TEKNIMEDIA: Teknologi Informasi dan Multimedia

PRINCIPAL COMPONENT ANALYSIS - CONTRAST LIMITED ADAPTIVE HISTOGRAM EQUALIZATION AND ATT-UNET FOR HAIR SEGMENTATION

2026-06-14T04:27:36+07:00

In the field of medical image analysis, artifacts such as dermal hair pose a major challenge to both visual interpretation and automated image processing during dermoscopic examinations. Hair covering the lesion area can obscure the lesion boundaries, reduce the quality of feature extraction, and lead to segmentation and classification errors. Recent studies have shown that dermal hair remains one of the most persistent artifacts affecting automated analysis, even in state-of-the-art segmentation models. These artifacts also degrade the performance of AI-based systems that rely on visual information. This study aims to improve the accuracy of hair segmentation in dermoscopic images through the application of effective and efficient preprocessing techniques. This study applies Principal Component Analysis (PCA) as a grayscale method to reduce the computational burden while preserving essential image features, and Contrast-Limited Adaptive Histogram Equalization (CLAHE) to enhance local contrast and highlight thin or low-contrast hair structures. The combination of PCA and CLAHE serves as a preprocessing stage to improve the quality of input images for deep learning-based segmentation models. The main contribution of this research is the integration of PCA-based grayscale methods with CLAHE in a single preprocessing pipeline before deep learning segmentation and the evaluation of their effects on the performance of the segmentation model. The evaluation is conducted using the AttU-Net architecture with Dice Similarity Coefficient (DSC) and Jaccard Index (JAC) metrics. The proposed PCA–CLAHE preprocessing achieves DSC and JAC values of 75.24% and 61.04%, respectively, outperforming the model without preprocessing. These results indicate that PCA–CLAHE effectively improves image quality and segmentation accuracy while maintaining computational efficiency.

DAMAGE DETECTION SYSTEM FOR FLOATING SOLAR PANELS AT SIDOBANDUNG POND BASED ON CONVOLUTIONAL NEURAL NETWORKS WITH AUGMENTED REALITY VISUALIZATION

2026-06-14T04:27:37+07:00

This study aims to develop an Augmented Reality (AR) application integrated with a Convolutional Neural Network (CNN) as an interactive system for detecting damage in floating solar power plant (PLTS) panels at Embung Sidobandung in order to maintain the efficiency of the photovoltaic energy system. Conventional manual inspection methods are considered inefficient and prone to errors due to human factors. Therefore, a deep learning approach is employed to automatically and interactively detect and classify solar panel damage. AR technology is utilized to display panel condition information directly through a mobile device camera, enabling real-time damage monitoring. The dataset consists of 615 solar panel images, including 472 images of physical damage and 143 images of electrical damage. Experimental results show that the system is capable of classifying solar panel damage types in real time, achieving a precision of 93.48%, recall of 89.58%, and an F1-score of 91.49% for physical damage, and a precision of 70.59%, recall of 80.00%, and an F1-score of 75.00% for electrical damage, with an overall accuracy of 87.30%. Although the developed application provides interactive and informative visualization, varying lighting conditions in aquatic environments and differences in image acquisition angles remain challenges that affect system accuracy. Overall, the integration of CNN and AR has the potential to serve as an effective and efficient solution for developing damage detection systems for floating solar power plant (PLTS) panels.

HYBRID CNN AND LEXICON-BASED MODEL FOR DETECTING PUBLIC SENTIMENT TO-WARDS DANANTARA

2026-06-17T19:57:30+07:00

Danantara is a sovereign wealth fund that has received widespread attention on social media, including on "X." Various public opinions about Danantara have emerged, both positive and negative. To understand public perception more accurately, an effective sentiment analysis method is needed. Lexicon-Based methods are often used but have limitations in capturing context, while Convolutional Neural Networks (CNNs) are able to recognize patterns in text but require a lot of data. Therefore, this study aims to develop a hybrid CNN and Lexicon-based model to improve the accuracy of public sentiment analysis towards Danantara. Data will be collected from social media "X" using scraping or API methods, then go through a preprocessing process before being analyzed using a hybrid approach. Model evaluation is carried out by comparing the performance of CNN, Lexicon-based, and hybrid models using metrics such as accuracy, precision, recall, and F1-Score. The results of this study provide a more accurate model in understanding public opinion and provide insights for policymakers in designing more effective communication strategies with an accuracy rate of 86%. In addition, researchers will also contribute to the development of Natural Language Processing (NLP) and deep learning-based sentiment analysis.

MARKET SUPPLY AND DEMAND PREDICTION USING RANDOM FOREST AND PRINCIPAL COMPONENT ANALYSIS METHOD

2026-06-14T04:27:37+07:00

Inaccurate forecasting of stock requirements and market demand is a major challenge faced by PT Bintang Jaya, which may lead to excess inventory (overstock) or stock shortages (stockout). This issue occurs because the company’s stock planning process still relies on manual approaches and historical experience without optimal utilization of data analytics. Therefore, this study aims to apply machine learning–based prediction techniques to estimate stock needs and market demand more accurately. The methods used in this research include the Random Forest algorithm as the baseline model, and Random Forest combined with Principal Component Analysis (PCA) as a hybrid model to evaluate the impact of dimensionality reduction on prediction performance. The dataset consists of historical sales transaction records from PT Bintang Jaya during the 2022–2024 period, which were processed through data preprocessing, monthly aggregation, and time series feature engineering. The results show that the Random Forest model provides more stable demand predictions and is closer to the actual values compared to the hybrid RF+PCA model. The application of PCA did not improve prediction performance due to the characteristics of the dataset, which is relatively low-dimensional and non-linear. Overall, the baseline Random Forest model demonstrates good and stable performance, indicated by consistent MAE and RMSE values and a coefficient of determination (R²) of approximately 0.69, meaning that the model explains around 69% of the demand variation based on the historical features.

COMPARISON OF THE EFFECTIVENESS OF WHATSAPP GROUPS AND WEBSITES AS INFORMATION MEDIA FOR TAM-BASED PPDB

2026-06-14T04:27:37+07:00

The development of digital technology has encouraged schools to utilize various platforms to deliver information, particularly in the new student admission process, commonly referred to as PPDB. MIN 7 Tangerang uses WhatsApp Groups and the school website as the primary media for disseminating PPDB information; however, the effectiveness of these two media has not been determined. This study aims to compare the effectiveness of WhatsApp Groups and the website as media for PPDB information using the Technology Acceptance Model (TAM) approach. The study employed a quantitative method supported by qualitative data. Data were collected from 136 parents of admitted students who passed the PPDB selection for the 2024/2025 academic year, all of whom had accessed and used both media. The research instrument consisted of a Likert-scale questionnaire based on TAM constructs along with additional variables. Cronbach’s Alpha was used to test the reliability of the instrument, while validity testing was conducted using Exploratory Factor Analysis (EFA). Furthermore, a paired sample t-test was applied to compare two related conditions, and regression analysis was employed to examine the effect of Perceived Ease of Use (PEOU) and Perceived Usefulness (PU) on technology acceptance. The results indicate that WhatsApp Groups are superior in terms of ease of use, whereas the website demonstrates superiority in terms of usefulness, trust, and user satisfaction. The regression analysis confirms that both PEOU and PU significantly influence technology acceptance, with PU exerting a more effect. In conclusion, WhatsApp Groups and the website play complementary roles in delivering PPDB information.

THE ROLE OF PERSONALIZATION, RECOMMENDATION SYSTEMS, INFORMATION QUALITY, AND E-SERVICE QUALITY IN IMPROVING SHOPEE USER SATISFACTION: AN SEM-PLS APPROACH

2026-06-14T04:27:37+07:00

User satisfaction has become a crucial factor in the success of e-commerce platforms amid increasingly fierce competition, particularly for Shopee as the platform with the highest number of visits in Indonesia. This study aims to analyze the influence of personalization, recommendation systems, information quality, and electronic service quality (e-service quality) on Shopee user satisfaction through the mediation of Perceived Usefulness (PU) and Perceived Ease of Use (PEOU) within the Technology Acceptance Model (TAM) framework. The research method employed is a quantitative approach with a survey of 430 active Shopee users in the Jabodetabek area who have completed at least two transactions in the last three months. Data were analyzed using Partial Least Square-based Structural Equation Modeling (SEM-PLS) with SmartPLS software. The results show that the recommendation system is the strongest predictor of PU, while e-service quality is the main determinant of PEOU. PU has the most dominant direct influence on user satisfaction, followed by PEOU. All mediation paths proved to be significant, with the recommendation system having the strongest indirect effect through PU. The research model can explain 73.1% of the variance in user satisfaction. It can be concluded that the integration of intelligent technology and basic service quality simultaneously shapes perceptions of usefulness and ease of use, which become the main pillars of e-commerce user satisfaction in Indonesia.

DEVELOPMENT AND EXPERIMENTAL EVALUATION OF A REAL-TIME EGG QUALITY DETECTION SYSTEM BASED ON DIGITAL IMAGE PROCESSING AND YOLO MODEL ON EDGE DEVICES

2026-06-14T04:27:37+07:00

The demand for chicken eggs as a nutritious protein source continues to rise, yet automated inspection technology in small-scale farms remains limited due to high costs. This study develops a standalone, low-cost real-time egg quality detection prototype based on the edge computing paradigm. The system is implemented on a Raspberry Pi 4 Model B using the YOLOv8n deep learning model to classify eggs into three categories: Good, Cracked, and Broken. Visual data is acquired via a Raspberry Pi Camera Module 3 supported by controlled white LED ring lighting at a fixed distance of 20 cm to mitigate environmental light variations. Experimental evaluation using 1,080 samples indicates that the system achieves an optimal accuracy of 91.30% at 320x320 resolution under controlled lighting. Technically, the system demonstrates stable performance with CPU usage ranging from 43% to 76%, while maintaining temperatures at 48-51⁰C. Despite a processing speed of 0.5-0.6 FPS, the system's independence from cloud connectivity makes it a highly applicable objective inspection solution for small-scale farmers in regions with limited digital infrastructure.

DIGITAL ENGINEERING OF ECONOMIC SYSTEMS FOR CIRCULAR ECONOMY DEVELOPMENT BASED ON ISLAMIC ECONOMIC PRINCIPLES IN EAST LOMBOK, INDONESIA

2026-06-14T04:27:37+07:00

The governance of conventional communal waste banks is frequently characterized by managerial inefficiencies and a heavy reliance on manual record-keeping, which fosters information asymmetry. From the perspective of Islamic economics, such conditions manifest as gharar (informational uncertainty), potentially undermining amanah (trust) and ‘adl (fairness in valuation), while simultaneously neglecting the circularity of organic materials—thereby contravening the principle of Hifzh al-Bi’ah (environmental preservation).

This study aims to engineer, develop, and evaluate a digital platform ecosystem for waste bank management (SiRKAH) that inherently integrates circular governance with the principles of Maqashid Sharia. Employing the Design Science Research Methodology (DSRM), the system artifact was validated by a multidisciplinary expert panel using the Content Validity Index (CVI) and its efficacy was assessed through field-based operational comparisons (pre- versus post-implementation). The expert validation results confirm a highly robust architectural validity, with an S-CVI score of 0.939. Operationally, the implementation of a digital ijab-qabul protocol (dual authentication) and an immutable ledger has demonstrated significant efficacy disruption: reducing transaction duration by up to 94.2%, completely eliminating (100%) recording errors and balance disputes, and increasing user participation rates by 46.4%. This study contributes to the Green Information Systems (Green IS) literature through the conceptualization of a “Gharar-Free System Architecture,” demonstrating that the fusion of digital technological engineering with religio-communal ethical values constitutes a robust instrument for restoring public trust and accelerating the adoption of a circular economy.

DATA QUALITY IMPROVEMENT: CASE STUDY FAST PAYMENT SYSTEM INFRASTRUCTURE

2026-06-14T04:27:38+07:00

This study aims to assess the data quality of the Fast Payment System infrastructure application. The assessment employs a specialized framework for financial data, namely the QAFD (Quality Assessment Framework for Data), which focuses on five data quality dimensions. Given the high transaction volume and operational criticality of fast payment systems, ensuring reliable data quality is essential to support reporting, analysis, and policy communication. Data quality assessment was conducted through two approaches: objective and subjective assessment, applied to 34 variables. The objective assessment was based on quantitative measurements of the data, while the subjective assessment involved user or stakeholder perceptions of data quality. The objective assessment results showed that the Accuracy dimension reached 99.99%. The Completeness dimension for mandatory data variables was recorded at 84.41%. For the Uniqueness dimension, the variable sending_customers_id_number_hash achieved 87.34%, while receiving_customers_id_number_hash reached 99.96%. Meanwhile, both the Currency and Timeliness dimensions achieved a 100% rate. A comparison between the objective and subjective assessment results indicated discrepancies in the Completeness and Uniqueness dimensions, while the other dimensions were aligned. These findings indicate that data quality challenges extend beyond technical processing aspects and reflect the importance of continuously improving metadata clarity and business rule alignment governing variable usage and identifier relationships. This study contributes by providing empirical evidence of QAFD implementation in fast payment operational data and emphasizes the value of proactive data governance through metadata enhancement and strengthened validation mechanisms to support reliable reporting and institutional credibility.

KPI-BASED ADAPTIVE SCORING MODEL OF MSME PERFORMANCE USING QUANTILE CLIPPING AND DATA-DRIVEN WEIGHTING

2026-06-14T04:27:38+07:00

Monitoring the performance of Micro, Small, and Medium Enterprises (MSMEs) is a challenge for local governments due to the large number of business actors, heterogeneity of business scale, and variations in financial indicators that often contain extreme values. These conditions cause Key Performance Indicators (KPI)-based evaluations to be susceptible to bias when using conventional normalization and weighting. This study aims to develop an adaptive scoring model for MSME performance based on KPIs that is robust to outliers and more objective in indicator weighting. The proposed method integrates quantile clipping at the P5–P95 percentiles to stabilize the KPI distribution, followed by min–max normalization to the range 0–100. Furthermore, KPI weights are determined in a data-driven manner using standard deviation (adaptive weighting) to represent the indicator's contribution based on actual data variations. Experiments were conducted on a dataset of 1,000 MSMEs in Serang City using three main KPIs, namely ROI, Profit Margin, and Growth Rate. The results show that the adaptive weights obtained are ROI 0.308, Profit Margin 0.353, and Growth Rate 0.339. A ranking comparison between fixed weighting and adaptive weighting yielded a Spearman correlation of 0.9879, and two entities changed in the Top 10. These findings indicate that the adaptive method maintains ranking stability while increasing evaluation objectivity. The proposed model is computationally efficient and has potential for application in multi-indicator-based performance monitoring systems.

INTERNET OF THINGS–BASED CHINNING UP AND PULL UP COUNTING SYSTEM USING INFRARED AND EMG SENSORS

2026-06-14T04:27:38+07:00

This study aims to design and implement an automatic counting tool for Chinning Up and Pull Up exercises based on the Internet of Things (IoT) to improve accuracy and efficiency in physical fitness assessment. The system uses SHARP GP2Y0A21 infrared sensors to detect the number of repetitions of the movement and Electromyography (EMG) sensors to measure biceps and triceps muscle activity during exercise. Data from both sensors is processed by an ESP32 microcontroller and sent in real-time to the Firebase Realtime Database, then displayed through a web interface based on the Laravel Framework. This system consists of three user roles, namely admin, supervisor, and athlete, each with different management, monitoring, and exercise result access functions. Testing was conducted using Blackbox Testing and User Acceptance Test (UAT) to assess system performance and user experience. The test results showed that the system had a sensor reading accuracy rate of 96.8% for infrared sensors and 94.5% for EMG sensors in detecting real-time muscle movements and activities. The measurement results can be accessed directly and responsively through the website, facilitating the process of monitoring athlete performance. Thus, this system has proven to be effective in supporting objective, efficient, and integrated physical training evaluation.

ANALYSIS OF FACTORS INFLUENCING CONTINUANCE INTENTION OF DIGITAL BANK USERS IN JABODETABEK: A STUDY ON USERS OF JENIUS, NEOBANK, DIGIBANK, AND TMRW APPLICATIONS

2026-06-14T04:27:38+07:00

The rapid development of digital technology has accelerated the transformation of the banking industry, particularly through the emergence of digital banking applications that enable customers to conduct financial transactions more efficiently and conveniently. This study aims to analyze the factors influencing the continuance intention of digital banking users by examining the roles of perceived ease of use, perceived usefulness, customer service quality, perceived security, and perceived risk in shaping customer satisfaction and customer trust. This research employs a quantitative approach using the Partial Least Squares Structural Equation Modeling (PLS-SEM) method. Data were collected through questionnaires distributed to active users of several digital banking applications. The analysis results indicate that perceived ease of use and customer service quality play important roles in increasing customer satisfaction. In addition, perceived risk is identified as a crucial factor in shaping customer trust toward digital banking services. Furthermore, customer satisfaction is found to be a major determinant in encouraging users to continue using digital banking applications, while customer trust also contributes positively to the intention to maintain long-term usage. The structural model demonstrates strong explanatory capability in explaining users’ continuance intention. These findings highlight that improving user experience, strengthening service quality, and managing security and risk effectively are essential strategies for digital banking providers to maintain customer satisfaction, build trust, and encourage sustainable usage of digital banking services in Indonesia.

EVALUATING OF DEEP LEARNING MODELS FOR EARLY DETECTION IN MEAT CLASSIFICATION: A STUDY ON BEEF AND PORK DETECTION

2026-06-14T04:27:38+07:00

Accurate classification of beef and pork images is crucial for developing reliable automated food inspection systems, particularly due to their visual similarity in color, texture, and muscle fiber patterns. This study aims to comparatively evaluate the performance of multiple deep learning models for binary meat image classification using RGB digital images. Four Convolutional Neural Network (CNN) architectures, namelyInceptionV3, VGG16, ResNet50, and Xception were assessed under identical preprocessing pipelines and hyperparameter settings to ensure a fair comparison. The dataset underwent cropping, resizing to 224×224 pixels, normalization, and augmentation to enhance variability and improve generalization performance. Model effectiveness was measured using accuracy, precision, recall, and F1-score on unseen test data. Experimental results show that InceptionV3 achieved the most balanced classification performance, with a test accuracy of 72% and an F1-score of 0.7. Although Xception obtained higher training accuracy, it exhibited overfitting during testing, while VGG16 and ResNet50 demonstrated comparatively lower classification capability. These findings indicate that InceptionV3 provides a more stable and generalizable architecture for beef and pork image classification. The study highlights the importance of cross-architecture evaluation in developing robust CNN-based systems for automated meat classification.

DESIGNING AN INTELLIGENT DIGITAL LIBRARY SYSTEM WITH LOCATION DETECTION AND INTEGRATED POINT MANAGEMENT

2026-06-14T04:27:38+07:00

This study designed and developed a web-based intelligent digital library system aimed at overcoming low reading interest and library visits. This system allows users to access digital book collections for free within the library environment, as well as regulate access to digital books outside the library through a point system mechanism as an incentive and control mechanism. The Prototype method was used in the development of this system, which involved the stages of requirements gathering, rapid design, and prototype evaluation. Functional testing of the system was carried out using the Blackbox Testing method. The results of the implementation show that this system has succeeded in providing easy access to digital reading materials, motivating interest in reading through a point system obtained from attendance and reading activities, and facilitating data management for librarians. The location detection feature via WiFi IP ensures that free digital book access is only available within the library zone, while access outside the zone requires point redemption. It is hoped that this system can increase users' interest in reading by combining easy access to digital information and optimising library technology.

INFORMATION SYSTEM SECURITY ANALYSIS USING SUPPORT VECTOR MACHINE ALGORITHM ON SHOPEE USERS

2026-06-15T13:09:09+07:00

The expansion of e-commerce in Indonesia has made information system security a crucial concern, especially on sites like Shopee that see a lot of user activity and transaction volumes. Potential security hazards, such as account misuse, unauthorized access, and suspicious activity, are increased by the volume of online transactions. Therefore, in order to comprehend the elements linked to security threats based on user characteristics and behavioral patterns, an analytical approach is necessary. The purpose of this study is to apply machine learning to examine security risk tendencies among Shopee users. A standardized questionnaire addressing demographic factors, usage frequency, security awareness levels, and experiences with questionable activity was used to gather data from 101 active users. Data cleaning, label encoding, Min–Max normalization, and feature selection were among the steps in the data processing procedure. The classification model used was the Support Vector Machine (SVM) technique with a Radial Basis Function (RBF) kernel. The creation of a security risk analysis model based on user perceptions and behavioral aspects rather than system log or transactional data is what makes this study unique. By using non-technical indications as predictive factors in e-commerce platforms, this method provides an alternate viewpoint for spotting possible security threats.

DESIGN AND IMPLEMENTATION OF AN INVENTORY INFORMATION SYSTEM TO IMPROVE THE EFFICIENCY AND EFFECTIVENESS OF INVENTORY MANAGEMENT AT INDAH JEJE SHOP

2026-06-14T04:27:38+07:00

Manual inventory management remains a common problem among small and medium-sized retail businesses, including Indah Jeje Shop. The use of manual systems causes difficulties in monitoring stock availability, increases the risk of recording errors, and hinders the preparation of accurate and timely inventory reports. This study aims to design and implement a web-based inventory information system to improve the efficiency and effectiveness of inventory management at Indah Jeje Shop. The system development method used in this study is the waterfall method, which consists of requirements analysis, system design, implementation, and testing stages. The system was developed using the Laravel framework with a MySQL database and applies role-based access control consisting of superadmin, admin, and owner. The results of system testing using the black-box testing method indicate that all system functions operate properly and meet user requirements. Furthermore, the User Acceptance Testing (UAT) results show a user acceptance rate of 81%, which falls into the good category. Based on these results, it can be concluded that the developed web-based inventory information system is able to support inventory management processes in a more structured, accurate, and efficient manner.

A MACHINE LEARNING SYSTEM ARCHITECTURE FOR PROACTIVE CUSTOMER CHURN PREDICTION

2026-06-14T04:27:38+07:00

The hyper-competitive credit card industry faces growing challenges from digital disruption and evolving consumer expectations, demanding a shift from reactive to proactive customer retention strategies. Traditional reactive approaches prove ineffective as customer decisions often reach irreversible stages before intervention. This study aims to develop and evaluate a comprehensive data-driven framework for predicting customer churn at PT XYZ, a leading Indonesian banking institution, and design a scalable system architecture with CRM integration and real-time analytics dashboard for operational deployment. Following the CRISP-DM framework, we comparatively evaluate Logistic Regression, Decision Tree, and Random Forest using a dataset of 11,314 customer records. Model performance evaluation encompasses multiple metrics including Accuracy, Precision, Recall, F1-Score, AUC. Random Forest algorithm demonstrated superior performance, achieving an AUC of 0.98 and accuracy of 97 percent. Feature importance analysis revealed customer transaction inactivity and credit utilization patterns as the most critical churn predictors, with transaction count contributing 41.59% importance score. The research successfully establishes a robust foundation for data-driven customer retention strategies, providing PT XYZ with a comprehensive blueprint for institutionalizing proactive retention strategies that can minimize revenue losses and secure competitive advantages in an increasingly dynamic market environment.

DESAINING INTERACTIVE DIGITAL FLIPBOOK WITH EDUCATIVE CHATBOT ON FOOD CHAIN MATERIAL FOR ELEMENTARY SCHOOL STUDENTS

2026-06-14T04:27:38+07:00

The advancement of digital technology demands the presence of innovative learning media that can support interactive and engaging learning processes, especially at the elementary school level. One of the materials that requires visualization support and good conceptual understanding is the food chain material in the Natural and Social Sciences (IPAS) subject. This study aims to design and develop learning media in the form of an interactive digital flipbook equipped with a simple educational chatbot on the food chain material for elementary school students. The research method used is the Research and Development (R&D) method by applying the ADDIE model which is limited to the analysis, design, and development stages. The developed media was then validated by two media experts and two material experts using an assessment instrument in the form of a questionnaire. The validation results showed that the flipbook learning media with an educational chatbot obtained a feasibility percentage of 93.3% from the media experts and 97.3% from the material experts, which is categorized as very feasible. Based on these results, the developed learning media is considered suitable for use as an alternative media in learning the food chain. It is hoped that this research can be a reference or reference material in the development of interactive digital learning media in elementary schools.

MANAGING IT PROJECTS FOR AI-DRIVEN PERSONAL-ADAPTIVE HOTEL INFORMATION SYSTEMS

2026-06-14T04:27:38+07:00

The rapid adoption of artificial intelligence (AI) in the hospitality industry has intensified expectations regarding service efficiency, personalization, and operational performance. Despite growing empirical evidence highlighting the potential benefits of AI-enabled systems, implementation outcomes across hotel contexts remain uneven. This inconsistency suggests that technological capability alone is insufficient to explain project success, underscoring the need to examine AI adoption through the lens of information technology (IT) project management. Accordingly, this study investigates how AI-driven IT projects contribute to operational efficiency in the hospitality sector and identifies managerial and organizational factors that differentiate successful implementations from those that underperform or fail. A structured literature review (SLR) was conducted to synthesize recent empirical and conceptual studies on AI implementation in hotel operations. The analysis focuses on operational performance outcomes across guest-facing and organizational domains, as well as contextual conditions shaping project execution. The results indicate that AI-driven IT projects are commonly associated with improvements in service responsiveness, personalization accuracy, internal workflow efficiency, and resource utilization. However, the magnitude and sustainability of these benefits vary considerably across implementation contexts. An aggregated analysis of operational outcomes reveals that projects achieving balanced improvements across both service and organizational dimensions tend to demonstrate more stable efficiency gains. The findings further highlight leadership commitment, stakeholder engagement, change management practices, and system integration depth as critical determinants of project success. By framing AI adoption as a socio-technical IT project rather than a standalone technological upgrade, this study contributes to the hospitality and information systems literature and offers actionable insights for managers seeking to align AI initiatives with organizational strategy and service delivery objectives

DEVELOPMENT OF A REGIONAL CREATIVE ECONOMY DIGITAL PLATFORM USING USER-CENTERED DESIGN CASE STUDY: EKRAFMAGELANG.ID

2026-06-14T04:27:39+07:00

The development of digital technology has driven significant transformations in the management and promotion of regional creative economies. However, the use of digital platforms by creative economy actors in the region still faces various obstacles, such as limited information presentation, unintegrated data management, and a less than optimal user experience. This study aims to develop a regional digital creative economy platform in Magelang Regency by applying the User-Centered Design (UCD) method so that the resulting system is in line with user needs and characteristics. The UCD method is implemented through several stages, including user-oriented planning, understanding the context of use, identifying user needs, designing solutions, and evaluating predetermined needs. Data collection was conducted through interviews with platform managers, creative economy actors, and users, followed by designing user flows and system mock-ups. The evaluation system was conducted using a usability test method to measure ease of use and interface comfort. The results of the study indicate that the developed digital platform is able to present product information and MSME profiles in a structured, easily accessible manner, and provides a good user experience. Thus, the application of the UCD method has proven effective in supporting the development of a regional digital creative economy platform and is expected to increase the promotion and visibility of creative economy actors in a sustainable manner.

IMPLEMENTASI DAN ANALISIS AUGMENTED REALITY (AR) SEBAGAI MEDIA PEMBELAJARAN DI WILAYAH 3T: STUDI KASUS PADA SMA NEGERI 1 ROTE BARAT

2026-06-14T04:27:39+07:00

This study explores the implementation and effectiveness of Augmented Reality (AR) technology as an innovative educational solution for overcoming learning challenges in Indonesia's 3T regions (Tertinggal, Terdepan, dan Terluar - Remote, Frontier, and Underdeveloped areas). Conducted at SMA Negeri 1 Rote Barat, East Nusa Tenggara, this research employs a mixed-methods approach combining qualitative and quantitative data collection through in-depth interviews, classroom observations, student questionnaires (n=60), and documentation analysis over six months (May-October 2025). The study develops a comprehensive research model incorporating Technology Acceptance Model (TAM) variables including Perceived Usefulness, Perceived Ease of Use, Interactivity, Technical Barriers, and Social-Cultural Support as independent variables, with Student Satisfaction as a mediating variable and Continuance Intention as the dependent variable. Results demonstrate that AR implementation significantly enhances student engagement, motivation, and understanding of complex concepts despite infrastructure limitations typical of 3T regions. Social-cultural adaptation and teacher support emerge as critical success factors. The research provides valuable insights for educational policymakers and institutions seeking to implement technology-enhanced learning in underserved areas while addressing unique challenges of remote educational contexts.

DEVELOPMENT OF AN INTERACTIVE AUGMENTED REALITY (AR) BASED MATHEMATICS TEACHING AID FOR VISUALIZING SURFACE AREA OF GEOMETRY

2026-06-14T04:27:39+07:00

Understanding the consept of surface area in geometry remains a challenge in mathematics learning, particularly due to students limited spatial visualization skills. This study aims to develop an interactive mathematics teaching aid based on Augmented Reality (AR) technology in the form of a mobile application. The development followed the ADDIE model (Analysis, Design, Development, Implementation, Evaluation) and resulted in an application featuring 3D visualizations, interactive nets of geometry, and audio explanations for four types of solids cube, cuboid, triangular prism, and cylinder. The application was validated by two media experts and two subject matter experts, and tested on seventh grade students of MTS Muta’alimin. The media expert validation showed a feasibility score of 82,5% (highly feasible category), the subject matter expert validation scored 92,5% (highly feasible category), and the student trial obtained an average score of 93,7% (highly feasible category). Based on these results, the application is considered highly feasible as an interactive learning media that effectively supports students in understanding surface area concepts in a more engaging and meaningful way.

PENGEMBANGAN AUGMENTED REALITY BOOK SEBAGAI MEDIA PEMBELAJARAN STRUKTUR DAN FUNGSI BAGIAN TUMBUHAN BUNGA SEMPURNA MENGGUNAKAN METODE MARKER BASED TRACKING

2026-06-14T04:27:39+07:00

Science learning plays an important role in developing students’ understanding of natural phenomena through conceptual mastery and application in everyday life. One science topic that requires strong conceptual and visual understanding is the structure and function of plant parts, particularly complete flowers. However, science learning on this topic still largely relies on textbooks and two-dimensional images as instructional media. These media are not able to clearly present the shape, position, and function of flower parts, causing students to have difficulty visualizing abstract concepts. As a result, students’ understanding of the material is not yet optimal. Therefore, this study aims to develop and implement an Augmented Reality Book (AR Book) using a marker-based tracking method as a learning medium, as well as to determine its feasibility and effectiveness in the learning process. This study employed a Research and Development (R&D) method using the ADDIE model, which consists of analysis, design, development, implementation, and evaluation stages. The research subjects were fourth-grade students of MIN 8 Aceh Besar. Data were collected through validation by media experts and subject-matter experts, student response questionnaires, and analysis of learning media implementation results. The results showed that the developed AR Book obtained a feasibility score of 90% from media experts and 91% from subject-matter experts, both of which fall into the “very good” category. In addition, the results of its implementation with students showed a percentage of 94%, also categorized as very good. Based on these findings, it can be concluded that the marker-based tracking AR Book is feasible and effective as a science learning medium and is able to improve students’ understanding, interest, and engagement in learning the structure and function of complete flower plant parts

PREDIKSI TARGET PENDAPATAN PAJAK DAERAH DI KABUPATEN SUMBAWA MENGGUNAKAN ALGORITMA EXTREME GRADIENT BOOSTING (XGBOOST)

2026-06-14T04:27:39+07:00

Locally Generated Revenue (PAD) is the primary source of funding for local governments, with local taxes being the largest component supporting revenue. In Sumbawa Regency, tax target determination is still based on historical potential and realization, resulting in suboptimal accuracy in determining tax targets. This study aims to develop a prediction model for local tax revenue targets using the XGBoost algorithm. Secondary data was obtained from the Sumbawa Regency Regional Revenue Agency (Bapenda), covering various types of local taxes for the 2021–2025 period. The research method uses the Cross Industry Standard Process for Data Mining (CRISP-DM) framework with stages of business understanding, data understanding, preprocessing, modeling, evaluation, and implementation. The evaluation results show an RMSE value of 743,314,988.84 or 743 million, MAPE of 5.34%, and R² of 0.9845, indicating low prediction errors and the model's ability to understand data patterns well. The model is then implemented into a Flask-based web system to support the data input process, model performance, and generate more accurate and data-based predictions of regional tax revenue targets, and has the potential to become a strategic tool in making decisions about determining regional tax targets.

SISTEM PENDUKUNG KEPUTUSAN PENENTUAN PENERIMA KIP KULIAH STMIK SYAIKH ZAINUDDIN NW

2026-06-16T19:56:46+07:00

KIP Kuliah is a government tuition assistance program for High School graduates who have good academic potential but face economic constraints. STMIK Syaikh Zainuddin NW Anjani faces challenges in the KIP Kuliah recipient selection process due to the large number of applicants and diverse assessment criteria, which risks subjectivity and inaccurate targeting. This study aims to build a Decision Support System (DSS) that can assist the campus in determining KIP Kuliah recipients objectively and efficiently. The method used in this system is the Simple Multi-Attribute Rating Technique (SMART) method, which works by assigning weights to each criterion and calculating the final value based on normalization. The criteria used include parental income, academic/non-academic achievements, home ownership status, parental dependents, and other economic conditions. The result of this research is a web-based application capable of processing applicant data and generating a ranking of potential KIP Kuliah recipients according to predetermined criteria weights. System testing shows that the SMART method is effective in providing transparent and accurate decision recommendations for the management of STMIK Syaikh Zainuddin NW.