian court records. During the development of the solution several machine learning algorithms were tested identifying the best performing, in this case Support Vector Machine, accuracy and precision as well as the response time and training for efficient process classification. |