JSAI (Journal Scientific and Applied Informatics)

Designing a Web-Based Asset Management Information System at PT Global Asia Sinergi

Ari Yanto — 2025-11-03

Manual asset management often causes problems such as data duplication, recording errors, and difficulties in tracking asset status. This study aims to design and develop an efficient, adaptive, and user-friendly web-based Asset Management Information System using the Agile (Scrum) development method and measuring its success through the System Usability Scale (SUS) evaluation. The study was conducted at PT Global Asia Sinergi with direct users as active participants in each development iteration. The development process was carried out through several sprints covering the planning, implementation, testing, and evaluation stages with users. The Agile evaluation results showed an average sprint success rate of 88%, with a stable velocity of 45 story points per iteration and positive user feedback (a score of 4.3 out of 5). Meanwhile, the SUS test results obtained an average score of 78.5, which falls into the “Good Usability” category, indicating that the system is easy to use, consistent, and reliable in supporting the company's operational activities. Thus, it can be concluded that the application of the Agile method in the development of an asset management information system can improve the efficiency of the development process while producing a system with a high level of usability in accordance with user needs.

Digital Transformation, Social Media, and Information Technology on the Competitiveness of Used Car MSMEs in Batam City

DESI RENATA SINURAT DESI — 2025-11-03

This study aims to analyze the influence of digital transformation, social media, and information technology on the competitiveness of used car MSMEs in Batam City. A quantitative research approach was employed with a sample of 89 respondents. The results of the classical assumption tests indicate that the data are normally distributed, as shown by the Kolmogorov–Smirnov Z value of 0.068 and a significance level of 0.200 (>0.05). The multicollinearity test results show no indication of multicollinearity, with Tolerance values greater than 0.10 and VIF values below 10 (X₁ = 1.685; X₂ = 1.062; X₃ = 1.614). The heteroscedasticity test results also confirm that the model is free from heteroscedasticity, as all independent variables have significance values greater than 0.05. The simultaneous F-test shows that digital transformation, social media, and information technology collectively have a significant effect on MSME competitiveness, with an F-value of 3.264 and a significance of 0.000 (<0.05). Partial t-tests indicate that all three variables have a positive and significant influence, with a significance value of 0.000. Digital transformation has the most dominant effect (B = 0.406, t = 2.908), followed by information technology (B = 0.306, t = 2.438) and social media (B = 0.066, t = 2.518). These findings demonstrate that implementing digital transformation supported by effective utilization of social media and information technology can enhance the competitiveness of used car MSMEs in Batam City by improving operational efficiency, service innovation, and digital marketing effectiveness.

Web-Based Company Profile Information System for Bengkel Tralis Mang BUUK Using the Likert Scale

Yanti Amaliasari — 2025-11-03

This study aims to measure user satisfaction with the Web-Based Company Profile Information System for Tralis Workshop, which was developed as a digital platform for information dissemination and business promotion. The measurement was conducted using the User Satisfaction method, which includes three main dimensions: system quality, information quality, and service quality. A quantitative descriptive approach was applied, and data were collected through a five-point Likert scale questionnaire distributed to 30 respondents, consisting of workshop owners, employees, and customers. The analysis results indicate that system quality achieved an average score of 4.35 or 87% (very satisfied category), information quality reached 4.21 or 84.2% (satisfied category), and service quality obtained 4.12 or 82.4% (satisfied category). Overall, the user satisfaction level reached an average score of 4.23 or 84.6%, categorized as satisfied. These findings demonstrate that the system successfully meets user expectations, particularly in terms of usability, access speed, and clarity of information. Furthermore, the implementation of this web-based information system effectively enhances the workshop’s professional image, strengthens its promotional reach, and serves as a model for user-oriented digital transformation among small and medium-sized enterprises (SMEs).

Fake News Detection Model Using BERT and Bi-LSTM Based on Discriminative Approach

Dwi Fitri Brianna — 2025-11-11

This study aimed to develop a text classification model for detecting hoaxes using a deep learning approach and text representation methods. The text data that had undergone preprocessing were then extracted using three approaches: Word2Vec, Doc2Vec, and Bidirectional Encoder Representations from Transformers (BERT). The research dataset consisted of 2,325 genuine news articles (label 0) and 2,287 fake news articles (label 1). In this study, BERT feature vectors with a dimension of 768 were combined with the Bidirectional Long Short-Term Memory (Bi-LSTM) algorithm to capture sequential dependencies in the text, along with the Support Vector Machine (SVM) algorithm as the final classifier. The training process was carried out on Dell Precision 7750 hardware using parameters of embedding dimension 128, 64 hidden units, a dropout rate of 0.3, and a learning rate of 0.001. Training and testing were conducted for 10 epochs with a batch size of 32. The results indicated that the Word2Vec and Bi-LSTM model achieved an accuracy of 87.4% with an F1-Score of 87.0%, while the Doc2Vec and Bi- LSTM model performed slightly lower with an accuracy of 85.6% and an F1- Score of 85.4%. The best performance was obtained by the BERT, Bi-LSTM, and SVM model, which achieved an accuracy of 93.8%, precision of 94.1%, recall of 93.5%, and an F1-Score of 93.7%.

Implementation of Agile Development Methods in the Development of Mobile-Based Work Stress Level Monitoring Applications

Putra Afiqi — 2025-11-03

This study aims to develop the WARASPADA (Work Stress Awareness and Prevention Application Dashboard) as a mobile-based work stress monitoring tool using the Agile Development methodology, with the Flutter framework for the frontend and Supabase as the backend. The instrument used in this research is the Work Stress Diagnosis Survey (SDS) issued by the Indonesian Ministry of Manpower. System evaluation was conducted using the Technology Acceptance Model (TAM), which measures four key variables: Perceived Usefulness (PU), Perceived Ease of Use (PEOU), Attitude Toward Using (ATU), and Behavioral Intention to Use (BIU). The results show that the overall user acceptance level reached 85.4%, categorized as very good. This finding indicates that users perceive the application as easy to use, functional, and effective for detecting and monitoring work stress levels. Therefore, the WARASPADA application is considered feasible and has strong potential to be implemented in workplace environments as a data-driven decision-support tool for human resource management.

Classification of Oil Palm Plant Diseases Based on Hybrid Deep Learning Using U-Net and ResNet-

Fakhri Lambardo — 2025-11-11

Oil palm (Elaeis guineensis) productivity was frequently constrained by foliar diseases, which were often difficult to detect at an early stage using conventional visual inspection methods. To address this challenge, the present study proposed a hybrid deep learning framework for automated oil palm leaf disease detection. A dataset comprising 1,200 oil palm leaf images, equally distributed across three disease classes (400 images per class), was utilized. The dataset was partitioned into training (70%), validation (15%), and testing (15%) subsets, with training and validation data obtained from public repositories, while testing data were collected directly to ensure model generalizability. The proposed hybrid architecture combined U-Net for precise leaf lesion segmentation, ResNet-50 as a deep feature extractor to capture high-level discriminative representations. U-Net segmentation enabled isolation of infected regions, while ResNet-50 provided robust feature embeddings that enhanced separability between visually similar disease classes. Experimental evaluation demonstrated that the baseline U-Net + SVM approach achieved an accuracy of 84.2%, precision of 82.5%, recall of 83.1%, and F1-score of 82.8%. In contrast, the hybrid U-Net + ResNet-50 + SVM method yielded superior results with 91.6% accuracy, 90.8% precision, 91.2% recall, and 91.0% F1-score, reflecting an improvement of approximately 7.4%.

Implementation of Push Notifications in Android-Based Patient Queuing Applications

Mohamad Styvani Hendi S — 2025-11-03

This study aims to implement and evaluate the push notification feature in an Android-based patient queue application to improve service efficiency and user satisfaction. The research process includes several stages: needs analysis, system design, feature implementation, functional testing, and user satisfaction testing using the User Satisfaction method. Data were collected through a Likert-scale questionnaire covering five aspects: ease of use, interface design, response speed, information relevance, and overall satisfaction. The results show that the application successfully provides real-time queue information through push notifications, enhancing user convenience. The average user satisfaction rate exceeded 80%, with the highest score of 88% achieved in the information relevance aspect. Reliability testing using Cronbach’s Alpha obtained a coefficient value of 0.87, indicating that the questionnaire instrument was reliable and internally consistent. Therefore, the implementation of push notifications has proven effective in improving user experience and service efficiency in digital patient queue management.

Evaluation of the Quality of Tomoro Coffee's Digital Services on the Satisfaction of the Productive Population Using the E-Service Quality Model

Vina Dewi Ramadhanty — 2025-11-06

The trend of drinking coffee among productive people has impacted the growth of coffee shops in Indonesia. To strengthen its market position, Tomoro Coffee utilizes mobile applications. However, there is dissatisfaction with the quality of Tomoro Apps' digital services in reviews. This study evaluated the quality of Tomoro's digital application services using the Electronic Service Quality model. The model focuses on comprehensive digital services from technical aspects and other aspects related to user satisfaction. The assessment was conducted on students of Jember University who made transactions through Tomoro Apps. Data was collected using a questionnaire involving 100 students as research samples. The data were processed using instrument testing, classical assumption testing, and multiple linear regression analysis. An analysis of the data showed that efficiency, responsiveness, and contact positively affected user satisfaction. From the data analysis, seven dimensions of e-service quality were known to have a significant positive and negative effect and were able to explain 88.3% of user satisfaction variability. Through this study, we found that there is still a gap between expectations and application performance. Therefore, improvements in digital service quality should be made in each dimension of electronic service quality, balanced and focused on service users, so that Tomoro Coffee continues to be a competitive advantage. These findings have managerial implications for Tomoro Coffee as recommendations on each dimension of Electronic Service Quality to improve the quality of digital services and strengthen its position in the market.

Topic Modeling Based on Computer Science Research Documents Using Text Mining

Bakhtiar Bakhtiar — 2025-11-11

This study aimed to develop a document clustering model using a combination of the IndoBERT model and the K-Means algorithm to group research abstracts in the field of computer science and technology. The data used consisted of 1000 research abstracts, divided into two parts: 80% for training data (800 abstracts) and 20% for testing data (200 abstracts). The IndoBERT model was used to represent the abstracts as embedding vectors, which were then processed with the K-Means algorithm to form 10 topic clusters, including artificial intelligence, computer systems and networks, programming, cybersecurity, and others. The training experiment used the training data to generate clusters and centroids for mapping new documents into the appropriate clusters. Evaluation was carried out using several metrics, including accuracy, cluster homogeneity, Davies-Bouldin Index, and Silhouette Score. The testing results showed that the developed model achieved an accuracy of 85%, indicating good performance in clustering the test data. The cluster homogeneity value of 0.90 indicated that documents that should belong to the same cluster were grouped together effectively. The Davies-Bouldin Index value was 0.34, while the Silhouette Score was 0.76.

Exam Cheating Detection Model Using SURF-CNN Approach Based on Digital Image Data

Uus Rusmawan — 2025-11-11

Detecting cheating during examinations was one of the main challenges in maintaining academic integrity in educational environments. This study developed an approach to detect cheating behavior by utilizing a combination of Speeded Up Robust Features (SURF) and Convolutional Neural Networks (CNN) based on digital images. The novelty of this research lay in the application of the SURF method as a feature extraction technique to detect suspicious objects and movements, which were then further analyzed using CNN for student behavior classification. The main objective of this study was to design, develop, and test a digital image–based exam cheating detection model capable of recognizing various types of behavior, such as looking around, glancing suspiciously, as well as non-cheating behaviors like focusing and boredom. The dataset used in this study consisted of 1,200 digital images categorized into six different behavior classes. The dataset was divided into three parts: 70% for training (840 images), 10% for validation (120 images), and 20% for testing (240 images). The experimental results showed that the SURF-CNN approach achieved better performance compared to the standard CNN. The SURF-CNN model achieved an accuracy of 91.80% on training data, 88.65% on validation, and 86.15% on testing, while CNN only achieved 88.20% on training, 85.25% on validation, and 83.15% on testing

Academic Fraud Detection Based on Handwritten Documents Using Siamese ResNet Approach

Azhar Andika Putra — 2025-11-11

Academic cheating, particularly involving the forgery of handwriting documents, has become a significant challenge in the field of education. One of the most difficult modes to detect is identity misuse, where an individual writes or completes tasks on behalf of someone else. This study aimed to develop an academic cheating detection model based on handwriting using the Siamese Network approach and cosine similarity. The experiment was conducted using an HP Z8 G5 device equipped with two NVIDIA RTX 6000 GPUs, and the dataset used came from Universitas Sjakhyakirti. The dataset consisted of 101,475 pairs of handwriting images, each labeled as 1 (similar) for pairs from the same writer and 0 (dissimilar) for pairs from different writers. The data was divided into 70% for training, 15% for validation, and 15% for testing. This research dataset was sourced from handwriting documents of 450 different students, consisting of 450 positive pairs (label = 1) and 101,025 negative pairs (label = 0). The model was evaluated using a cosine similarity threshold of 0.5, with training accuracy reaching 95.34%, validation accuracy at 84.12%, and testing accuracy at 83.87%. This study contributes to the development of a handwriting-based academic cheating detection system that can be implemented in higher education institutions.

Classification of Fish Species in South Sumatra Based on Underwater Imagery Using ResNet-50

Lemi Iryani — 2025-11-11

Fish species classification in underwater ecosystems posed a significant challenge, particularly due to poor lighting that affected the quality of underwater images and decreased the accuracy of species identification. This study aimed to improve the accuracy of fish species classification in South Sumatra based on underwater images by utilizing the Super-Resolution Generative Adversarial Network (SRGAN) to enhance image quality and ResNet-50 for species classification. The research employed a Dell XPS 13 9310 device with an Intel Core i7 processor and 16GB of RAM for model training. Fish image data were collected from Google Images and YouTube according to predefined fish species, including Oreochromis mossambicus (Mujair), Oreochromis niloticus (Nila), Johnius trachycephalus (Gulamah), Eleutheronema tetradactylum (Senangin), and Chanos chanos (Bandeng). The data was divided into 70% for training, 15% for validation, and 15% for testing. The experimental results showed that the developed model achieved a training accuracy of 94.10%, validation accuracy of 88.25%, and testing accuracy of 84.68%. This research contributed to the field of underwater image classification and can be applied to conservation and monitoring of fish species in aquatic ecosystems.

Blockchain-Based Authentication and Traceability System for Local Creative Batik and Weaving Products

Desi Ramayanti — 2025-11-11

This research aimed to develop a blockchain-based system to ensure product authentication and traceability for creative textile products, with a case study on local Batik and Weaving SMEs in Bandung City. The main issue addressed was the widespread counterfeiting of products and the lack of transparency in the supply chain, which harmed both artisans and consumers. To address this, a system architecture was designed that utilized Hyperledger Fabric as the private blockchain platform, Node.js for the backend server, and IPFS for off-chain data storage. So far, the research has reached a technology readiness level (TRL) of 4, where the system prototype was validated in a laboratory environment. Key components such as the registration and verification module for SMEs by administrators, the product data input system, and the mechanism for generating a unique QR code for each product were successfully implemented. The system was developed to enhance consumer trust, protect product authenticity, and increase the market value of local creative textile products.

Model Analisis Kinerja Sumber Daya Manusia (SDM) di Perguruan Tinggi Berbasis Multi-modal Data Menggunakan Machine Learning

Reni Utami — 2025-11-11

This research aimed to develop a machine learning-based model for human resource (HR) performance analysis in higher education institutions using a multi-modal approach, combining static data and text data analysis. For static data analysis, the random forest (RF) algorithm was employed to assess HR performance based on attributes such as years of service, training attended, and performance evaluations. The dataset for this experiment consisted of 250 data points, which were divided into 70% for training, 10% for validation, and 20% for testing. The experimental results with the RF model showed high accuracy in training (90.4%), although there was a performance drop during validation and testing, with accuracies of 85.7% and 82.5%, respectively. For the text data, which contained feedback with negative, positive, and neutral sentiments, the CNN-BiLSTM model achieved an accuracy of 92.6% in training, despite a decrease in validation (87.4%) and testing (84.4%) accuracies. The text dataset comprised 1,000 data points, divided into 70% for training, 10% for validation, and 20% for testing. The study recommends the application of a multi-model approach to assess HR performance using the RF algorithm for static data and the CNN-BiLSTM model for more complex data in future research.

Implementation of Token-Based Access Control and Content Protection in Web-Based Learning Management to Improve Data Security in Medical Education Institutions

Ahmad Sulaeman — 2025-11-17

Digital content security is a critical aspect in the implementation of online learning systems, especially in medical education institutions that manage sensitive video data. This study aims to develop and evaluate a web-based video content protection system by integrating three main methods: AES-128 encryption, Token-Based Access Control, and Dynamic Watermarking. The AES-128 encryption method is applied to secure video files so that only authorized users can decrypt and access them, while Token-Based Access Control serves as a dynamic session-based authentication mechanism to prevent unauthorized access. In addition, Dynamic Watermarking embeds user identity information into each video playback to trace and deter illegal content distribution. Experimental results show a Token Validation Success Rate of 100% and an Intrusion Detection Rate of 90%, indicating that the system performs well in verifying authentication and detecting unauthorized access. Performance testing of the HTTP Live Streaming (HLS) process achieved an average response time of 1.79 seconds with 1000 successful requests and no failures, demonstrating that the additional security layers did not significantly degrade system performance. Overall, this study concludes that the integration of AES-128 encryption, Token-Based Access Control, and Dynamic Watermarking provides an effective multilayer security approach for protecting video-based learning content and strengthening data security in medical education environments.

Sentiment Analysis of Yamet Clinic Health Services Instagram Comments With Naive Bayes

Reina Salsa Kinanty — 2025-11-19

The rapid advancement of information technology has driven the healthcare sector to adopt digital platforms in order to enhance service efficiency and accessibility. Yamet Clinic Palembang, as a therapy center for children with special needs, has utilized social media as its primary communication channel, although its official website has not yet been fully optimized. The main issue addressed in this study is the lack of understanding regarding public perception of the clinic’s digital services. Therefore, this research aims to analyze public sentiment toward Yamet Clinic Palembang’s services through comments posted on its official Instagram account. The Naïve Bayes algorithm was selected due to its strong performance in text classification tasks involving limited datasets and low computational complexity. The research process includes text preprocessing, feature weighting using Term Frequency–Inverse Document Frequency (TF-IDF), and the application of the Synthetic Minority Over-sampling Technique (SMOTE) to balance data distribution among sentiment categories. The experimental results indicate that the developed model successfully classified public opinions into three categories—positive, neutral, and negative—with an accuracy rate of 70%. The combination of TF-IDF, Naïve Bayes, and SMOTE proved effective in capturing public perceptions of digital health services. Practically, the findings provide valuable insights for Yamet Clinic Palembang in developing a more adaptive digital communication strategy and theoretically contribute to the advancement of sentiment analysis research within the healthcare service context in Indonesia.

Integration of Body Mass Index (BMI) and Total Daily Energy Expenditure (TDEE) Calculations in Healthy Lifestyle Monitoring Applications

Ulya Nidaul — 2025-11-21

This study aims to evaluate the performance of a healthy lifestyle monitoring application integrated with Body Mass Index (BMI) and Total Daily Energy Expenditure (TDEE) calculations. The evaluation focuses on three primary aspects: usability using the System Usability Scale (SUS), adherence through usage consistency monitoring, and effectiveness through pre–post statistical analysis. The study involved 30 respondents for usability assessment and 30 participants for adherence measurement and statistical testing. The findings show that the application demonstrates excellent usability with an SUS score of 82.4. User adherence reached 63.3%, surpassing the minimum target set for the study. Additionally, the statistical test resulted in p = 0.003, indicating a significant difference between pretest and posttest conditions. These results suggest that the application provides a positive user experience, maintains user engagement, and has a measurable impact on users’ understanding and behavior in monitoring their health status and daily energy needs.

Market Basket Analysis of Coffee Shop Customers Using the FP-Growth Algorithm

haditsah annur — 2025-11-21

One of the businesses that attracts economic development is a coffee shop. This business is very important and growing rapidly. Shopping cart analysis has the ability to provide information about which products are frequently purchased together. The use of shopping cart analysis is regulated in association rules which is a data processing process that provides records of purchase transactions that come out simultaneously at one time, the algorithm used to regulate these association rules is the FP-Growth algorithm. coffee shop customer shopping cart analysis uses the FP-Growth Algorithm. This research data was obtained from public data on the website https://www.kaggle.com/datasets/sryasuka/coffee-shop-dataset/data,, with a dataset of 1000 transactions, the data processing uses RapidMiner tools, after processing, 2 association rules were found using minimum support = 0.01 and minimum confidence = 0.7. It can be concluded that the results of the shopping cart analysis show that 1 item is most frequently purchased by customers, namely croissants and the purchase of 2 items, namely croissants and fries. So that the shopping basket analysis method with the FP-Growth algorithm can optimize item combination patterns and can improve sales strategies, thereby supporting coffee shop business decision making.

Implementation of SPBE-FORGE(6D) as an Enterprise Architecture Framework in Accelerating Local Government Digital Transformation

Setiawan Assegaff — 2025-11-25

Digital transformation within local governments requires a standardized architectural framework to ensure service integration, data consistency, and efficient information technology governance. This study examines the implementation of SPBE-FORGE(6D) as an Enterprise Architecture framework designed to support the acceleration of Indonesia’s Electronic-Based Government System (SPBE). The evaluation employed Enterprise Architecture Success Metrics, including architecture artefact quality, system interoperability, application efficiency, and data consolidation indicators. The results show that SPBE-FORGE(6D) improved architecture artefact quality by 35–36%, reduced application duplication by 41%, decreased data silos by 44%, and enhanced digital service integration by 150%. These findings demonstrate that SPBE-FORGE(6D) effectively strengthens IT governance, improves cross-domain architectural integration, and significantly contributes to accelerating local government digital transformation in an efficient, standardized, and sustainable manner.

Implementation of Fuzzy Mamdani for Laptop Quality Classification Based on Specification Parameters

Bambang Cahyono — 2025-11-27

This study implements the Mamdani fuzzy logic method to classify laptop quality based on key technical specifications, including processor performance, RAM capacity, storage type, and price. The fuzzy model was developed using linguistic membership functions and rule-based inference to represent expert judgment patterns. System performance was evaluated using the Mean Absolute Error (MAE) metric, which measures the average deviation between the model output and expert reference values. The results show that the Mamdani fuzzy model provides highly accurate and consistent classifications, indicated by a MAE value of 0.027 (2.7%). The error values for each laptop sample ranged from 0.02 to 0.04, demonstrating the model’s stability and effectiveness across various specification levels. These findings confirm that Mamdani fuzzy logic is suitable for laptop quality classification and can serve as a foundation for the development of intelligent recommendation systems for consumer electronics.

Implementation of the Analytical Hierarchy Process in a Multi-Criteria University Selection Recommendation System

Anton Topadang — 2025-11-29

Choosing a university is a multi-criteria decision-making process that requires a systematic approach to ensure objective and transparent recommendations. This study applies the Analytical Hierarchy Process (AHP) to determine the priority weights of criteria in university selection. The research stages include constructing a decision hierarchy, conducting pairwise comparisons, calculating weights, and performing consistency validation. The results indicate that criterion K1 is the most dominant factor with a weight of 0.447, followed by K2 with 0.227 and K5 with 0.161. Meanwhile, K3 and K4 have relatively lower weights. The Consistency Ratio (CR) value of 0.038 confirms that the judgments are consistent and meet the acceptable threshold of CR < 0.1. These findings demonstrate that AHP provides a robust and measurable evaluation framework for university recommendation systems. The proposed model supports users in making informed decisions based on systematic and justifiable preference weighting.

Designing Rawi Apps for Android Mobile

Siti Zahrotul Fajriyah — 2025-11-29

The digital transformation has shifted religious literacy consumption patterns towards mobile platforms requiring high accessibility. However, existing Rawi applications are predominantly image-based, limiting display flexibility and readability for users across different ages, especially the elderly. This study aims to design a Rawi application on Android Mobile implementing an offline-first approach to ensure fast and inclusive content access. The research method employs a Mixed Method with Design Thinking, comprising Empathize, Define, Ideate, Prototype, and Test stages. The application is designed based on the Flutter framework with imaging SQLite local database and Firebase synchronization. The design testing involves functional testing and non-functional testing (Heuristic Evaluation, Cognitive Walkthrough, and First-Click Testing). The results indicate that all functional requirements are met, and the design is considered intuitive with a First-Click success rate of 91.6%, exceeding the feasibility threshold of 87%. This study concludes that the digital text-based Rawi application design successfully enhances accessibility and user experience through personalization features such as font resizing, advanced searching, as well as bookmark and favorite management.