Arabic Text Classification in the Legal Domain

The field of law in Morocco is behind on progress to other fields, where the archiving of legal documents is done manually and the access to the information is increasingly complex. In the scope of this work, we investigate the application of text classification methods to support law professionals, by proposing a system able to automatically select the class of a judgement. We will build four models which are Naive Bayes, K-Nearest Neighbor, Support Vector Machine and Decision Tree to classify Arabic legal documents, which are tried by the Moroccan court.