Author : Manpreet Kaur and Navpreet Kaur

Text clustering refers to divide text collection into small clusters and require similarity as large as possible in same cluster. Textual clustering technique was introduced in the area of text mining. The two important goals in text clustering are achieving high performance or efficiency and obtaining highly accurate data clusters that are closed to their natural classes or textual document cluster quality. In order to obtain useful information quickly and accurately form the mass information, text clustering technique is an important research direction. The k-means clustering algorithm has limitations, which depends on the initial clustering center and needs to fix the number of clusters in advance. For these reasons a text clustering algorithm based on latest semantic analysis and optimization is proposed. Thus, a new clustering algorithm based on PBO and optimization has been proposed, which effectively solved the high dimensional and sparse problem and overcomes the dependency of the number of clusters and initial clustering center of k –means algorithm.

Keywords: Text clustering, Latent Semantic Analysis, K-means clustering algorithm, clustering optimization, GA, PSO

Article published in International Journal of Current Engineering and Technology, Vol.4, No.6 (Dec-2014)