Smart Learning Library

  • Beranda
  • Informasi
  • News
  • Bantuan
  • Pustakawan
  • Area Anggota
  • Pilih Bahasa :
    Arabic Bengali Brazilian Portuguese English Espanol German Indonesia Bahasa Jepang Melayu Persia Russian Thai Turkish Urdu

Search by:

All Author Subject ISBN/ISSN Advanced Search

Last search:

{{tmpObj[k].text}}
No image available for this title

SKRIPSI IF

Explainable AI untuk Cyber Threat Intelligence Scoring Berbasis Aplikasi Web dalam Klasifikasi Malicious domain

SAPUTRA, DICKY - Personal Name;



ABSTRAK

Domain memainkan peran penting dalam sebagian besar serangan siber, dan Security Operations Center (SOC) sering mengalami kesulitan dalam memproses data ancaman dalam jumlah besar yang melibatkan banyak domain. Untuk mengatasi permasalahan tersebut, dikembangkan sistem Cyber Threat Intelligence (CTI) Scoring berbasis web yang terintegrasi dengan machine learning untuk mengklasifikasikan domain berbahaya dan menghasilkan risk score. Namun, model machine learning umumnya bersifat black-box sehingga keputusannya sulit dipahami. Oleh karena itu, diperlukan Explainable Artificial Intelligence (XAI) untuk meningkatkan interpretabilitas model. Model klasifikasi dibangun menggunakan Extreme Gradient Boosting (XGBoost) dengan interpretabilitas melalui perbandingan dua metode XAI, yaitu SHapley Additive exPlanations (SHAP) dan Local Interpretable Model-agnostic Explanations (LIME), untuk menentukan metode yang paling sesuai. Model yang dihasilkan memperoleh accuracy sebesar 0,9316 dan ROC-AUC sebesar 0,9794 pada data pengujian. Untuk kelas malicious, model memperoleh precision sebesar 0,9030, recall sebesar 0,9252, dan F1-score sebesar 0,9140, yang menunjukkan keseimbangan antara kemampuan mendeteksi ancaman dan meminimalkan ancaman yang terlewat meskipun distribusi data tidak seimbang, dengan waktu komputasi 0,0560 detik. SHAP menunjukkan kinerja yang lebih baik dibandingkan LIME dari aspek efisiensi waktu, faithfulness, dan stability, sedangkan pengujian consistency menunjukkan korelasi yang signifikan meskipun tingkat kesepakatan menurun pada kasus False Positive. Model dan metode XAI diimplementasikan ke dalam sistem CTI Scoring berbasis web yang mampu memproses dan menjelaskan hasil klasifikasi domain secara cepat (real-time) berdasarkan input yang diberikan pengguna. Hasil penelitian menunjukkan bahwa model mampu mengklasifikasikan domain berbahaya secara efektif dengan interpretabilitas yang baik, serta mengidentifikasi SHAP sebagai metode XAI dengan performa terbaik.
Kata Kunci: machine learning; explainable artificial intelligence; klasifikasi malicious domain; XGBoost; risk scoring.

ABSTRACT

Domains play a critical role in most cyberattacks, and Security Operations Centers (SOC) often struggle to process large volumes of threat data involving numerous domains. To address this, a web-based Cyber Threat Intelligence (CTI) Scoring system integrated with machine learning was developed to classify malicious domains and generate risk scores. However, machine learning models often remain black-box, making their decisions difficult to understand; thus, Explainable Artificial Intelligence (XAI) is required to improve model interpretability. The classification model was built using Extreme Gradient Boosting (XGBoost), with interpretability provided through a comparison of two XAI methods, SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME), to determine the most suitable method for this task. The resulting model achieved an accuracy of 0.9316 and a ROC-AUC of 0.9794 on the test data. For the malicious class, the model obtained a precision of 0.9030, a recall of 0.9252, and an F1-score of 0.9140, indicating a balance between detection accuracy and missed threats despite an imbalanced data distribution, with a computation time of 0.0560 seconds. SHAP outperformed LIME in time efficiency, faithfulness, and stability, with a consistency test showing significant correlation between the two methods, although agreement decreased for False Positive cases. The model and XAI method were implemented into a web-based CTI Scoring system capable of processing and explaining domain classification results in real time. These results show that the model effectively classifies malicious domains while offering improved interpretability, with SHAP identified as the best-performing XAI method for this task.
Keywords—machine learning; explainable artificial intelligence; malicious domain classification; XGBoost; risk scoring


Ketersediaan
S260908014607.2 SAP ePerpustakaan STMIK AMIKBANDUNGTersedia
Informasi Detil
Judul Seri
-
No. Panggil
607.2 SAP e
Penerbit
: ., 2026
Deskripsi Fisik
-
Bahasa
ISBN/ISSN
-
Klasifikasi
NONE
Tipe Isi
-
Tipe Media
-
Tipe Pembawa
-
Edisi
-
Subyek
SKRIPSI IF
Info Detil Spesifik
-
Pernyataan Tanggungjawab
-
Versi lain/terkait

Tidak tersedia versi lain

Lampiran Berkas
Komentar

You must be logged in to post a comment

Smart Learning Library
  • Information
  • Services
  • Librarian
  • Member Area

About Us

As a complete Library Management System, SLiMS (Senayan Library Management System) has many features that will help libraries and librarians to do their job easily and quickly. Follow this link to show some features provided by SLiMS.

Search

start it by typing one or more keywords for title, author or subject

Keep SLiMS Alive Want to Contribute?

© 2026 — Senayan Developer Community

Powered by SLiMS
Select the topic you are interested in
  • Computer science, information & general works
  • Philosophy & psychology
  • Religion
  • Social sciences
  • Language
  • Pure Science
  • Applied sciences
  • Arts & recreation
  • Literature
  • History & geography
Advanced Search