This dataset provides a starting point to develop a classifier for Moroccan Darija in Arabic writing.
This dataset contains 1.5k classified records, 85% in Moroccan Darija and 15% in Modern Standard Arabic.
Classification has been performed using internal tools by the Sawalni AI project, and is based on the Goud/Goud-sum dataset.
Curated by: Omar Kamali
Funded by: Omar Kamali