Sentiment analysis of Facebook users towards COVID-19 vaccination

Abstract

Social networks have become the first-leading virtual space for expressing and sharing people's opinions. Therefore, many sentiment analysis practitioners are focusing on gathering and analyzing the content generated by different social network users. Believes, thoughts, and events that create controversy are the resources for getting various kinds of feedback, making them the main fuel for different areas of research such as sentiment analysis, hate speech detection, and fake news identification. In this paper, we present a sentiment analysis applied to comments that are publicly shared on Facebook. These comments are expressed in Moroccan Arabic and convey the viewpoints of Moroccan citizens toward the COVID-19 vaccination. In doing so, we collected comments from the Facebook pages of official Moroccan newspapers. Then, we manually annotated the comments compiled into two categories, namely positive and negative. Furthermore, the TF-IDF and Chi-square test were used for data extraction and selection. Finally, we implemented three classifiers namely Naïve Bayesian (NB), Support Vector Machines (SVM), and Random Forests (RF). The results showed that the RF-based classifier achieved the best performance in terms of accuracy and F1-score metrics.