Sentiment Analysis of Arabic Tweets on Online Learning During the COVID-19 Pandemic: A Machine Learning and LSTM Approach

Abstract

In response to the unprecedented shift towards online learning during the COVID-19 pandemic, this study presents an innovative analysis of sentiments expressed in Arabic tweets. Utilizing a dataset comprising approximately 100,000 posts from the social media platform X (formerly known as Twitter), collected between 2020 and 2021, sentiments are explored surrounding two prevalent online learning hashtags: (Madrasti platform, translating to 'My School platform') and (Distance Learning). These hashtags were predominantly used by educational professionals, students, and teachers, reflecting their experiences with online education and the Madrasti platform. The dataset was initially imbalanced, with a significant skew towards negative sentiments. To address this imbalance and enhance the reliability of the analysis, Synthetic Minority Over-sampling Technique (SMOTE) and random under-sampling methods were employed. The balanced dataset was then subjected to sentiment analysis using different supervised machine learning (ML) algorithms, including Support Vector Machine (SVM), K nearest neighbor (KNN), and Random Forest, along with the long short-term memory (LSTM) as a deep learning (DL) algorithm. The experiments are conducted using a 10-fold cross-validation approach. The results showed a marked improvement in the precision, recall, and F-measure of the ML algorithms when applied to the balanced dataset, as opposed to theoriginal imbalanced one. The performance of traditional ML classifiers was much betterthan that observed for LSTM. This research offers a detailed analysis of sentiments relatedto online learning during the pandemic and critically assesses different ML techniques inprocessing Arabic language data. The study's innovative approach to balancing the datasetand its extensive evaluation of different algorithms contribute significantly to sentimentanalysis and opinion mining, particularly in the context of online education during a globalhealth crisis.