Implementation of OLAP and K-Medoids Clustering for Accreditation Data Analysis of Study Programs

Abstract

At present to maintain the quality of the data in the study program is very much needed, given the current accreditation based on PDDIKTI(Pangkalan Data Pendidikan Tinggi) data which must be reported every semester. Higher education data are generally still separate, there is no valid data warehouse and system for data analysis so that it complicates data quality control. The purpose of this study is to apply data warehouse, OLAP (Online Analytical Processing) and k-medoids clustering for data prediction and control according to the study program's accreditation self-evaluation report matrix. This study uses data the last 3 years for prospective new students, students, students, study periods, achievements and lecturers to be analyzed using OLAP and the k-medoids cluster. The results showed OLAP can see information in an informative, real-time data accreditation matrix and the k-medoids cluster produces accurate cluster estimates with an evaluation value of Davies Bouldin Index of 0.2927 and said to be a good cluster