Public Opinion and Emotion Analysis on the Israel-Hamas War Using Machine Learning
摘要
At the beginning of October 2023, Israel and Hamas went to war. The battle resulted in both geopolitical fallout and casualties, sharply intensifying the Israeli-Palestinian conflict. Social media had a critical role in shaping public opinion and sentiment around the conflict, which attracted worldwide attention. The quick spread of information shaped public opinion, the dynamic framing of images, and the amplification of voices on social media sites. The research used machine learning techniques, specifically sentiment analysis and emotion analysis, to obtain insights into the thoughts and sentiments in social media. The study covered 80 days and concentrated on Reddit discussions about the Israel-Hamas conflict. The sentiment analysis showed that neutral, positive, and negative feelings were in order of dominance. Hugging Face's Transformers emotion analysis revealed a majority of negative feelings, with disgust ranking highest and being followed by sadness, happiness, fear, anger, and surprise. The complicated emotional responses underscored people's wide range of emotions throughout the war. Applying machine learning in sentiment analysis, especially on platforms like Reddit, proves beneficial for analyzing large-scale datasets. Further research exploring the temporal evolution of sentiments during different phases of conflicts could enhance our understanding of the dynamic nature of public sentiment in geopolitical events.