Tamanho da fonte: 

"The Brazilian judicial system is currently one of the largest in the world with more than 77 million cases awaiting decision. One of the most slow steps is the identification of relevant information within the initial petition, with the aid of intelligence techniques, it would be possible to get around this problem. This article aims to build models capable of classifying health demands in the judiciary between requests for ICU, surgery, examination and drug supply. In addition to identifying information related to demand, such as diseases of the author and identification of the drug or surgery requested. For this, information retrieval models based on dictionaries, rules or hybrids were used. Within this article, the models obtained are evaluated on the recall rate. Among the models, the best result had a 90.22% recall rate, while the worst had a 37.03% recall rate."