Comparison between the NB and SVM Methods for Multiclass Arabic Sentiment Analysis

Sentiment analysis gives the ability of monitoring public opinions on different topics. It can be, for example, very helpful in business future plans. Despite the Arabic language being one of the most spoken languages, Arabic sentiment analysis researches face numerous challenges.The main objective of the present paper is to provide a study of Arabic sentiment analysis by comparing the Vector Support Machine (VSM) and Naïve Baye (NB) approaches as a function of dataset classes.