cover
Contact Name
Yosep Septiana
Contact Email
yseptiana@itg.ac.id
Phone
+6282124588750
Journal Mail Official
algoritma@itg.ac.id
Editorial Address
Jl. Mayor Syamsu No.1, Jayaraga, Kec. Tarogong Kidul, Kabupaten Garut, Jawa Barat 44151
Location
Kab. garut,
Jawa barat
INDONESIA
Jurnal Algoritma
ISSN : 14123622     EISSN : 23027339     DOI : https://doi.org/10.33364/algoritma
Core Subject : Science,
Jurnal Algoritma merupakan jurnal yang digunakan untuk mempublikasikan hasil penelitian dalam bidang Teknologi Informasi (TI), Sistem Informasi (SI), dan Rekayasa Perangkat Lunak (RPL), Multimedia (MM), dan Ilmu Komputer (Computer Science).
Articles 1,145 Documents
Penilaian Risiko Fraud Transaksi Digital menggunakan Hybrid Machine Learning dengan Clustering dan Klasifikasi Hendra Wijaya; Naek Parulian Hutagalung; Mira Afrina; Ali Ibrahim; Fathoni
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.23-1.3398

Abstract

Credit card transaction fraud detection is commonly treated as a binary classification problem, whereas operational risk management requires more detailed risk-level information to support investigation prioritization. This study proposes a hybrid machine learning framework for transaction risk stratification. In the first stage, the K-Means algorithm was applied to the training set to discover latent risk structures and generate cluster-based risk labels. Subsequently, a Random Forest model was trained to predict risk levels for new transaction data. To maintain evaluation objectivity, the dataset was divided into training, validation, and testing sets, and data leakage prevention mechanisms were implemented. The testing results show that the model was able to consistently classify two levels of risk with stable precision, recall, and F1-score values. In the binary fraud detection scenario, the model achieved an accuracy of 0.8831. These findings indicate that separating latent risk exploration from predictive classification can produce a more informative risk representation compared to conventional binary approaches. However, this study is still limited to a single public dataset and one classification model. Therefore, the generalizability and potential performance improvements of the model still need to be evaluated by experimenting with other algorithms.
Implementasi CRM Berbasis Web Untuk Meningkatkan Kepuasan Pelanggan di Toko Adi Glass Muhammad Fahrozi; Jhonson Efendi Hutagalung; Ruri Ashari Dalimunthe
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.23-1.3401

Abstract

Advances in information technology in the era of digital transformation are driving businesses to leverage technology to enhance competitiveness and service quality. Adi Glass, a business in the home furnishings sector, still faces challenges in managing customer data and sales transactions, which are currently handled manually, thereby reducing service effectiveness and sales performance. This study aims to design and implement a web-based Customer Relationship Management (CRM) system and evaluate its impact on sales performance and customer satisfaction. The method used is a mixed-methods approach involving interviews, observations, literature review, and questionnaires. The results show that the CRM system is capable of integrating customer and sales data and supports structured promotion management. The evaluation indicates a 30% increase in data management efficiency, a 25% increase in service speed, and a 27% increase in customer satisfaction. The contribution of this research lies in the development of an integrated web-based CRM system with promotional features as a sales improvement strategy for small businesses.
Klasifikasi Pola Serangan Keamanan Jaringan pada UNSW-NB15 Menggunakan Pendekatan Machine Learning Afrizal Dzikri Arifah; Diah Risqiwati
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.23-1.3403

Abstract

The increasing complexity and volume of network traffic have created challenges for attack detection systems in recognizing diverse attack patterns and handling imbalanced class distributions. This study aims to detect network attacks on the UNSW-NB15 dataset using supervised learning methods, namely Logistic Regression and Random Forest. The research stages include data preprocessing, model development, and evaluation using Stratified 5-Fold Cross Validation with Accuracy, Precision, Recall, F1-Score, and AUC metrics. The results show that Random Forest achieved better performance than Logistic Regression, with an accuracy of 0.9515, precision of 0.9631, and AUC of 0.9924. Cross-validation results also showed that the average accuracy of Random Forest was 0.9504, higher than Logistic Regression at 0.9038. Feature analysis indicates that attributes based on Time to Live, traffic, data volume, and temporal characteristics contribute significantly to the detection process. Therefore, Random Forest demonstrates more optimal and stable performance in detecting network attacks on the UNSW-NB15 dataset, while Logistic Regression remains relevant as a simple and interpretable comparison model.
Seleksi Kelapa Unggul untuk Ekspor Menggunakan Metode SAW Shafiq Aiman Manurung; jhonson Efendi Hutagalung; Sri Rezki Maulina Azmi
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.23-1.3407

Abstract

The growth of the coconut industry in Indonesia presents significant opportunities for increasing exports of dried coconuts; however, the selection process for high-quality coconuts at CV. Sejati Kilang is still conducted manually and subjectively, leading to inconsistencies in quality assessment. This study aims to design and implement a decision support system based on the Simple Additive Weighting (SAW) method for the objective and efficient selection of high-quality coconuts based on five criteria: size, moisture content, cleanliness, maturity level, and physical condition. The system was developed as a web-based application using PHP and MySQL. Test results showed that Hybrid Coconuts received the highest score of 0.95, followed by Pandan Wangi Coconuts at 0.86 and Local Coconuts at 0.81. This study contributes to the development of a web-based superior coconut selection system integrated with the SAW method, resulting in a more structured, objective, and consistent evaluation process that supports rapid decision-making. However, this study is still limited in the number of alternatives and criteria, so further development is needed to improve accuracy and enable application on a larger scale.
Peningkatan Knowledge Capture dan Knowledge Sharing dalam KMS Tools dengan Kaizen Form Faisal Hakim Amrullah; Hendry; Irwan Sembiring
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.23-1.3409

Abstract

This study discusses the improvement of knowledge capture and knowledge sharing through the strengthening of a Kaizen-based Knowledge Management System (KMS) in the footwear manufacturing industry. The main problems include the suboptimal management of tacit knowledge and the limitations of document search based on simple keywords. This study applies an information retrieval method using TF-IDF and Cosine Similarity on 800 validated Kaizen documents through preprocessing, weighting, and document similarity measurement stages. The test results show that the proposed method performs better than conventional keyword-based search, with a precision value of 0.60, recall of 0.75, and F1-score of 0.67. The contribution of this study lies in the application of information retrieval methods to improve the effectiveness of knowledge retrieval in a Kaizen-based KMS, thereby supporting continuous improvement and organizational learning.
Model Rekomendasi Karier Lulusan Sekolah Menengah Kejuruan Berdasarkan Kompetensi dan Bakat Menggunakan Perbandingan Algoritma Apriori dan FP-Growth Tarwan; Eko Aji Putra; Arief Wibowo
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.23-1.3410

Abstract

The increasingly dynamic development of the job market requires Vocational High School (SMK) graduates to possess adaptive abilities, not only in mastering vocational competencies but also in determining appropriate career paths. However, in reality, many SMK graduates still experience difficulties in choosing careers that align with their competencies and talents. This condition highlights the need for a systematic approach capable of providing data-driven career recommendations. This study aims to develop a data-based career recommendation system using the Apriori and FP-Growth algorithms to identify relationship patterns among vocational competencies, students’ academic talents, and alumni tracer study data. The study offers a new approach to career recommendation systems for Vocational High Schools by integrating students’ academic data and alumni post-graduation histories (tracer studies) within a single pattern analysis framework. In addition to generating association rules that can easily be used as a basis for decision-making, the system also incorporates validation from guidance and counseling teachers (BK teachers) to strengthen data-driven career decisions. Talents are classified into two categories, namely exact/science-oriented and non-exact/non-science-oriented, based on comparisons between average Mathematics grades and non-science subject grades from semesters 1 to 6. Alumni tracer data include post-graduation status (employment, higher education, or others), job relevance, competency certificates, and the positions or work sectors pursued. Subsequently, each student and alumni entity was transformed into transactional data analyzed using the Apriori and FP-Growth algorithms to discover association rules between student profiles and career recommendations. The analysis results indicate strong relationships between combinations of talents and vocational competencies with specific career choices. The inclusion of data from guidance and counseling teachers serves as qualitative input that strengthens the validity of the system’s results. This system can be utilized by schools, guidance counselors, and students as a decision-support tool for making more objective and data-driven career decisions. Therefore, the system supports a vocational education direction that is more integrated with labor market needs.
Implementasi Algoritma First Come First Serve (FCFS) pada Sistem Manajemen Pengaduan Online APTIKA Rizka Annisa; Anita Ahmad Kasim
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.23-1.3412

Abstract

The high demand for digital services, which is not yet supported by an integrated complaint management system, has led to low transparency and slow handling of reports. This study developed a web- and mobile-based complaint management system by implementing the First Come First Serve (FCFS) algorithm for queue management. The research method uses a mixed method with black-box functional testing and performance evaluation through a pre-test and post-test scheme analyzed using the Paired Sample T-Test. The results show an acceleration in resolution time across all service categories with efficiency levels ranging from 60.9% to 91.3%. Statistically, the significance value (p-value < 0.05) proves an improvement in performance compared to the manual method. The contribution of this research is the integration of FCFS and Role-Based Access Control (RBAC) into a structured and efficient multi-service-based government complaint system. The limitations of this study lie in the use of simulation data in the post-test as well as pre-test data that is estimative with a limited sample, so the results do not yet fully represent real operational conditions and still require further testing.
A Analisis Sentimen Publik terhadap Program Makan Bergizi Pemerintah (MBG) pada Media Sosial X Menggunakan Metode Support Vector Machine Windi Haria Ningsi; Ucta Pradema Sanjaya
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.23-1.3413

Abstract

The Nutritious Meal Program (MBG) is one of the government’s initiatives to improve the nutritional quality of society. This study aims to analyze public sentiment toward the MBG program on social media platform X in order to provide a quantitative overview of public perception. The data used consisted of 1,938 public posts containing keywords related to the MBG program. The analysis stages included text preprocessing, namely data cleaning, tokenization, stopword removal, and stemming. Furthermore, feature representation was carried out using the TF-IDF method, while sentiment classification was performed using the Support Vector Machine (SVM) algorithm to categorize the data into positive, neutral, and negative sentiments. The results indicate that the classification model achieved an accuracy rate of 96.69 percent, demonstrating excellent model performance. Based on the classification results, sentiment distribution was dominated by negative sentiment, followed by neutral and positive sentiments, indicating that public responses toward the MBG program tend to be critical. These findings suggest that although the MBG program has received significant public attention, there are still various criticisms and feedback that can serve as evaluation material for the government in improving the effectiveness and implementation of the program.
Desain dan Implementasi Smart Contract untuk Pengelolaan Persetujuan Akses Data Pasien Berbasis Blockchain Anggie Ciecilia Saragih; Bayu Angga Wijaya; Jon Kevin Sihombing; Febryco Rives; Soeli Yanto Rotua Marbun
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.23-1.3421

Abstract

This study aims to design and implement blockchain-based smart contracts for secure, transparent, and patient-oriented patient data access consent management. The research method employed a systems approach combining qualitative and quantitative methods through waterfall development stages. The system was developed using the Ethereum Sepolia Testnet blockchain and Solidity-based smart contracts. The implementation results demonstrate that blockchain technology is capable of permanently and transparently recording all patient data consent transactions. The smart contract successfully implemented a Role-Based Access Control (RBAC) mechanism, allowing patients to grant and revoke access permissions for doctors or healthcare institutions. The testing results indicate that the access validation mechanism functioned properly, although there are limitations related to scalability and gas costs on public blockchains. Security evaluation was limited to functional testing and access validation, indicating the need for further testing such as penetration testing and smart contract vulnerability analysis. Overall, this study proves that blockchain technology and smart contracts are capable of improving security and trust in digital healthcare data management, while also supporting the future development of artificial intelligence-based Decision Support Systems.
Penerapan Algoritma XGBoost dengan SMOTE untuk Klasifikasi Kanker Payudara pada Dataset Wisconsin Adrianus Anggoro; Imam Tahyudin; Ades Tikaningsih
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.23-1.3423

Abstract

Penelitian ini bertujuan untuk mengembangkan model deteksi kanker payudara menggunakan algoritma Extreme Gradient Boosting (XGBoost) pada Breast Cancer Wisconsin Diagnostic Dataset. Dataset terdiri dari 569 sampel dengan 30 fitur medis yang merepresentasikan karakteristik morfologi tumor, dengan dua kelas target yaitu benign (jinak) dan malignant (ganas). Tahapan penelitian meliputi pembersihan data, imputasi nilai hilang, normalisasi fitur, serta penerapan teknik Synthetic Minority Over-sampling Technique (SMOTE) untuk menangani ketidakseimbangan kelas. Model XGBoost dievaluasi menggunakan metrik akurasi, precision, recall, dan F1-score. Hasil pengujian menunjukkan bahwa model mencapai akurasi sebesar 94,55%, dengan nilai recall kelas malignant sebesar 95,24%, yang mengindikasikan kemampuan tinggi dalam mendeteksi kanker ganas. Confusion matrix menunjukkan hanya 2 kasus false negative, menandakan sensitivitas model yang sangat baik terhadap kelas minoritas. Dibandingkan dengan model tanpa SMOTE, penerapan SMOTE terbukti meningkatkan recall pada kelas malignant secara signifikan. Hasil penelitian ini menunjukkan bahwa algoritma XGBoost dengan penanganan imbalance class efektif digunakan sebagai sistem pendukung keputusan dalam diagnosis kanker payudara dan berpotensi membantu deteksi dini secara lebih akurat.