Enhancing Social Media Trends using Clustering Algorithm and Swarm Intelligence for Large-Scale Data
Main Article Content
Abstract
The unprecedented growth of social media platforms has led to the generation of vast, complex, and unstructured datasets, creating both opportunities and challenges in extracting meaningful insights. Identifying emerging trends from this data is critical for applications ranging from marketing and political campaigning to public health awareness and disaster response. Traditional clustering algorithms such as K-Means and DBSCAN have been widely used for social media analysis, but they often encounter limitations in scalability, accuracy, and efficiency when dealing with highdimensional, large-scale datasets. To address these challenges, this study proposes a hybrid framework that integrates clustering algorithms with swarm intelligence techniques—specifically, Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO). Using large datasets derived from Twitter and Reddit, the framework was tested on over 3.5 million posts and tweets. The performance was evaluated using cluster quality metrics (Silhouette Score, Davies-Bouldin Index), execution time, and scalability across increasing dataset sizes. Results indicate that KMeans optimized with PSO achieved an 18% improvement in cluster quality and a 25% reduction in computation time compared to conventional clustering methods. Similarly, DBSCAN enhanced with ACO outperformed baseline DBSCAN in cluster cohesion and noise reduction. These findings highlight the potential of hybrid clustering and swarm intelligence models in effectively detecting social media trends, enabling faster, more accurate, and scalable analytics. The proposed approach offers a robust pathway toward real-time social media monitoring systems, contributing significantly to big data analytics, trend forecasting, and decision-making in dynamic digital ecosystems.