Sentiment analysis algorithms: evaluation performance of the Arabic and English language

Abstract

Usage of social media like Facebook, WhatsApp, Twitter, and Blogs is rapidly increasing in recent years. These platforms allow people to freely write comments and share their opinions, ideas and suggestions that can be either positive, negative or neutral comments on various topics such as politics, business, advertisement, and entertainment. Several, Machine Learning $(ML)$ algorithms such as Naive Bayes $NB$ and Decision Tree $DT$ are used with sentiments analysis technique in different languages to understand the opinions of people in social media. In this paper, we evaluate and discussed the application of $NB$ and $DT$ in sentiment analysis using a multi-dataset in different languages to understand which can give a better result when used with $ML$ algorithms. Multi-language dataset such as English, modern standardArabic and dialectArabic are collected for the experiment. We evaluate is based on two parameters which are accuracy and runtime. The result of our experiment shows some significant.