Hybridization of a Social Spider Optimization Algorithm with Differential Evolution for Text Document Clustering Using Single Cluster Approach

Aasheesh Shukla and Vishal Goyal

Text documents can be clustered based on their content by Text Document Clustering. The field like unsupervised document organization, automatic topic extraction, information retrieval and text data mining uses this as a fundamental operation. The effective navigation of information, summarization as well as the organization is depends on the document clustering algorithms with high quality and speed. The web news articles, research paper abstract can be clustered. In this paper, intelligent hybridization of Social Spider Optimization (SSO) algorithm and Differential Evolution (DE) named as SSO-DE is proposed. The effective exploitation and exploration can be balanced by this approach. The search scope of SSO algorithm is controlled and adjusted dynamically by the introduction of a weighting factor which changes with iteration. A mutation operator is used after the completion of social-spiders search. This is used to jump out of potential local optima as well as to improve the global search ability. The effectiveness of the algorithm in Evolution for Text document clustering is shown by the experimental results

Volume 11 | 10-Special Issue

Pages: 642-646

DOI: 10.5373/JARDCS/V11SP10/20192853