Claim Missing Document
Check
Articles

Found 10 Documents
Search

Sistem Segmentasi Keluhan Air Bersih di PT. Suryacipta Swadaya Menggunakan K-Medoids Clustering Reza Noviandi; Yulison Herry Chrisnanto; Herdi Ashaury
Prosiding SISFOTEK Vol 4 No 1 (2020): Vol 4 No 1 (2020): SISFOTEK 2020
Publisher : Ikatan Ahli Informatika Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (341.728 KB)

Abstract

Suryacipta Swadaya is a company that manages industrial estates. PT. Suryacipta Swadaya manages water in the industrial area to be distributed to 73 companies in the industrial area. Water in the industrial area sometimes has problems that companies complain about. There are 7 types of complaints with an average of complaints from 2017 to 2019, which is 8929 complaints per year. Water complaints for each company vary in the amount of distribution, however the water needs of each company are different even for the same company that has a diversity of water needs each year. Based on this, PT. Suryacipta Swadaya has difficulty mapping the water complaints group based on company needs, so that it is effective in handling water complaints in the industrial area. In this study the grouping of water complaints will be assisted using the K-Medoids method in the process of grouping water complaints so that the handling process can be handled immediately.
Klasifikasi Kerpibadian Berdasarkan Big Five Personality Menggunakan Metode Fuzzy Decision Tree Dengan Algoritma C4.5 Siti Aisah; Fajri Rakhmat Umbara; Herdi Ashaury
Jurnal Teknologi Informatika dan Komputer Vol 8, No 1 (2022): Jurnal Teknologi Informatika dan Komputer
Publisher : Universitas Mohammad Husni Thamrin

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37012/jtik.v8i1.1110

Abstract

Kepribadian merupakan sifat alami seseorang mencangkup tingkah laku, cara berbicara, bertindak dalam menghadapi sesuatu. Pada setiap orang memiliki kepribadian yang berbeda-beda dalam soal kualitas dan kuantitas. Banyak penelitian sudah dilakukan untuk klasifikasi kepribadian yang menggunakan media sosial sebagai sumber informasi. Memahami kepribadian seseorang dapat melalui tweet pada pengguna twitter dengan hasil klasifikasi gabungan antara perilaku sosial dan linguistik. Nilai akurasi penilaian kepribadian dapat ditingkatkan jika menggunakan salah satu dari pembobotan yang mempengaruhi kata-kata dari suatu tweet pengguna. Namun data yang digunakan pada penelitian ini memiliki atribut tipe diskrit dan kontinu sehingga membutuhkan metode yang dapat menangani permasalahan pada klasifikasi kepribadian. Pada penelitian ini klasifikasi kepribadian berdasarkan Big Five Personality sehingga menemukan kepribadian baru seperti Extraverted, Serious, Dependable, Lively dan Responsible. Data yang digunakan akan diektraksi menggunakan metode Fuzzy Decision Tree dengan Algoritma C4.5. Data kepribadian yang digunakan terdiri dari 8 variabel. Model Fuzzy Decision Tree diimplementasikan menggunakan pruning atau pemangkasan, sehingga ada kemungkinan aturan yang dihasilkan memiliki akurasi yang tinggi.
Reverse Engineering Pada Basis Data Relasional Untuk Optimasi Performansi Query Fajri Umbara; Herdi Ashaury; Puspita Nurul Sabrina
Jurnal ICT: Information Communication & Technology Vol. 23 No. 1 (2023): JICT-IKMI, Juli 2023
Publisher : LPPM STMIK IKMI Cirebon

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Pemanfaatan basis data sudah dimulai sejak komputer generasi ke 2 diciptakan, yaitu sekitar tahun 1960. Semakin bertambahnya tahun, penggunaan basis data sudah memiliki perkembangan dan didukung dengan teknologi yang sudah canggih pada masa sekarang. Basis data yang sering digunakan dalam masalah transaksi adalah basis data dengan skema relasional atau sering disebut dengan Relational Database dan diimplementasikan dengan baik oleh aplikasi Relational Database Management System (RDBMS). Permasalahan yang sering dihadapi adalah pengimplementasian basis data relasional yang tidak baik, terutama dari sisi relasi, kardinalitas, dan strukturnya secara umum sehingga apabila dilakukan suatu proses query, maka hasilnya tidak optimal. Hal ini terjadi karena akses dari basis data yang secara paralel dapat diakses oleh banyak pengguna dalam satu waktu. Pada penelitian ini dilakukan evaluasi secara benchmark hasil query basis data relasional sebelum dilakukan desain ulang dan setelah dilakukan desain ulang menggunakan teknik reverse engineering. Hasil yang didapat adalah peningkatan performansi rata-rata sekitar 82.2442% setelah dilakukannya perbaikan struktur desain dan implementasi tabel untuk hasil query yang sama pada aplikasi RDBMS Mysql.
IMPLEMENTATION OF OAUTH 2.0 BASED ON LARAVEL FRAMEWORK IN A CASE STUDY OF CLIENT INFORMATION MANAGEMENT SYSTEM Arthur Oliviana Zabka; Asep Id Hadiana; Herdi Ashaury
Journal of Informatics and Communication Technology (JICT) Vol. 5 No. 2 (2023)
Publisher : PPM Telkom University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52661/j_ict.v5i2.219

Abstract

The swift growth of the internet and its utilization by businesses for operational purposes, such as developing information systems and utilizing cloud-based data storage, has been remarkable. The discussed client data recording system is designed to facilitate the recording of wedding photography bookings, easing the workload for employees. However, the rapid expansion of the internet has also introduced security concerns, particularly regarding unauthorized access due to weak website authorization and authentication. Consequently, ensuring and effectively managing access rights to information systems becomes crucial. This study aims to implement secure website login authorization using the OAuth 2.0 method with Laravel Passport in the client data recording information system. Post-authentication, the authorization in the context of OAuth2, used within Laravel Passport, provides users with access tokens to reach the primary interface. This process involves an API that both furnishes and safeguards the intended resources. Upon authentication and receipt of a valid access token from the OAuth2 system, users can utilize the token to access the API. The research outcomes enhance the security of information system access rights, aiming to reduce unauthorized breaches in websites storing vital data, thus ensuring the safety and protection of stored client data. Testing results using SQL Injection yielded 4418 messages sent and 2209 task IDs, with a current fuzzer count of 0, signifying that the system remained secure and impervious to SQL Injection attacks.
Peningkatan UI/UX Learning Management System Melalui Redesain Berbasis Design Thinking Harfin Aqbil Falah; Fajri Rakhmat Umbara; Herdi Ashaury
Jurnal Algoritma Vol 23 No 1 (2026): Jurnal Algoritma
Publisher : Institut Teknologi Garut

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33364/algoritma/v.23-1.2454

Abstract

The Learning Management System (LMS) at SMKN 4 Padalarang faces various problems in terms of interface and user experience due to the use of default Moodle templates that are not designed according to user needs. An unintuitive interface, confusing navigation, and unattractive visual aesthetics are obstacles to the effectiveness of online learning. This study aims to evaluate the effectiveness of the Design Thinking method in improving the quality of the LMS UI/UX through a user-centered approach. The testing methods used include the System Usability Scale (SUS), User Experience Questionnaire (UEQ), and A/B Testing. Before the redesign, the average SUS score was only 45, which falls into the “Awful” (F) category, and the UEQ results showed an average value below -0.8, which is a negative evaluation. After implementing the Design Thinking method, the SUS score increased to 85, which falls into the “Excellent” (A) category, and the A/B Testing results showed that users preferred the new design, especially in terms of navigation, visuals, and ease of use. These findings indicate that the Design Thinking approach is not only effective in improving usability and functionality, but also capable of increasing user satisfaction and comfort in using the LMS.
Perbandingan IBAC dan ABAC Sebagai Strategi Keamanan Siber Berbasis Zero Trust Architecture (ZTA) Erwin Maulana; Herdi Ashaury; Melina
Jurnal Algoritma Vol 23 No 1 (2026): Jurnal Algoritma
Publisher : Institut Teknologi Garut

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33364/algoritma/v.23-1.3127

Abstract

The rapid development of web technologies has increased exposure to various types of cyberattacks, making traditional perimeter-based security approaches less effective. Zero Trust Architecture (ZTA), with its principle of “never trust, always verify,” offers a more adaptive approach to addressing modern threats. This study aims to analyze and compare the effectiveness of two access control models within the ZTA framework—Identity-Based Access Control (IBAC) and Attribute-Based Access Control (ABAC)—in enhancing web application security. The research employs a quantitative experimental approach by developing two ZTA-based web application prototypes implementing IBAC and ABAC. Three common cyberattacks—SQL Injection, Cross-Site Scripting (XSS), and Brute Force—are tested using Detection Rate as the evaluation parameter. The results show that both models achieve a 100% Detection Rate against all simulated attacks, indicating that they provide an equivalent level of protection within the context of this experiment. Nevertheless, conceptual analysis reveals that ABAC offers advantages in terms of granularity and contextual awareness, as it is capable of dynamically evaluating user attributes, object attributes, and environmental conditions. These characteristics make ABAC more aligned with the principle of continuous verification in Zero Trust, particularly for long-term implementation in modern web applications.
DETEKSI PENIPUAN TRANSAKSI PADA PLATFORM MARKETPLACE MENGGUNAKAN ALGORITMA ISOLATION FOREST Rakha Aryagunawibawa; Puspita Nurul Sabrina; Herdi Ashaury
JOURNAL SAINS STUDENT RESEARCH Vol. 4 No. 5 (2026): Oktober: Jurnal Sains Student Research
Publisher : CV. KAMPUS AKADEMIK PUBLISING

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61722/jssr.v4i5.12743

Abstract

The increasing use of marketplace platforms has raised the risk of transaction fraud. Fraud detection is challenging because fraudulent transactions are far less frequent than legitimate transactions, creating an imbalanced dataset. This study analyzes the capability of Isolation Forest for fraud detection, evaluates the influence of preprocessing and hyperparameter settings, and compares its performance with Local Outlier Factor (LOF). The experiment uses the public Fraudulent E-Commerce Transactions dataset from Kaggle containing 1,048,575 transactions: 996,055 legitimate transactions (94.99%) and 52,520 fraudulent transactions (5.01%). The workflow includes data cleaning, feature selection, feature engineering, One-Hot Encoding, StandardScaler, an 80:20 stratified train-test split, baseline modeling, hyperparameter tuning, threshold tuning, evaluation, and comparison with LOF. Isolation Forest is trained in an unsupervised manner, while the Is Fraudulent label is used only as evaluation ground truth. The final configuration uses contamination 0.05, 250 estimators, max_samples 0.80, threshold 0.45, and random_state 42. The model achieves Accuracy 0.9820, Precision 0.8201, Recall 0.8200, F1-Score 0.8200, ROC-AUC 0.9187, and PR-AUC 0.6458. The confusion matrix contains 197,321 True Negatives, 8,613 True Positives, 1,890 False Positives, and 1,891 False Negatives. Isolation Forest outperforms LOF on most metrics, although LOF provides slightly higher Recall. The detection result is intended as an early indicator for transactions requiring further review.
PREDIKSI FRAUD TRANSAKSI DOMPET DIGITAL MENGGUNAKAN ALGORITMA CATBOOST Mochamad Vikry Firdaus; Puspita Nurul Sabrina; Herdi Ashaury
JOURNAL SAINS STUDENT RESEARCH Vol. 4 No. 5 (2026): Oktober: Jurnal Sains Student Research
Publisher : CV. KAMPUS AKADEMIK PUBLISING

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61722/jssr.v4i5.12752

Abstract

The growth of digital transactions via digital wallet and mobile money services has heightened the need for fraud detection systems capable of accurately identifying suspicious transactions. A major challenge in fraud detection is class imbalance, where the volume of fraudulent transactions is significantly lower than that of legitimate ones. This study aims to analyze and compare the performance of Random Forest and CatBoost algorithms in detecting fraud within digital transactions using the PaySim dataset. Random Forest served as the benchmark model, while CatBoost was the primary model evaluated through both baseline configurations and hyperparameter tuning. Class imbalance was addressed using class weighting, and the dataset was partitioned using an 80:20 train-test split combined with stratified sampling. Model performance was evaluated using Accuracy, Precision, Recall, F1-Score, ROC-AUC, and PR-AUC, alongside an analysis of False Positives and False Negatives. The results indicate that while Random Forest achieved a Recall of 98.78%, it produced 18,284 False Positives, resulting in a Precision of 8.15%. The CatBoost Baseline model achieved a Precision of 79.19%, Recall of 85.27%, F1-Score of 82.12%, ROC-AUC of 99.92%, and PR-AUC of 90.93%. Following hyperparameter tuning, the CatBoost Tuned model improved its Recall to 92.08%, ROC-AUC to 99.96%, and PR-AUC to 94.46%, while reducing the number of False Negatives from 242 to 130 transactions; however, Precision dropped to 71.46% and the F1-Score to 80.48%. These findings demonstrate that CatBoost Tuned possesses superior fraud detection capabilities—particularly in maximizing Recall and minimizing False Negatives—whereas CatBoost Baseline offers a better balance between Precision and Recall at the classification threshold used in this study.
PREDIKSI RATING FILM DENGAN PERBANDINGAN ALGORITMA CATBOOST DAN RANDOM FOREST Muhammad Irmanto; Puspita Nurul Sabrina; Herdi Ashaury
JOURNAL SAINS STUDENT RESEARCH Vol. 4 No. 6 (2026): Desember : Jurnal Sains Student Research
Publisher : CV. KAMPUS AKADEMIK PUBLISING

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61722/jssr.v4i6.12775

Abstract

The increasingly rapid development of the film industry has caused the number of films produced and published to continue to rise, resulting in an ever-growing volume of available film data. This condition drives the need for methods capable of utilizing film metadata to predict rating values. This study aims to build a film rating prediction model and compare the performance of the CatBoost and Random Forest algorithms using the TMDB 5000 Movies dataset obtained from Kaggle. The initial dataset consists of 4,803 film records with vote_average as the target variable and eight predictor attributes, namely budget, runtime, release_year, genres, original_language, main_country, main_language, and production_company. The research stages include data quality checking, preprocessing, JSON attribute extraction, feature engineering, categorical data transformation, an 80:20 data split, modeling, hyperparameter tuning, evaluation using MAE, RMSE, and R², as well as 5-fold cross-validation. Random Forest uses One-Hot Encoding for the retained categorical features, while CatBoost processes certain categorical features natively. The hold-out results show that CatBoost achieved an MAE of 0.6596, an RMSE of 0.9727, and an R² of 0.3725, whereas Random Forest achieved an MAE of 0.6604, an RMSE of 0.9778, and an R² of 0.3660. In the 5-fold cross-validation, Random Forest achieved a Mean MAE of 0.6709, a Mean RMSE of 0.9942, and a Mean R² of 0.3061, while CatBoost achieved a Mean MAE of 0.6719, a Mean RMSE of 0.9948, and a Mean R² of 0.3048. Thus, the two algorithms have relatively comparable performance; CatBoost is slightly better on the hold-out, while Random Forest is slightly better on the cross-validation. Both models were then integrated into a web application using Flask as the machine learning service and CodeIgniter 4 as the interface.
Prediksi Pengagguran Menggunakan Decision Tree Dengan Algoritma C5.0 Pada Data Penduduk Kecamatan Caringin Kabupaten Bogor Muhammad Dzatul Kahfi; Fajri Rakhmat Umbara; Herdi Ashaury
Informatics and Digital Expert (INDEX) Vol. 4 No. 2 (2022): INDEX, November 2022
Publisher : LPPM Universitas Perjuangan Tasikmalaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36423/index.v4i2.913

Abstract

Tingkat kesejahteraan dalam kehidupan bermasyarakat dapat dilihat dari tingkat penganggurannya. Pemerintah daerah biasanya mengadakan sebuah program untuk membantu mengurangi jumlah pengangguran, entah itu dengan mengadakan sebuah pelatihan atau hal lain yang dapat mendorong kreativitas masyarakat dan meningkatkan kemampuan hardskill agar dapat bersaing di dunia kerja. Ada banyak penelitian yang memprediksi tingkat pengangguran dan juga ada penelitian yang menggunakan algoritma C5.0 untuk melakukan prediksi, namun belum ada penelitian yang menggabungkan subjek dan metode tersebut. penelitian ini bertujuan untuk membuat sebuah model prediksi menggunakan algoritma C5.0 terhadap data penduduk kecamatan caringin dan mencari skenario dengan hasil akurasi yang paling tinggi. namun terdapat beberapa permasalahan yang harus dihadapi seperti bagaimana tingkat akurasi Model klasifikasi Decision Tree dengan algoritma C5.0 terhadap dataset penduduk Kecamatan Caringin dan Bagaimana resio data latih data uji dan penggunaan pruning memengaruhi tingkat akurasi prediksi yang akan dilakukan. Penelitian ini dievaluasi menggunakan beberapa skenario rasio data latih dan data uji yang berbeda beda dan penggunaan pruning yang berbeda. Hasil dari penelitian ini adalah model prediksi pengangguran berhasil dibuat dengan tingkat akurasi paling tinggi yaitu pada skenario data latih 70% dan data uji 30% dengan menerapkan teknik post pruning.