Topic Prediction and Trend Analysis in Social Media Data Mining

To enhance the accuracy of topic prediction and trend analysis in social media data mining, an improved block-based single-pass clustering algorithm is employed to handle large-scale data. By incorporating a weighted time decay model and a multi-center network prediction method, the study investigates topic popularity value calculation and trend forecasting. Results indicate that the improved algorithm significantly boosts processing efficiency and prediction accuracy, particularly demonstrating high stability and reliability in dynamic data environments.