Claim Missing Document
Check
Articles

Found 37 Documents
Search

Analisis Sentimen Ulasan Aplikasi CapCut Berbasis IndoBERT dengan Validasi Silang Tasya Nurdin; Dodo Zaenal Abidin; Kurniabudi Kurniabudi
Prosiding Seminar Nasional Ilmu Teknik Vol. 2 No. 2 (2025): Desember: Prosiding Seminar Nasional Ilmu Teknik
Publisher : Asosiasi Riset Ilmu Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61132/prosemnasproit.v2i2.157

Abstract

This study conducts sentiment analysis of Indonesian user reviews of the CapCut application using IndoBERT and compares two evaluation schemes: a single 80/20 train–test split and stratified 5-fold cross-validation (k=5). A total of 1,048,575 reviews were collected from the Google Play Store through web scraping and labeled into three sentiment classes based on rating: negative (1–2), neutral (3), and positive (4–5). After preprocessing—cleaning, case folding, banned-word removal, normalization—and duplicate removal, 517,962 reviews were retained. IndoBERT Base P1 was fine-tuned using fixed hyperparameters (batch size 32, learning rate 2e-5, up to 4 epochs, early stopping patience 2), while undersampling was applied to the training set to address class imbalance. Performance was assessed using accuracy, precision, recall, F1-score, and ROC-AUC, supported by confusion matrix and ROC-curve visualizations. The single split achieved an accuracy of 0.756, whereas cross-validation produced a mean accuracy of 0.740. Across both schemes, the positive class achieved the best performance (F1-score 0.850; ROC-AUC 0.918–0.919), while the neutral class remained the most challenging (precision 0.198–0.206; F1-score 0.280–0.283). Overall, cross-validation is recommended for reporting because it reduces dependence on a single partition and provides a more representative estimate across multiple splits.
Model Prediksi Pelunasan Haji Berbasis XGBoost Dengan Interpretasi Shap: Studi Prediksi Pelunasan Haji dengan XGBoost dan SHAP di Provinsi Jambi Yan Apriadi; Dodo Zaenal Abidin; Jasmir Jasmir
Prosiding Seminar Nasional Ilmu Teknik Vol. 2 No. 2 (2025): Desember: Prosiding Seminar Nasional Ilmu Teknik
Publisher : Asosiasi Riset Ilmu Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61132/prosemnasproit.v2i2.178

Abstract

This study develops an interpretable machine learning model to predict the settlement status of Hajj fees in Jambi Province, Indonesia. Utilizing the XGBoost algorithm on a dataset of 4,332 prospective pilgrims from 2025, the research addresses the critical challenge of class imbalance where only 28.5% of samples are labeled "Unsettled". The baseline XGBoost model achieved a ROC-AUC of 0.7778, with a recall of 0.3482 for the minority class. SHAP (SHapley Additive exPlanations) analysis was employed to interpret model predictions, revealing that financial features specifically NILAI_VA (Virtual Account Value), JML_SETORAN (Deposit Amount), and JML_PELUNASAN (Settlement Amount) are the most significant factors influencing repayment risk, with negative SHAP values indicating increased default probability. The findings demonstrate that an interpretable XGBoost framework can provide both predictive accuracy and actionable insights for policymakers, enabling targeted interventions such as flexible payment schemes and enhanced financial monitoring for high-risk pilgrims..
Social Feature Integration for Entertainment Hoax Detection: Machine Learning and DistilBERT-Fusion: Dodo Zaenal Abidin; Agus Siswanto; Chindra Saputra; Bhetantio Bhetantio
Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi) Vol 10 No 4 (2026): August 2026
Publisher : Ikatan Ahli Informatika Indonesia (IAII)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29207/resti.v10i4.7208

Abstract

Detecting entertainment hoaxes remains a pressing challenge in the social media era, where short texts, provocative headlines, and rapid dissemination complicate verification. This study compares the effectiveness of classical machine learning models and the transformer DistilBERT, both with and without the integration of social features. The modified GossipCop dataset, consisting of 22,140 entries, includes news titles as the primary text representation and social features such as tweet count, tweet density, and viral indicators. Text was represented using TF-IDF for classical models and DistilBERT tokenization for the transformer, with performance evaluated under stratified 10-fold cross-validation. Results show that incorporating social features consistently improves classical models, with XGBoost + Social achieving the best performance (PR-AUC 0.84; F1-score 0.76), surpassing DistilBERT-Fusion (PR-AUC 0.79). McNemar’s test confirmed significant differences in error distributions, strengthening the reliability of these findings. To reduce reliance on a single empirical dataset, the social feature integration pipeline was additionally validated on the PolitiFact benchmark from a different domain, where incorporating minimal social signals yielded measurable and statistically significant gains over text-only models (Wilcoxon p = 0.0137), confirming the robustness of the social feature effect beyond the entertainment domain. Overall, the results highlight that for short-text entertainment news under the conditions examined in this study, boosting models enriched with social signals can outperform transformer-based approaches. While DistilBERT-Fusion provided competitive results, its improvement over text-only DistilBERT remained limited due to the short-text nature of the dataset, indicating opportunities for richer fusion strategies in future research.
Design of an Android-Based Augmented Reality Virtual Menu Application for Ejji Coffee Shop in Jambi City FA Bambang Sukoco; M Rakesh; Dodo Zaenal Abidin
Jurnal Manajemen Teknologi Dan Sistem Informasi (JMS) Vol 6 No 2 (2026): JMS Vol 6 No 2 September 2026
Publisher : LPPM Universitas Dinamika Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33998/jms.2026.6.2.2537

Abstract

Kedai Kopi Ejji is located in Jambi city, specifically in Beringin. Many customers are confused by the coffee menu options due to a lack of explanations, forcing them to call the waiter for assistance. Augmented reality (AR) technology can be an effective solution. AR provides an immersive experience and can be used as an attractive promotional tool. With AR, the menu, which initially consists of text and images, can be visualized interactively. It combines virtual objects with real ones, capturing the attention of consumers, especially in areas where this technology is still new. the use of AR can increase sales revenue in several ways. First, AR enhances the customer experience with 3D visualizations of the menu, making customers feel more connected to the products. It also reduces confusion, helping them make decisions more quickly. additionally, the interactive elements of AR can encourage impulsive purchases, increasing the average transaction value. AR also improves service efficiency, reduces the workload of waitstaff, and speeds up the service process. The implementation of AR provides a competitive advantage, attracting more new customers. The AR application can be promoted through social media, encouraging customers to share their experiences, which positively impacts visitor numbers and revenue. The AR application serves as an innovative alternative to conventional menus.
Analisis Kepuasan Pengguna Aplikasi EMOS Menggunakan Model Delone dan Mclean Pada PT.Enseval Putera Megatrading Cabang Jambi Susanto Hertio Bagus Saputro; Dodo Zaenal Abidin; Joni Devitra
Jurnal Informatika Dan Rekayasa Komputer(JAKAKOM) Vol 6 No 1 (2026): JAKAKOM Vol 6 No 1 APRIL 2026
Publisher : LPPM Universitas Dinamika Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33998/jakakom.2026.6.1.2339

Abstract

Abstract− This study aims to evaluate user satisfaction with the EMOS (Enseval Mobile Order System) application at PT. Enseval Putera Megatrading Cabang Jambi, using the DeLone and McLean model. The main focus of this research is to identify the impact of system quality, information quality, and service quality on application usage, as well as their effect on user satisfaction and perceived benefits. The study uses a quantitative approach with a questionnaire distributed to 150 users, and data analysis was conducted using Structural Equation Modeling (SEM) based on Partial Least Squares (PLS) with the SmartPLS software. The findings indicate that system quality, information quality, and service quality significantly influence the usage of the EMOS application and user satisfaction. Specifically, service quality plays a crucial role in enhancing user satisfaction and the perceived benefits, such as increased operational efficiency and ease of ordering. These findings provide recommendations for PT. Enseval Putera Megatrading to improve the EMOS platform's quality to better meet user needs and support the company's operational efficiency.
Perancangan Sistem Informasi Santri Berbasis Web Dengan Menerapkan Algoritma AES (Advanced Encryption Standard) Pada Yayasan Griya Fadhilah Al-Qur’an Mohammad Badri; Dodo Zaenal Abidin; Joni Devitra
Jurnal Informatika Dan Rekayasa Komputer(JAKAKOM) Vol 6 No 1 (2026): JAKAKOM Vol 6 No 1 APRIL 2026
Publisher : LPPM Universitas Dinamika Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33998/jakakom.2026.6.1.2349

Abstract

Information systems are now an important part of an organization because they support decision making, simplify activity processes, and help manage data. The process of managing student data at the Griya Fadhilah Al-Qur'an Foundation is still done manually, so there is a risk of damage, loss, unauthorized access, or even misuse of data. Attendance is also still done manually using paper, which is susceptible to damage and loss. Filling in assessments and monitoring student development using Microsoft Excel is also susceptible to input errors, calculations, and data loss. This study aims to design a web-based student information system and implement the AES (Advanced Encryption Standard) algorithm in the system at the Griya Fadhilah Al-Qur'an Foundation. The system development method used is the Waterfall method, with UML modeling tools, including use case diagrams and class diagrams. This system was developed using the PHP programming language with the Laravel framework, and the MySQL database. The results of this study are a web-based student information system by implementing the AES algorithm that can help the foundation manage student data and improve data security. By implementing AES, student data, assessments and student development can be converted into characters that cannot be read or interpreted directly, which is expected to increase security and prevent unauthorized access or misuse of data.
Comparison of Machine Learning Algorithms for Classifying Scientific Journal Article Titles in the Computer Science Field Muhammad Fadhillah; Dodo Zaenal Abidin; Jasmir
Jurnal Informatika Dan Rekayasa Komputer(JAKAKOM) Vol 6 No 2 (2026): JAKAKOM Vol 6 No 2 SEPTEMBER 2026
Publisher : LPPM Universitas Dinamika Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33998/jakakom.2026.6.2.2591

Abstract

The increasing number of scientific publications in the computer science field poses challenges for automatic classification of article titles, which tend to be short, ambiguous, and contain domain-specific technical terms. This study aims to compare the performance of three machine learning algorithms—k-Nearest Neighbor (kNN), Naïve Bayes, and Random Forest—in classifying the titles of computer science journal articles into five main topics: AI & Machine Learning, UI/UX, IT Framework, IoT, and Cybersecurity. A dataset of 2,449 article titles was collected from the SINTA portal and underwent text preprocessing steps including tokenization, case normalization, stopword removal, and stemming, followed by numerical representation using TF-IDF weighting. Models were trained and tested using separate datasets and evaluated based on accuracy, precision, recall, and F1-score. The results show that kNN (k=51) achieved the highest accuracy of 89.61% with consistent precision, recall, and F1-score across all categories. Random Forest ranked second with 81.46% accuracy, showing strong performance in categories such as Cybersecurity and AI & Machine Learning, although it suffered a significant drop in recall for Cybersecurity. Naïve Bayes recorded the lowest accuracy (71.35%) but achieved the fastest execution time (12 seconds). These findings emphasize that algorithm selection should consider data characteristics, particularly for short-text classification such as article titles, and indicate that kNN is the most stable and accurate choice in this context.