The welfare classification of Indonesian National Civil servant using TOPSIS and k-Nearest Neighbour (KNN)

Public services to citizen should be improved and developed to achieve better government and satisfying citizen as its main beneficiary. The quality of public services is mainly depended on government employee or civil servant as actor who is conducted all process of services and is supported by many factors. One of significant factors which influences to employee to give better public services is ensuring welfare of civil servant. To formulate policy or regulation to support welfare of civil servant, stakeholders needed the classification of welfare status of civil servant. This research attempted to define welfare criteria and classify civil servant data based on welfare measurement by utilizing Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and k-Nearest Neighbours algorithm (k-NN). This research used qualitative approach and quantitative approach in three governmental organizations, i.e. National Civil Service Agency (BKN; Badan Kepegawaian Negara), Ministry for Administrative and Bureaucracy Reform (KemenPAN&RB; Kementrian Pendayagunaan Aparatur Negara dan Reformasi Birokrasi) and National Institute for Administration (LAN; Lembaga Administrasi Negara). As the result, fifteen welfare criteria of civil servant is identified, i.e. current functional position, marital status, current structural position, health allowance, functional allowance, location of office, credit score, job performance, position allowance, appreciation, structural allowance, current age, number of children, year of work, and month of work. This is also successfully classified welfare status of civil servant by using k-NN with 80.87% accuracy prediction.