Sentiment analysis is defined as an analysis of text to determine the sentiment expressed within it. This text emphasizes the significance of sentiment analysis in web mining and data classification, with detailed illustrations on sentiment analysis of the Arabic language. This study proposed a sentiment analysis framework to review the Arabic text. Two textual representations were explored: term frequency-inverse document frequency (TF-IDF) and word embedding via Word2vec. Various methods have been suggested for categorizing sentiments in Arabic text based on a dependable dataset, including Long Short-Term Memory (LSTM), hybrid LSTM-CNN, Convolutional Neural Network (CNN), Logistic Regression (LR), Decision Tree (DT), Support Vector Machine (SVM), Multinomial Naïve Bayes (MNB), and Random Forest (RF). The findings indicated that these methods enhanced Accuracy, precision, Recall, and F1-score. The LR and SVM classifiers accomplished the highest Accuracy with 87%, while the other classifiers (LSTM), (CNN-LSTM), (CNN), (MNB), (RF), and (DT) achieved accuracies with 86.41%, 86.10%, 85.26%, 85%, 84% and 81% respectively