Muhammad Nur Faiz
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Muhammad Nur Faiz has a total of 47 co-authors. Here are the top co-authors by number of shared publications:
- Linda Perdana Wanti
Politeknik Negeri Cilacap
4 shared publications Abdul Rohman Supriyono
2 shared publicationsAnton Yudhana
Universitas Ahmad Dahlan
2 shared publicationsArif Wirawan Muhammad
2 shared publications- Arif Wirawan Muhammad
IT Telkom Purwokerto
2 shared publications Laura Sari
2 shared publications
Latest Publications
- Open Access
- 30 June 2026
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 8 (1) , 246-252
A Reproducible Explainable NLP Workflow for Workplace Sexism Detection: Classification Performance, Rationale Faithfulness, and Sanity Checks
Workplace sexism often appears as indirect, deniable language (e.g., patronizing compliments, competence-doubting questions), making automated detection and organizational response difficult. This study evaluates a transparent, explanation-ready NLP pipeline on the Sexist Workplace Statements (SWS) dataset (1,137 items) with its binary labels: certain sexism vs. ambiguous/neutral. Using the provided fixed stratified split (1,023 train; 114 test), we train a TF–IDF (word 1–2, character 3–5 n-grams) logistic regression baseline and report performance stability across five random seeds. To audit model evidence, sparse token rationales are extracted from linear feature contributions and quantify faithfulness with ERASER-style comprehensiveness (logit drop when rationales are removed) and sufficiency (logit change when only rationales are kept), benchmarked against random-token rationales. The baseline achieves 0.768 ± 0.006 accuracy and 0.759 ± 0.007 macro-F1, with errors concentrated in the ambiguous/neutral class. Faithfulness tests show that model-selected rationales substantially affect the sexism logit (comprehensiveness 1.335 ± 0.001), while remaining insufficient in isolation (|sufficiency| 1.075 ± 0.006). Sanity checks reveal modest sensitivity to gender-term swaps and reduced rationale overlap underweight randomization. Overall, results motivate cautious deployment: explanation-driven auditing can surface shortcut risks and clarify where binary labels blur neutral language and deniable sexism, pointing to future work on finer-grained annotation and human rationale collection.
- 30 June 2026
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 8 (1) , 246-252
Open AccessA Reproducible Explainable NLP Workflow for Workplace Sexism Detection: Classification Performance, Rationale Faithfulness, and Sanity Checks
Workplace sexism often appears as indirect, deniable language (e.g., patronizing compliments, competence-doubting questions), making automated detection and organizational response difficult. This study evaluates a transparent, explanation-ready NLP pipeline on the Sexist Workplace Statements (SWS) dataset (1,137 items) with its binary labels: certain sexism vs. ambiguous/neutral. Using the provided fixed stratified split (1,023 train; 114 test), we train a TF–IDF (word 1–2, character 3–5 n-grams) logistic regression baseline and report performance stability across five random seeds. To audit model evidence, sparse token rationales are extracted from linear feature contributions and quantify faithfulness with ERASER-style comprehensiveness (logit drop when rationales are removed) and sufficiency (logit change when only rationales are kept), benchmarked against random-token rationales. The baseline achieves 0.768 ± 0.006 accuracy and 0.759 ± 0.007 macro-F1, with errors concentrated in the ambiguous/neutral class. Faithfulness tests show that model-selected rationales substantially affect the sexism logit (comprehensiveness 1.335 ± 0.001), while remaining insufficient in isolation (|sufficiency| 1.075 ± 0.006). Sanity checks reveal modest sensitivity to gender-term swaps and reduced rationale overlap underweight randomization. Overall, results motivate cautious deployment: explanation-driven auditing can surface shortcut risks and clarify where binary labels blur neutral language and deniable sexism, pointing to future work on finer-grained annotation and human rationale collection.
- Open Access
- 30 June 2026
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 8 (1) , 216-224
Optimized Skill Mastery Prediction for Adaptive Test Decision-Making
Adaptive testing systems require accurate and timely estimation of students' skill mastery levels to support decision-making in question selection and assessment flow determination. However, uncertainty in students' knowledge levels and variations in learning behavior often limit the performance of conventional predictive models. This study proposes an optimized predictive modeling framework for skill mastery to enhance decision-making quality in adaptive testing. The proposed framework integrates a machine learning-based skill mastery prediction model with an optimization mechanism to improve model accuracy and stability, while accommodating the sequential nature and uncertainty of student responses. Learning interaction data is used to dynamically model the development of skill mastery levels, which are then utilized as decision-support input in the adaptive testing system. The proposed predictive skill mastery model shows strong and consistent performance with an AUC value of 0.822, Average Precision of 0.868, accuracy of 0.757, and a precision balance of 0.834, recall of 0.788, and F1-score of 0.810, supported by well-calibrated probabilities and the ability to respond adaptively to student learning dynamics, making it suitable for use to support decision-making in adaptive test decision-making systems. The results of this study confirm the potential of integrating predictive analytics and optimization techniques in developing intelligent adaptive assessment systems.
- 30 June 2026
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 8 (1) , 216-224
Open AccessOptimized Skill Mastery Prediction for Adaptive Test Decision-Making
Adaptive testing systems require accurate and timely estimation of students' skill mastery levels to support decision-making in question selection and assessment flow determination. However, uncertainty in students' knowledge levels and variations in learning behavior often limit the performance of conventional predictive models. This study proposes an optimized predictive modeling framework for skill mastery to enhance decision-making quality in adaptive testing. The proposed framework integrates a machine learning-based skill mastery prediction model with an optimization mechanism to improve model accuracy and stability, while accommodating the sequential nature and uncertainty of student responses. Learning interaction data is used to dynamically model the development of skill mastery levels, which are then utilized as decision-support input in the adaptive testing system. The proposed predictive skill mastery model shows strong and consistent performance with an AUC value of 0.822, Average Precision of 0.868, accuracy of 0.757, and a precision balance of 0.834, recall of 0.788, and F1-score of 0.810, supported by well-calibrated probabilities and the ability to respond adaptively to student learning dynamics, making it suitable for use to support decision-making in adaptive test decision-making systems. The results of this study confirm the potential of integrating predictive analytics and optimization techniques in developing intelligent adaptive assessment systems.
- Open Access
- 24 May 2026
- Published by Society of Visual Informatics in JOIV : International Journal on Informatics Visualization
- Vol. 10 (3) , 1323
Enhancing DDoS Detection Using SMOTE with PCA - Random Forest Feature Analysis
Distributed Denial of Service (DDoS) attacks are one of the biggest threats in network security that can cause significant disruption to online services. Early and accurate detection of DDoS attacks is very important to maintain service availability. In this study, we propose an approach to optimize DDoS attack detection by combining the Random Forest algorithm and the Synthetic Minority Over-sampling Technique (SMOTE) using the CIC-DDoS2019 dataset. The CIC-DDoS2019 dataset has imbalanced data, namely 50,006,249 DDoS attack data and 56,863 benign/normal. The Synthetic Minority Oversampling Technique (SMOTE) approach is used to address imbalanced data by generating synthetic samples from the minority class (DDoS attacks) to make the data more proportionally balanced, thereby improving the performance of the detection model. Principal Component Analysis (PCA) aims to identify underlying patterns and structures in high-dimensional data by transforming it into a lower-dimensional space while retaining the most important information for feature selection. The Random Forest Algorithm is chosen for its strong ability to handle complex data and to provide feature importance estimates. The combination of these techniques is expected to improve Accuracy. The experimental results in this study compare the number of features with and without SMOTE. The experimental results show that the accuracy without PCA and SMOTE is 98.98%, whereas using PCA with 10 selected features and SMOTE yields a detection accuracy of 99.84%.
- 24 May 2026
- Published by Society of Visual Informatics in JOIV : International Journal on Informatics Visualization
- Vol. 10 (3) , 1323
Open AccessEnhancing DDoS Detection Using SMOTE with PCA - Random Forest Feature Analysis
Distributed Denial of Service (DDoS) attacks are one of the biggest threats in network security that can cause significant disruption to online services. Early and accurate detection of DDoS attacks is very important to maintain service availability. In this study, we propose an approach to optimize DDoS attack detection by combining the Random Forest algorithm and the Synthetic Minority Over-sampling Technique (SMOTE) using the CIC-DDoS2019 dataset. The CIC-DDoS2019 dataset has imbalanced data, namely 50,006,249 DDoS attack data and 56,863 benign/normal. The Synthetic Minority Oversampling Technique (SMOTE) approach is used to address imbalanced data by generating synthetic samples from the minority class (DDoS attacks) to make the data more proportionally balanced, thereby improving the performance of the detection model. Principal Component Analysis (PCA) aims to identify underlying patterns and structures in high-dimensional data by transforming it into a lower-dimensional space while retaining the most important information for feature selection. The Random Forest Algorithm is chosen for its strong ability to handle complex data and to provide feature importance estimates. The combination of these techniques is expected to improve Accuracy. The experimental results in this study compare the number of features with and without SMOTE. The experimental results show that the accuracy without PCA and SMOTE is 98.98%, whereas using PCA with 10 selected features and SMOTE yields a detection accuracy of 99.84%.
- Open Access
- 30 December 2025
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 7 (2) , 395-405
Development of a Hybrid CNN–SVM-Based Acute Lymphoblastic Leukemia Detection System on Hematology Image Data
Acute Lymphoblastic Leukemia (ALL) is among the most common pediatric blood cancers and progresses rapidly, necessitating early and accurate detection. Manual diagnosis via microscopic analysis of blood samples is time-consuming and highly dependent on specialist expertise. This study proposes a hybrid model that combines a Convolutional Neural Network (CNN) with a Support Vector Machine (SVM) to automatically detect ALL from blood-cell images. The CNN performs deep feature extraction from images, while the SVM serves as the classifier to determine ALL status. The dataset comprises microscopic images labeled as ALL or normal and is processed through preprocessing steps such as augmentation and normalization. The adopted CNN produces optimized feature representations. Experimental results show that the hybrid CNN–SVM model with an RBF kernel achieves the best performance, with an accuracy of 96.4%, precision of 95.8%, recall of 96.1%, and an F1-score of 96.0%, surpassing pure CNN-based baselines. Training converged at the 41st epoch, with a training accuracy of 97.2%, validation accuracy of 95.9%, training loss of 0.09, and validation loss of 0.11, indicating stable learning without overfitting. The model’s ROC curve lies well above the chance diagonal, with an Area Under the Curve (AUC) of 0.914, means there is a 91.4% chance the model assigns a higher score to a truly positive (leukemia) image than to a negative (normal) image.These findings suggest that the CNN–SVM hybrid approach enhances leukemia detection performance compared with conventional CNN-only methods and holds promise as a fast, accurate, and efficient image-based decision-support tool for early leukemia diagnosis in digital hematology.
- 30 December 2025
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 7 (2) , 395-405
Open AccessDevelopment of a Hybrid CNN–SVM-Based Acute Lymphoblastic Leukemia Detection System on Hematology Image Data
Acute Lymphoblastic Leukemia (ALL) is among the most common pediatric blood cancers and progresses rapidly, necessitating early and accurate detection. Manual diagnosis via microscopic analysis of blood samples is time-consuming and highly dependent on specialist expertise. This study proposes a hybrid model that combines a Convolutional Neural Network (CNN) with a Support Vector Machine (SVM) to automatically detect ALL from blood-cell images. The CNN performs deep feature extraction from images, while the SVM serves as the classifier to determine ALL status. The dataset comprises microscopic images labeled as ALL or normal and is processed through preprocessing steps such as augmentation and normalization. The adopted CNN produces optimized feature representations. Experimental results show that the hybrid CNN–SVM model with an RBF kernel achieves the best performance, with an accuracy of 96.4%, precision of 95.8%, recall of 96.1%, and an F1-score of 96.0%, surpassing pure CNN-based baselines. Training converged at the 41st epoch, with a training accuracy of 97.2%, validation accuracy of 95.9%, training loss of 0.09, and validation loss of 0.11, indicating stable learning without overfitting. The model’s ROC curve lies well above the chance diagonal, with an Area Under the Curve (AUC) of 0.914, means there is a 91.4% chance the model assigns a higher score to a truly positive (leukemia) image than to a negative (normal) image.These findings suggest that the CNN–SVM hybrid approach enhances leukemia detection performance compared with conventional CNN-only methods and holds promise as a fast, accurate, and efficient image-based decision-support tool for early leukemia diagnosis in digital hematology.
- Open Access
- 30 June 2025
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 7 (1) , 105-120
Sentiment Analysis Using Stacking Ensemble After the 2024 Indonesian Election Results
Sentiment analysis is a text processing technique aimed at identifying opinions and emotions within a sentence. Machine learning is commonly applied in this area, with algorithms such as Naïve Bayes, Support Vector Machine (SVM), and Random Forest being frequently used. However, achieving optimal accuracy remains a challenge, particularly when dealing with unstructured text data, such as content from social media platforms. This research seeks to improve sentiment analysis performance by implementing a stacking ensemble learning approach, which combines the predictive strengths of several base models. The base models selected for this study are Naïve Bayes, SVM, and Random Forest, while Random Forest also serves as the meta-model to generate final predictions. The study focuses on sentiment analysis in a specific context—public opinion following the announcement of the Indonesian presidential election results in 2024. The dataset comprises 6,737 tweets collected from the X platform using web scraping techniques in 2024. Results show that individual models achieved varying levels of accuracy: Naïve Bayes at 66.84%, SVM at 77.74%, and Random Forest at 74.78%. In contrast, the stacking ensemble model achieved a significantly higher accuracy of 81.53%. This improvement highlights the effectiveness of ensemble learning in integrating different algorithmic perspectives to enhance predictive performance. By leveraging the complementary strengths of each base model, stacking not only boosts accuracy but also increases model robustness, making it highly suitable for real-world sentiment analysis applications that involve noisy and informal textual data from social media.
- 30 June 2025
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 7 (1) , 105-120
Open AccessSentiment Analysis Using Stacking Ensemble After the 2024 Indonesian Election Results
Sentiment analysis is a text processing technique aimed at identifying opinions and emotions within a sentence. Machine learning is commonly applied in this area, with algorithms such as Naïve Bayes, Support Vector Machine (SVM), and Random Forest being frequently used. However, achieving optimal accuracy remains a challenge, particularly when dealing with unstructured text data, such as content from social media platforms. This research seeks to improve sentiment analysis performance by implementing a stacking ensemble learning approach, which combines the predictive strengths of several base models. The base models selected for this study are Naïve Bayes, SVM, and Random Forest, while Random Forest also serves as the meta-model to generate final predictions. The study focuses on sentiment analysis in a specific context—public opinion following the announcement of the Indonesian presidential election results in 2024. The dataset comprises 6,737 tweets collected from the X platform using web scraping techniques in 2024. Results show that individual models achieved varying levels of accuracy: Naïve Bayes at 66.84%, SVM at 77.74%, and Random Forest at 74.78%. In contrast, the stacking ensemble model achieved a significantly higher accuracy of 81.53%. This improvement highlights the effectiveness of ensemble learning in integrating different algorithmic perspectives to enhance predictive performance. By leveraging the complementary strengths of each base model, stacking not only boosts accuracy but also increases model robustness, making it highly suitable for real-world sentiment analysis applications that involve noisy and informal textual data from social media.
- Open Access
- 25 May 2025
- Published by LPPM Institut Teknologi Telkom Purwokerto in Journal of Informatics Information System Software Engineering and Applications (INISTA)
- Vol. 7 (2) , 173-182
Classification of DDoS Attacks based on Network Traffic Patterns Using the k-Nearest Neighbor (k-NN) Algorithm
Many server attacks disrupt industrial or business operations. Attacks that flood bandwidth with simultaneous requests can overwhelm a system, leading to significant downtime and financial losses. Additionally, breaches that compromise sensitive data can damage a company's reputation and erode customer trust. DDoS attacks, or Distributed Denial of Service attacks, are among the most common types of server attacks. DDoS has been proven to cause server downtime, and one effective way to mitigate this attack is to detect and classify it using a machine learning approach. The K-Nearest Neighbor (KNN) algorithm, a simple yet effective classification method based on similarity measures, is known for its high accuracy. The current research builds upon two stages: the feature extraction stage and the classification stage, with the ultimate goal of improving the accuracy of DDoS identification using the CICDDoS2019 dataset. Based on this premise, the detection accuracy can be improved by enhancing these two stages. At a value of k equal to 3, this study produces an accuracy of 99.73%.
- 25 May 2025
- Published by LPPM Institut Teknologi Telkom Purwokerto in Journal of Informatics Information System Software Engineering and Applications (INISTA)
- Vol. 7 (2) , 173-182
Open AccessClassification of DDoS Attacks based on Network Traffic Patterns Using the k-Nearest Neighbor (k-NN) Algorithm
Many server attacks disrupt industrial or business operations. Attacks that flood bandwidth with simultaneous requests can overwhelm a system, leading to significant downtime and financial losses. Additionally, breaches that compromise sensitive data can damage a company's reputation and erode customer trust. DDoS attacks, or Distributed Denial of Service attacks, are among the most common types of server attacks. DDoS has been proven to cause server downtime, and one effective way to mitigate this attack is to detect and classify it using a machine learning approach. The K-Nearest Neighbor (KNN) algorithm, a simple yet effective classification method based on similarity measures, is known for its high accuracy. The current research builds upon two stages: the feature extraction stage and the classification stage, with the ultimate goal of improving the accuracy of DDoS identification using the CICDDoS2019 dataset. Based on this premise, the detection accuracy can be improved by enhancing these two stages. At a value of k equal to 3, this study produces an accuracy of 99.73%.
- Open Access
- 1 March 2025
- Published by Universitas Sahid Surakarta in JURNAL PENGABDIAN TEKNOLOGI TEPAT GUNA
- Vol. 6 (1) , 110-124
Pengembangan Desa Digital Dengan Penerapan Sistem Informasi Kearsipan Untuk Meningkatkan Pelayanan Publik Pada Pemerintah Desa Banjarwaru Kecamatan Nusawungu Kabupaten Cilacap
Pelayanan Publik yang efektif dan efisien menjadi salah satu tolak ukur masyarakat dalam menilai kinerja penemrintah daerah secara kasat mata. Pelayanan publik seperti pembuatan surat dan pencarian data arsip sangat berpengaruh terhadap kualitas layanan pada Kantor Desa Banjarwaru. Proses pengarsipan surat dan administrasi lainnya di Desa Banjarwaru saat ini belum terdigitalisasi dan terdokumensi dengan baik. Hal tersebut sering membuat terjadinya kesalahan dalam penyimpanan data surat masuk dan membutuhkan ruang yang cukup luas untuk menyimpannya, selain itu juga seringnya terjadi kesalahan dalam penulisan nomor surat keluar karena pencarian nomor surat terakhir masih dicari manual pada buku besar. Guna untuk mewujudkan visi dan misi Desa Banjarwaru, maka sangat diperlukan Penerapan Sistem Informasi Kearsipan untuk pengembangan Desa Digital. Sistem informasi arsip persuratan mempunyai tujuan mengubah metode penyimpanan surat atau administrasi lainnya dengan proses pengarsapan secara digital sehingga mengurangi penggunaan kertas dan lebih efektif dan akurat.
- 1 March 2025
- Published by Universitas Sahid Surakarta in JURNAL PENGABDIAN TEKNOLOGI TEPAT GUNA
- Vol. 6 (1) , 110-124
Open AccessPengembangan Desa Digital Dengan Penerapan Sistem Informasi Kearsipan Untuk Meningkatkan Pelayanan Publik Pada Pemerintah Desa Banjarwaru Kecamatan Nusawungu Kabupaten Cilacap
Pelayanan Publik yang efektif dan efisien menjadi salah satu tolak ukur masyarakat dalam menilai kinerja penemrintah daerah secara kasat mata. Pelayanan publik seperti pembuatan surat dan pencarian data arsip sangat berpengaruh terhadap kualitas layanan pada Kantor Desa Banjarwaru. Proses pengarsipan surat dan administrasi lainnya di Desa Banjarwaru saat ini belum terdigitalisasi dan terdokumensi dengan baik. Hal tersebut sering membuat terjadinya kesalahan dalam penyimpanan data surat masuk dan membutuhkan ruang yang cukup luas untuk menyimpannya, selain itu juga seringnya terjadi kesalahan dalam penulisan nomor surat keluar karena pencarian nomor surat terakhir masih dicari manual pada buku besar. Guna untuk mewujudkan visi dan misi Desa Banjarwaru, maka sangat diperlukan Penerapan Sistem Informasi Kearsipan untuk pengembangan Desa Digital. Sistem informasi arsip persuratan mempunyai tujuan mengubah metode penyimpanan surat atau administrasi lainnya dengan proses pengarsapan secara digital sehingga mengurangi penggunaan kertas dan lebih efektif dan akurat.
- Open Access
- 30 January 2025
- Published by Politeknik Negeri Cilacap in Infotekmesin
- Vol. 16 (1) , 153-159
Perbandingan Pendekatan Machine Learning dalam Deteksi Serangan DDoS Jaringan Komputer
Distributed Denial of Service (DDoS) attacks are a serious threat to computer network security. This study offers a comprehensive evaluation by considering accuracy, detection time, and model complexity in simulation scenarios. Using the CICDDoS2019 dataset, which includes modern attack variations and complete features, this research compares the effectiveness of Naïve Bayes (NB), Random Forest (RF), and Decision Tree (DT) algorithms in detecting DDoS attacks. The results show that RF achieves the highest accuracy (99.95%), while DT excels in recall (99.83%). These findings provide a foundation for developing hybrid ML-DL models to enhance real-time attack detection. However, limitations such as using a single dataset and offline simulations restrict the generalizability of results to real-world network conditions. This study highlights opportunities for more comprehensive future research in real-world scenarios.
- 30 January 2025
- Published by Politeknik Negeri Cilacap in Infotekmesin
- Vol. 16 (1) , 153-159
Open AccessPerbandingan Pendekatan Machine Learning dalam Deteksi Serangan DDoS Jaringan Komputer
Distributed Denial of Service (DDoS) attacks are a serious threat to computer network security. This study offers a comprehensive evaluation by considering accuracy, detection time, and model complexity in simulation scenarios. Using the CICDDoS2019 dataset, which includes modern attack variations and complete features, this research compares the effectiveness of Naïve Bayes (NB), Random Forest (RF), and Decision Tree (DT) algorithms in detecting DDoS attacks. The results show that RF achieves the highest accuracy (99.95%), while DT excels in recall (99.83%). These findings provide a foundation for developing hybrid ML-DL models to enhance real-time attack detection. However, limitations such as using a single dataset and offline simulations restrict the generalizability of results to real-world network conditions. This study highlights opportunities for more comprehensive future research in real-world scenarios.
- Open Access
- 30 December 2024
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 6 (2) , 173-180
Website Security Analysis Using Vulnerability Assessment Method
In today’s digital era, ensuring website security is crucial, especially in the education sector which is frequently targeted by cyber attacks. This research aims to test security of the Universitas Internasional Batam (UIB) website using OWASP ZAP and Nessus. The method will be used in this research was vulnerability assessment. It will involve gathering information with the tools such as, Nmap, whois and nslookup. OWASP ZAP detected 11 vulnerabilities, categorized into 6 medium level and 5 low level, including Content Security Policies (CSP) and anti-clickjacking headers. Otherwise, Nessus only detected one medium level vulnerability, the absence of HTTP Strict Transport Security (HSTS). The difference in detection results from the tools that OWASP ZAP is better at finding web application weakness that are consistent with the OWASP Top Ten 2021, while Nessus specifically targets server and network configuration. For educational institutions, these results emphasize the importance of conducting regular vulnerability assessment to protect sensitive data. Recommended action include implementing CSP to prevent Cross-site scripting (XSS) and other injection attacks, enforcing HSTS to secure communication, and its recommend to updating software to mitigate the unknown vulnerabilities. By adopting these measures, institutions can reduce their exposure to cyber attacks, its also can maintain user trust, and strengthen overall security. This research provides a pratical framework for stregthening the security of educational websites against evolving threats. These findings highlight that the importance of using multiple tools can provide a more comprehensive view of security gaps.
- 30 December 2024
- Published by Politeknik Negeri Cilacap in Journal of Innovation Information Technology and Application (JINITA)
- Vol. 6 (2) , 173-180
Open AccessWebsite Security Analysis Using Vulnerability Assessment Method
In today’s digital era, ensuring website security is crucial, especially in the education sector which is frequently targeted by cyber attacks. This research aims to test security of the Universitas Internasional Batam (UIB) website using OWASP ZAP and Nessus. The method will be used in this research was vulnerability assessment. It will involve gathering information with the tools such as, Nmap, whois and nslookup. OWASP ZAP detected 11 vulnerabilities, categorized into 6 medium level and 5 low level, including Content Security Policies (CSP) and anti-clickjacking headers. Otherwise, Nessus only detected one medium level vulnerability, the absence of HTTP Strict Transport Security (HSTS). The difference in detection results from the tools that OWASP ZAP is better at finding web application weakness that are consistent with the OWASP Top Ten 2021, while Nessus specifically targets server and network configuration. For educational institutions, these results emphasize the importance of conducting regular vulnerability assessment to protect sensitive data. Recommended action include implementing CSP to prevent Cross-site scripting (XSS) and other injection attacks, enforcing HSTS to secure communication, and its recommend to updating software to mitigate the unknown vulnerabilities. By adopting these measures, institutions can reduce their exposure to cyber attacks, its also can maintain user trust, and strengthen overall security. This research provides a pratical framework for stregthening the security of educational websites against evolving threats. These findings highlight that the importance of using multiple tools can provide a more comprehensive view of security gaps.
- Open Access
- 1 April 2024
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 34 (1) , 224-232
Clustering method for criminal crime acts using K-means and principal component analysis
Criminality is an act of violating the values and norms of society that causes a lot of harm. Much of the criminal data is often just a collection of data that has no information. Analysis of crime data is key in efforts to reduce crime rates that provide an overview of the incidence of crime, patterns, levels of vulnerability, and the level of security of an area. This research proposes data analysis that provides an understanding of crime using data mining techniques, especially the K-means cluster method, both traditional and with principal component analysis (PCA) dimension reduction. Before the PCA process, the values are transformed first with Z score normalization. From the processing through the davies bouldin index (DBI) performance test with 3 clusters, it is concluded that traditional K-means produces a DBI Index value of 0.019 and K-means PCA of 0.299. Meanwhile, to see the optimal cluster, several iterations were performed and resulted in the most optimal DBI index of 4 clusters in K-means of 0.014 and K-means PCA of 0.172. From the performance test value, it means that in the context of clustering the traditional criminal K-means data is declared more optimal than K-means PCA.
- 1 April 2024
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 34 (1) , 224-232
Open AccessClustering method for criminal crime acts using K-means and principal component analysis
Criminality is an act of violating the values and norms of society that causes a lot of harm. Much of the criminal data is often just a collection of data that has no information. Analysis of crime data is key in efforts to reduce crime rates that provide an overview of the incidence of crime, patterns, levels of vulnerability, and the level of security of an area. This research proposes data analysis that provides an understanding of crime using data mining techniques, especially the K-means cluster method, both traditional and with principal component analysis (PCA) dimension reduction. Before the PCA process, the values are transformed first with Z score normalization. From the processing through the davies bouldin index (DBI) performance test with 3 clusters, it is concluded that traditional K-means produces a DBI Index value of 0.019 and K-means PCA of 0.299. Meanwhile, to see the optimal cluster, several iterations were performed and resulted in the most optimal DBI index of 4 clusters in K-means of 0.014 and K-means PCA of 0.172. From the performance test value, it means that in the context of clustering the traditional criminal K-means data is declared more optimal than K-means PCA.