Arabic tweeps dialect prediction based on machine learning approach

In this paper, we present our approach for profiling Arabic authors on twitter, based their tweets. We consider here the dialect of an author as important trait to be predicted. For purpose, many indicators, feature vectors and machine learning-based classifiers were implemented. The results these compared find out best prediction model. model was obtained using random forest classifier with full forms stems vector.