Ontology-based Semantic Text Documents Clustering Using Particle Swarm Optimization

Abstract

The World Wide Web, the largest shared information source is growing exponentially and the amount of business news on the web is overwhelming and need to be handled properly. As such, grouping the web document into cluster for speedy information retrieved becomes imperative. Clustering technique organizes a large quantity of unordered text documents into a small number of meaningful and coherent clusters, thereby providing a basis for intuitive and informative navigation and browsing mechanisms. The quality of clustering result depends greatly on the representation of documents and the clustering algorithm. In traditional document representation methods, the frequency count of the document terms is used for the feature vector representing the documents. But traditional document representation methods cannot identify related terms semantically. Documents written in human language contain contexts and the words used to describe these contexts are generally semantically related. Motivated by this fact, domain ontology is developed to promote the enrichment of semantic representation of terms. Then, Particle Swarm Optimization (PSO) clustering algorithm is used to cluster the web documents efficiently. The paper constitutes the comparative results of using PSO algorithm only and PSO algorithm with Ontology for clustering web documents. According to the analytical results, the representation of terms by using Ontology is significantly efficient and the implementation of PSO algorithm achieves better performance in intra cluster and inters cluster similarity.