The Influence Select Feature on The Clustering Algorithm

Abstract

 The methods available for structuring the collections are: Classification methods and clustering methods. The following is a summary of the basic principles of the text mining process. Then some of the important methods for classifying the texts are evaluated together. Clustering is the process of organizing anarchy into groups whose components are similar. A cluster is an irregular set of similarities that are heterogeneous with other components of the cluster. The goal of clustering is to achieve a steady and reliable correlation, and to identify the logical connection between them. Therefore, clustering algorithms can be used in a wide range of subject areas. Since clustering results can be varied with the number of terms used, several empirical methods are proposed to diagnose the approximate number of terms that can be expected to provide an appropriate distribution of data among clusters and to define the upper and lower limits of the clustering algorithm. Our goal is to study data mining on different data, the results of this method show that the meta-heuristic method is suitable for the meta cluster algorithm compared to other clusters.