Implementasi Support Vector Machine untuk Klasifikasi Laporan Gangguan Layanan Pelanggan Korporat

Syahrul Ramadhan (Universitas Bina Darma, Palembang)
Heri Suroyo (Universitas Bina Darma, Palembang)

TIN: TERAPAN INFORMATIKA NUSANTARA
Vol 6 No 12 (2026): May 2026

This study aims to develop a classification model based on Support Vector Machine (SVM) to identify and categorize service completion levels of corporate customer service disruptions at PT. Telkom Indonesia, South Sumatra Telecommunication Region. Research data was obtained directly from the Helpdesk Assurance CCAN Unit during the period of October 2023 to January 2024, comprising a total of 854 entries. The target variable used is the service completion level, while the feature variables include service type, repair type, and work duration. The entire research process follows the Cross Industry Standard Process for Data Mining (CRISP-DM) framework, covering business understanding, data understanding, data preparation, modeling, evaluation, and deployment stages. Data preparation included feature selection, label encoding, scale normalization, and an 80:20 data split. The model was built using SVM with a linear kernel and active probability configuration. Evaluation results demonstrate an accuracy of 98.25%, a log loss value of 0.0200, and a perfect Area Under Curve value of 1.00 for all classes. The work duration variable proved to be the most influential factor with a correlation value of 0.91 against the target variable. The resulting model has been saved in binary file format, making it ready for practical implementation in real operational environments.