<![CDATA[Advertising tax constitutes a significant component contributing to the Local Own-Source Revenue of Bandar Lampung. A persistent problem identified is the high volume of advertising tax arrears, reflecting non-compliance in payment among tax objects. This research aims to perform segmentation and analysis of advertising tax non-compliance by employing the K-Prototypes Clustering algorithm. The data for this study were sourced from the archives of the Regional Revenue Agency of Bandar Lampung (Badan Pendapatan Daerah) for the billing period of April 2025, encompassing a total of 949 tax objects. The research methodology includes data pre-processing (categorical data transformation and numerical variable normalization using the z-score method), descriptive analysis, application of the K-Prototypes Clustering algorithm, and result evaluation utilizing the Average Silhouette Width (ASW). The clustering results yield three primary clusters, which represent taxpayer categories with varying degrees of non-compliance, analyzed based on the variables: principal arrears, payment delay, interest, tax year, tax month, and Technical Implementation Unit (UPTD). The evaluation using ASW produced a value of 0.422, indicating that the K-Prototypes Clustering algorithm is sufficiently effective in grouping the advertising tax objects.]]>
Copyrights © 2026
