International Journal Of Imaging

Abstract

Deep Learning-based social media Trend Analyzer to Predict the Trends Over Time

Author: Sarasani Veda Reddy, Peechara Sai Mani Teja , Vallabhaneni Dhrushya Sree,Mr. Nagaraju

Abstract:

Social media platforms like Twitter generate over 500 million tweets per day, with a 72% increase in global user engagement since 2019. Analyzing and predicting trends over time remains a significant challenge due to the dynamic nature of user interactions and content generation. Predicting what content will become a trend remains a complex task. With millions of tweets posted every day, identifying potential trends manually is nearly impossible. The dynamic and fast-paced nature of Twitter interactions makes it difficult to rely on traditional methods like keyword analysis or manual monitoring. As more companies and influencers seek to capitalize on trending topics for engagement, a reliable automated system for predicting trends over time is essential. This research proposes a novel Deep Learning (DL) classification model that predicts trends based on real-time Twitter data. The preprocessing pipeline includes text cleaning, tokenization, and feature extraction, followed by tr