NLP-Based Sentiment Analysis of the Palestinian-Israeli Conflict

The resurgence of sentiment analysis (SA) has been driven by recent advancements in artificial intelligence (AI), making it a powerful tool for gauging public opinion across various domains. In light of the global attention surrounding the war in Gaza, this study applies and compares several natural language processing (NLP) techniques to analyze public sentiment expressed in social media comments about the conflict. A diverse range of classification approaches was explored, including traditional machine learning and deep learning approaches, as well as fine-tuning a transformer-based model, DistilBERT. The analysis revealed a predominance of negative sentiment in response to the war. Notably, the DistilBERT model outperformed other classifiers, achieving an accuracy of 89%, slightly ahead of the LSTM model at 88%. These results underscore the effectiveness of modern NLP techniques in sentiment analysis and provide valuable insights that can guide future research in this rapidly evolving field.