relation: http://repository.upi.edu/146227/
title: IMPLEMENTASI ALGORITMA TEXTCNN PADA SISTEM DETEKSI SQL INJECTION BERBASIS APLIKASI WEB
creator: Aulia Putri Cendekia, -
creator: Muhammad Taufik Dwi Putra, -
creator: Devi Aprianti Rimadhani Agustini, -
subject: L Education (General)
subject: QA75 Electronic computers. Computer science
subject: QA76 Computer software
description: Serangan SQL Injection (SQLi) menduduki peringkat teratas OWASP Top 10 pada tahun 2025, serangan ini dapat mengancam keamanan aplikasi web dengan potensi kebocoran jutaan data sensitif. Penelitian ini mengembangkan sistem deteksi SQLi berbasis web menggunakan algoritma Text Convolutional Neural Network (TextCNN) yang mampu menganalisis pola berbahaya dalam kueri SQL secara real-time. Penelitian ini menggunakan metode D&D dalam pengembangan model hingga model TextCNN dapat dilatih dengan dataset berlabel kueri normal dan serangan SQLi dengan proses tokenisasi dan embedding hingga lapisan konvolusional. Pengujian dilakukan dengan mengintegrasi model kedalam aplikasi web untuk dapat melihat model mendeteksi serangan dengan respons real-time. Evaluasi metrik dilakukan agar dapat menunjukkan keberhasilan model dalam mendeteksi, sedangkan blackbox digunakan untuk mengevaluasi aplikasi web  sistem deteksi SQL Injection. Hasil evaluasi menunjukkan akurasi rata-rata 98%, dengan precision, recall, dan F1-score tinggi. Sistem efektif mendeteksi first-order SQLi (boolean-based, time-based, union-based) dengan FP/FN rendah. Selain itu, model berhasil diintegrasikan secara real-time kedalam aplikasi web.  ----------  SQL Injection (SQLi) attacks ranked at the top of the OWASP Top 10 in 2025. These attacks pose a serious threat to web application security, with the potential to expose millions of sensitive data records. This study develops a web-based SQLi detection system using the Text Convolutional Neural Network (TextCNN) algorithm, which is capable of analyzing malicious patterns in SQL queries in real time. The research adopts the Design and Development (D&D) methodology to develop the model, enabling the TextCNN to be trained on labeled datasets consisting of normal queries and SQLi attacks through tokenization, embedding, and convolutional layers. Testing is conducted by integrating the model into a web application to evaluate its real-time detection capabilities. Evaluation metrics are used to demonstrate the model’s detection performance, while black-box testing is applied to assess the SQL Injection detection system within the web application. The evaluation results show an average accuracy of 98%, with high precision, recall, and F1-score values. The system effectively detects first-order SQLi attacks (boolean-based, time-based, and union-based) with low false positive and false negative rates. Furthermore, the model is successfully integrated into a web  application for real-time operation.
date: 2025-12-15
type: Thesis
type: NonPeerReviewed
format: text
language: id
identifier: http://repository.upi.edu/146227/7/S_TEKOM_2006303_TITLE.pdf
format: text
language: id
identifier: http://repository.upi.edu/146227/1/S_TEKOM_2006303_CHAPTER1.pdf
format: text
language: id
identifier: http://repository.upi.edu/146227/2/S_TEKOM_2006303_CHAPTER2.pdf
format: text
language: id
identifier: http://repository.upi.edu/146227/3/S_TEKOM_2006303_CHAPTER3.pdf
format: text
language: id
identifier: http://repository.upi.edu/146227/4/S_TEKOM_2006303_CHAPTER4.pdf
format: text
language: id
identifier: http://repository.upi.edu/146227/6/S_TEKOM_2006303_CHAPTER5.pdf
format: text
language: id
identifier: http://repository.upi.edu/146227/5/S_TEKOM_2006303_Appendix.pdf
identifier:   Aulia Putri Cendekia, - and Muhammad Taufik Dwi Putra, - and Devi Aprianti Rimadhani Agustini, -  (2025) IMPLEMENTASI ALGORITMA TEXTCNN PADA SISTEM DETEKSI SQL INJECTION BERBASIS APLIKASI WEB.  S1 thesis, Universitas Pendidikan Indonesia.   
relation: https://repository.upi.edu