Development of Text Clustering Method with K-Means for Analysis of Text Data

Abstract

Clustering is a widely used unsupervised data mining technique. In clustering, the main aim is to put similar data objects in one cluster and dissimilar in another cluster. The k-implies is the most famous clustering algorithm because of its effortlessness. But the performance of the k-means clustering algorithm depends upon the parameter selection. Parameter selection like number of cluster and initial cluster center are key of k-means algorithm. Distance augmentation method, density method quadratic clustering methods are utilized to initial cluster selection. This paper examination five unique methods, for example, improved k-means text clustering algorithm, revisiting k-means, LMMK algorithm, SELF-DATA architecture, Clustering Approach for Relation e.t.c. But these techniques have some limitations. To improve these approach, this paper has proposed the development of text clustering method with k-means for analysis of text data.