Comparative Analysis on Social Media Sites Using Sentiment Analysis

Abstract-This paper evaluates user views and emotional tone in postings across many social media sites by means of a comparative analysis utilising sentiment analysis. Understanding the mood underlying user-generated material has become vital for companies, marketers, and academics as social media is playing more and more influence on public debate. Focussing on sites like Twitter, Facebook, and Instagram, the paper uses sentiment analysis methods on social media data. The performance of these models in terms of accuracy, precision, recall, and F1 score is compared using machine learning models including Support Vector Machines (SVM), Light GBM (LGBM), and Long Short-Term Memory (LSTM). The results expose how sentiment patterns vary on different platforms, therefore offering understanding of public opinion dynamics, brand perception, and content engagement. Following LGBM in precisely identifying sentiment, the study emphasises SVM and LSTM’s efficiency and analyses the ramifications of these results for content development, market research, and social media monitoring.