Uncovering Hidden Patterns in Social Media Data Using Advanced Network Analytics and Sentiment Analysis

Abstract

In the digital era, social media platforms have emerged as critical spaces for public discourse, marketing, and socio-political mobilization. This study explores the intersection of network analytics and sentiment analysis to uncover latent structures and emotional trends in large-scale social media datasets. By integrating graph-based community detection with transformer-based sentiment classification models, the paper presents a comprehensive methodology to reveal underlying patterns that are not readily visible through traditional metrics. Empirical analysis was conducted using Twitter data collected during the 2024 Indian general elections, uncovering polarized communities and significant temporal sentiment shifts. The findings have implications for public opinion modeling, misinformation detection, and social network research.