Shafa Aurelliza Arian, Putri Rahel Alifia, Bagus Prihantoro, Muhammad Iqbal Disriansyah, Ken Ditha Tania, Alsella Meiriza, Ahmad Rifai

Perbandingan Model Klasifikasi Supervised Machine Learning dalam Knowledge Discovery Layanan TI Pertamina Prabumulih

Introduction

Perbandingan model klasifikasi supervised machine learning dalam knowledge discovery layanan ti pertamina prabumulih. Bandingkan model klasifikasi Supervised Machine Learning untuk knowledge discovery layanan TI Pertamina Prabumulih. SVM unggul, optimalkan manajemen layanan TI.

Abstract

Pengelolaan data layanan Teknologi Informasi (TI) di Pertamina Prabumulih memerlukan pendekatan analitik untuk meningkatkan efektivitas penanganan dan mendukung pengambilan keputusan berbasis data. Penelitian ini bertujuan membandingkan performa beberapa model klasifikasi supervised machine learning pada layanan TI periode 2020–2025 menggunakan pendekatan knowledge discovery melalui teknik data mining terhadap 8.627 data awal. Tahapan penelitian meliputi preprocessing, pelabelan kelas, penanganan ketidakseimbangan data dengan Synthetic Minority Over-sampling Technique (SMOTE), serta pembagian data dengan skenario 70:30, 80:20, dan 90:10. Proses klasifikasi dilakukan menggunakan algoritma Naïve Bayes, Support Vector Machine (SVM), dan Random Forest. Evaluasi model menggunakan metrik accuracy, precision, recall, dan F1-score. Hasil menunjukkan bahwa skenario 90:10 memberikan performa terbaik, dengan Support Vector Machine (SVM) mencapai accuracy 0,8806, precision 0,8757, recall 0,8941, dan F1-score 0,8788, melampaui algoritma lainnya. Kategori desktop hardware teridentifikasi sebagai kasus terbanyak selama periode penelitian. Temuan ini dapat dimanfaatkan sebagai dasar strategis untuk prioritas penanganan layanan, alokasi sumber daya, serta penguatan Knowledge Management guna peningkatan layanan TI secara terarah, efektif, dan berkelanjutan.

Review

The study "Perbandingan Model Klasifikasi Supervised Machine Learning dalam Knowledge Discovery Layanan TI Pertamina Prabumulih" addresses a critical need for data-driven insights in IT service management, particularly within a large organization like Pertamina Prabumulih. By focusing on an analytical approach to manage IT service data, the research aims to enhance handling effectiveness and support informed decision-making. The scope involves a comprehensive comparison of several supervised machine learning classification models applied to 8,627 IT service data records spanning from 2020 to 2025, framed within a knowledge discovery paradigm using data mining techniques. The methodology employed is robust and well-articulated in the abstract. It encompasses essential data preprocessing, class labeling, and crucially, addresses data imbalance using the Synthetic Minority Over-sampling Technique (SMOTE). The researchers systematically evaluated model performance across various data split scenarios (70:30, 80:20, and 90:10) using Naïve Bayes, Support Vector Machine (SVM), and Random Forest algorithms. Model evaluation was thorough, utilizing standard metrics: accuracy, precision, recall, and F1-score. A significant finding is that the 90:10 data split scenario consistently yielded the best performance, with Support Vector Machine (SVM) emerging as the superior algorithm, achieving an accuracy of 0.8806, precision of 0.8757, recall of 0.8941, and an F1-score of 0.8788. Furthermore, the analysis successfully identified "desktop hardware" as the most frequent service category issue during the investigated period. The findings of this research offer substantial practical implications for Pertamina Prabumulih. The identified superior classification model and the insight into prevalent service issues provide a strategic foundation for prioritizing service handling, optimizing resource allocation, and strengthening Knowledge Management practices. This data-driven approach promises to foster more targeted, effective, and sustainable improvements in IT service delivery. The study successfully demonstrates the utility of supervised machine learning in transforming raw IT service data into actionable intelligence, contributing valuable insights for organizational efficiency and strategic planning in the realm of IT operations.

Full Text

Comments

You need to be logged in to post a comment.