A Unified AI Framework for Multi-Platform Trend Analysis and Forecasting

This paper presents a unified artificial intelligence framework for multi-platform trend analysis and forecasting by integrating heterogeneous data from social media platforms such as Twitter/X, Reddit, Instagram, and YouTube along with Google Trends. Unlike existing approaches that focus on isolated analytical tasks, the proposed system combines topic modeling (NMF), sentiment analysis (VADER), geo-analytics, influencer detection, and LSTMbased time-series forecasting within a single scalable pipeline. The framework processes largescale real-time data to extract meaningful insights, including emerging topics, sentiment dynamics, and regional engagement patterns. Experimental evaluation demonstrates that the system achieves sentiment classification accuracy of 85–88%, topic coherence of 0.60–0.65, and forecasting RMSE of 0.12, outperforming baseline models such as ARIMA by 12–15%. The proposed approach provides a comprehensive and scalable solution for real-time digital intelligence, enabling researchers, policymakers, and businesses to effectively analyze and predict emerging trends.