Hybrid Reduction Dimension on Clustering Text of English Hadith Translation

Abstract

Clustering results are strongly influenced by the selected technique and data dimensions. Large data dimensions become the main problem that must be considered. Therefore, a dimensional reduction is needed to select sub-feature that provides important information. One of the dimensions reduction methods is the hybrid method. The hybrid method combines the method of feature selection and feature extraction to select informative sub-feature. Furthermore, the simplest clustering technique is the k-means algorithm, which divides n data into k cluster based on the centroid. This study carried out clustering using the k-means algorithm after reducing dimensions on 892 English translation hadith documents. The clustering results are validated using the silhouette coefficient and Davies Bouldin index (DBI). Experimental results show that dimensional reduction can improve the cluster quality.