Hybrid PSO and GA Models for Document Clustering

This paper presents Hybrid Particle Swarm Optimization (PSO) -Genetic Algorithm (GA) approaches for the document clusteringproblem. To obtain an optimal solution using Genetic Algorithm, operation such as selection, reproduction, and mutation proceduresare used to generate for the next generations. In this case, it ispossible to obtain local solution because chromosomes or individualswhich have only a close similarity can converge. In standard PSOthe non-oscillatory route can quickly cause a particle to stagnate andalso it may prematurely converge on suboptimal solutions that arenot even guaranteed to local optimal solution. This work proposeshybrid models that enhance the search process by applying GAoperations on stagnated particles and chromosomes. GA will becombined with PSO for improving the diversity, and the convergencetoward the preferred solution for the document clustering problem. The approach efficiency is verified and tested using a set ofdocument corpus. Our results indicate that the approaches arefeasible alternative to solve document clustering problems