Utilization of Data Mining Classification Technique for Civil Servant Mutation Pattern: A Case Study of Pangkajene and Kepulauan District Government

Abstract

Pangkajene and Kepulauan (Pangkep) District is an area located in South Sulawesi Province, Indonesia. Regional Civil Servants, Education, and Training (BKPPD) responsible for managing the civil servants (PNS) of Pangkep District. BKPPD provides mutation services to civil servants ranging from recruitment, placement, transfer, education and training, discipline, dismissal, and retirement. Currently, BKPPD has difficulty in conducting mutations, determining which civil servants should be transferred because the absence of a reference mutation pattern. This study aims to obtain mutation patterns using data mining based on historical data on the employment service application system (SAPK). We use three classification algorithms, which are Decision Tree, Naïve Bayes, and Support Vector Machine (SVM) for revealing the mutation pattern in the mutation history data. We find that the decision tree yields the highest accuracy compared to Naive Bayes and SVM with a value of 72.76%. This research also recommends that the mutation pattern may be implemented by BKPPD to design the civil servants redistribution planning of Pangkep District Government.