<![CDATA[This study investigates public sentiment patterns during the 2024 U.S. Presidential Race through machine learning analysis of social media data from X (formerly Twitter). Using a dataset of 500 annotated tweets collected from Kaggle, we employ BERT-based sentiment analysis, temporal engagement tracking, and Latent Dirichlet Allocation (LDA) topic modeling to examine discourse across five major candidates. The analysis reveals predominantly positive sentiment (54.2%) in political discussions, with established party candidates receiving higher positive engagement. Temporal analysis demonstrates strong correlations between major campaign events and public engagement, with presidential debates generating peak interaction levels. Topic modeling identifies five key themes driving voter discourse: economic policy, healthcare, climate change, social justice, and foreign policy. Positive content consistently achieved 20-30% higher engagement rates than negative content, though negative sentiments showed sharp spikes during controversies. Our findings contribute to understanding digital political discourse dynamics and offer practical insights for campaign strategy in the social media era. The study's limitations include platform-specific constraints and a two-month observation period, suggesting opportunities for cross-platform analysis in future research.]]>
Copyrights © 2024
