Application of Data Mining Using the K-Means Algorithm in Rural and Urban Land and Building Tax (PBB-P2) Receivables Data in Bantul Regency

Abstract

The rural and urban land and building tax (PBB-P2) receivable amount in Bantul is quite large. As of December 31, 2019, there were 3,344,145 PBB-P2 objects, worth for IDR 114,984,991,600. This number tends to increase from year to year, showing that PBB-P2 receivable collection process has not optimal. This study discusses data mining applications for data management using the K-Means clustering method. This paper uses the PBB-P2 existing receivable data, given by the Directorate General of Taxation before PBB-P2 becomes local taxes. The data is between the years of 1994 and 2012. Data is grouped based on the village area and the category of the receivables number. We cluster it into three types, namely the high, medium, and low accounts receivable clusters. Data mining is expected to improve the PBB-P2 receivable data management in the Bantul Regency to make better decision-making. This study’s results make it easier to analyze PBB-P2 receivable data pattern based on the grouping of village areas and the receivable number category. The analysis results are expected to provide input for BKAD Bantul to identify certain villages and categories of receivables that need more attention in the PBB-P2 collecting process in the Bantul Regency.