K-MEANS AND K-MEDOIDS FOR INDONESIAN TEXT SUMMARIZATION
K. K. Purnamasari
Department of Informatics Engineering
Faculty of Engineering and Computer Science
Universitas Komputer Indonesia
Jalan Dipatiukur No . 112-116, Bandung, Indonesia
email : ken.kinanti[at]email.unikom.ac.id
Abstract
Text summarization is the task for taking the main idea in documents. Manually, the summarization performed by humans takes a very long time. There is a need for an in-depth understanding of the contents to obtain the proper summary results. This reason encourages the emergence of various studies about building automatic summarization tools. The system has been built using many methods, such as Clustering. K-Means and K-Medoids are some of the Clustering methods that have been used to summarize text automatically. Several studies using these three methods showed good performance results, with accuracy greater than 50%. However, to be used appropriately, it is necessary to compare which method has the highest accuracy. Based on 50 testing document, average rate of accuracy is 51.16% for K-Means and 63.35% for K-Medoids.
Keywords: kmeans, kmedoids, clustering, summarization, unsupervised learning
Topic: Informatic and Information System