We collected data from twitter which is in the context of the proverb and not including the proverb itself, then made clean and preprocessing on the data . We run our classifier and reached to accuracy over 65% and we are currently improving the model.