Improving Sentiment Analysis of Arabic Tweets by One-way ANOVA

Abstract

Social media is an indispensable necessity for modern life. As a result, it is full of people’s opinions, emotions, ideas, and attitudes, whether positive or negative. This abundance of views creates many opportunities for applying sentiment analysis to the education sector, which reflects how countries and cultures develop. In this research, a real-world Twitter dataset was collected, containing approximately 8144 tweets related to Qassim University, Saudi Arabia. The main aim of this experimental study was to explore the possibility of using a one-way analysis of variance (ANOVA) as a feature selection method to considerably reduce the number of features when classifying opinions conveyed through Arabic tweets. The primary motivation for this research was that no previous studies had examined one-way ANOVA comprehensively to tackle the curse of dimensionality and to enhance classification performance in sentiment analysis for Arabic tweets. Therefore, various experiments were conducted to investigate the effects of one-way ANOVA and to select important features concerning the performance of different supervised machine learning classifiers. Support Vector Machine and Naïve Bayes achieved the best results with one-way ANOVA as compared to the baseline experimental results in the collected dataset. Furthermore, the differences between all results have been statistically analyzed in this study. As further evidence, one-way ANOVA with Support Vector Machine represented an excellent combination across different Arabic benchmark datasets, with its results outperforming other studies.