Segmentation of Public Perceptions of Integrity Zones Using the K-Means Algorithm and Agglomerative Clustering

Abstract

The development of Integrity Zones toward Corruption-Free Zones (WBK) and Clean and
Service-Oriented Bureaucracy Zones (WBBM) is part of bureaucratic reform efforts to
achieve public services that are clean, accountable, and focused on public satisfaction. Public
perception of the implementation of Integrity Zones is a key indicator in evaluating the quality
of services provided by government agencies. However, survey results are generally analyzed
descriptively, making it difficult to identify the characteristics of specific community groups
based on their perception patterns. This study aims to segment public perceptions of Integrity
Zones using the K-Means and Agglomerative Clustering algorithms. The research data were
derived from the 2025 Survey on the Assessment of Integrity Zone Development Toward
WBK/WBBM at the Medan Health Quarantine Center, involving 1,046 respondents. The
survey included the Anti-Corruption Perception Index (IPAK), consisting of five indicators,
and the Service Quality Perception Index (IPKP), consisting of eight indicators. The research
stages included data cleaning, variable transformation and normalization, determining the
optimal number of clusters, the clustering process, and evaluating cluster quality using the
Silhouette Score and dendrogram visualization. The research results are expected to identify
the characteristics of each public perception group so that they can serve as considerations for
improving service quality and strengthening the implementation of the Integrity Zone within
the Medan Health Quarantine Center.