The purpose of this study is to analyze whether the combination of data mining methods with clustering and classification techniques can be applied to the case of mapping the average number of years of schooling in Indonesia. The data source used in the study is secondary data obtained from the Central Statistics Agency (abbreviated BPS-RI) on the average length of school by province consisting of 34 records (2015-2019). The method used is a combination of kmedoids (clustering) and C4.5 (classification) methods where k-medoids are used to map clusters. The results of the cluster will be processed with C4.5 to see the value of the cluster in the form of a decision tree. The labels used in mapping clustering are high cluster for the average length of school (C1) and low cluster for the average length of school (C2) area. The average length of schooling is one indicator for the dimension of knowledge. The three dimensions are 1) Longevity and healthy living, 2) Knowledge and 3) Decent standard of living. These three dimensions are ways in which the population can access the results of development in obtaining income, health, education, and so on, which is called the Human Development Index. The results of cluster mapping mentioned that there were 9 provinces in the low cluster (26%). The low cluster is Kep. Bangka Belitung, Central Java, East Java, West Nusa Tenggara, East Nusa Tenggara, West Kalimantan, Gorontalo, West Sulawesi and Papua. Based on the decision tree value using the C4.5 method that the low cluster has values <8,763 and> 7,730. This means that for these low clusters the average length of schooling is to junior high school.

Original languageEnglish
Pages (from-to)1811-1816
Number of pages6
JournalARPN Journal of Engineering and Applied Sciences
Issue number17
Publication statusPublished - Sept 2021


  • C4.5 method
  • Indonesia
  • average length of schooling
  • data mining
  • k-medoids method


Dive into the research topics of 'Implementation of the k-medoids method in the old school expectations in indonesia by utilizing educational data mining'. Together they form a unique fingerprint.

Cite this