Section Article

Optimising Document Clustering using a Hybrid Genetic–Particle Swarm Model
Author(s): Harmanpreet Kaur

Abstract
Document clustering has become an essential research field in the broader domain of information retrieval text mining machine learning and natural language processing. The exponential growth of unstructured textual data across digital platforms has created a strong demand for automated techniques capable of grouping documents meaningfully without human supervision. Traditional clustering algorithms such as K-means and hierarchical clustering face significant limitations when applied to high-dimensional sparse and semantically complex text data. These algorithms often converge to local optima depend heavily on initial centroids and struggle to capture nonlinear structures within textual feature spaces. Evolutionary optimisation models particularly Genetic Algorithms (GA) and Particle Swarm Optimisation (PSO) have gained increasing attention in computational intelligence due to their global search capability adaptability and robustness in large-scale optimisation problems. A hybrid integration of GA and PSO offers a powerful approach to document clustering by combining GA’s mutation-based diversity with PSO’s velocity-guided convergence behaviour. This hybrid model creates a dynamic search process that balances exploration and exploitation yielding optimised cluster structures and improved semantic coherence among grouped documents.