MACHINE LEARNING AND NLP-BASED SENTIMENT ANALYSIS OF SOCIAL MEDIA POSTS: UNCOVERING PUBLIC EMOTIONS

Social media serves as a crucial platform for sharing opinions and feelings, making it essential for sentiment analysis1 . This paper explores the use of machine learning (ML) and natural language processing (NLP) techniques to categorize sentiments in social media posts as positive, negative, or neutral. It merges lexicon-based approaches, such as VADER2 , with more sophisticated models like RoBERTa3 to tackle issues such as sarcasm, ambiguity, and linguistic variety. The research compares traditional and contemporary methods, emphasizing the greater accuracy of transformer-based model4 . Potential applications include monitoring public opinion, managing brands, and tracking sentiment in real-time5 . By combining ML advancements with specialized knowledge, this study illustrates the capability of sentiment analysis to reveal public emotions and enhance social media analytics.