Jurnal Teknologi Informasi dan Pendidikan

The Role of Digital Platforms and Online User Interaction in Shaping Tourist Emotional Satisfaction

2026-01-23T07:49:02+00:00

Tourist experiences in the digital era are shaped not only by on-site physical interactions but are increasingly influenced by digital platforms, online reviews, and social media. Although the literature on digital tourism is growing rapidly, a gap remains in understanding how digital interactions holistically and integratively influence tourists' emotional satisfaction, particularly in coastal destinations. This study aims to examine the role of digital and information technology-based factors in shaping tourists' emotional satisfaction in coastal destinations by developing an integrated model that links the physical and social environments, digital experience quality, electronic word of mouth (e-WOM), digital destination image, and tourist engagement. Data were collected from 339 tourists who actively used digital platforms during their visits to a coastal destination in Indonesia. A mixed-method analytical approach was employed, combining Partial Least Squares Structural Equation Modeling (PLS-SEM) to test linear relationships and fuzzy-set Qualitative Comparative Analysis (fsQCA) to uncover complex configurational pathways leading to high emotional satisfaction. The findings indicate that digital experience quality and e-WOM significantly strengthen digital destination image, which in turn enhances tourist engagement and emotional satisfaction. Tourist engagement emerged as the strongest direct predictor of emotional satisfaction. The configurational analysis revealed multiple alternative pathways leading to high satisfaction, with digital destination image identified as a necessary condition across all configurations. This study offers theoretical contributions by integrating environmental psychology and digital tourism literature into a comprehensive model and by applying a mixed-method approach (SEM-fsQCA) capable of capturing causal complexity. Practically, the findings underscore the importance for destination managers to not only manage physical attractions but also to proactively build a strong digital destination image, encourage positive e-WOM, and create interactive and engaging digital experiences to enhance tourists' emotional satisfaction.

Lightweight YOLO Models for Robust Facial Expression Detection

2026-03-31T05:47:45+00:00

Facial expression recognition is a fundamental component of artificial intelligence systems, particularly in human–machine interaction. However, achieving robust detection accuracy remains challenging due to variations in lighting, facial orientation, and limited training data diversity. While recent lightweight YOLO architectures—YOLOv8n, YOLOv10n, and YOLO11n—have demonstrated strong performance in general object detection, comparative studies evaluating these models specifically for facial expression detection remain limited. This study addresses this gap by systematically comparing these three nano-variant models on a dataset of 2,000 labeled facial images across four expression categories: flat face, angry, sad, and smile. The dataset was divided into training (70%), validation (20%), and test (10%) subsets. Experiments were conducted under two scenarios—with and without data augmentation—using identical training configurations. Augmentation techniques included mosaic composition, HSV variation, geometric transformations, and flipping. Results show that augmentation improved the F1 score of YOLOv10n from 0.68 to 0.72 and YOLO11n from 0.65 to 0.72, with the latter achieving the highest overall precision of 0.82. YOLOv8n exhibited stable performance with an F1 score of 0.75 under both conditions. Confidence threshold optimization revealed distinct optimal operating points for each model, ranging from 0.1 to 0.6, confirming that per-model threshold tuning is necessary to maximize detection performance. These findings provide practical guidance for selecting and configuring lightweight YOLO models for facial expression detection in resource-constrained environments.

Design of Mobile Learning Application Interface Using Kansei Engineering Method (Case Study: Majelis Daur Ulang)

2026-03-07T08:30:47+00:00

Kansei Engineering is a method for realizing certain product designs based on a systematic exploration of human feelings and sensations (sight, touch, smell, hearing, taste). In mobile applications, the most important factor apart from the technical aspects of the system is the design. Design is an important factor in a mobile application because it becomes a liaison between the user and the existing system. The Mobile Learning application is not only enough to run the application and there are no errors, but the application must be built according to the wishes and interests of the user, this study aims to determine the emotional factors of the user, apply Kansei Engineering in designing and make recommendations for application display design according to Kansei Engineering. This methodology refers to Kansei Engineering Type I. This study uses Kansei Word to detect the user's feelings when looking at the specimen design. The list of Kansei Words used is 10 words related to the Majelis Daur Ulang mobile learning application. There are 5 specimens of similar Mobile Learning applications used. This study involved 32 participants, using multivariate statistical analysis, namely Cronbach's Alpha (CA), Correlation Coefficient Analysis (CCA), Principal Component Analysis (PCA), Factor Analysis (FA) and Partial Least Square (PLS). This research resulted in 2 recommendations for the design Majelis Daur Ulang Mobile Learning display, namely "Professional" and "Unique”.

Analysis and Prediction of Land Cover Change in Palangka Raya City Using a Cellular Automata–Neural Network Model Based on MOLUSCE

2026-03-10T07:38:02+00:00

Land cover change represents a dynamic phenomenon driven by human activities and rapid regional growth, particularly in Palangka Raya City. This study aimed to analyze the historical dynamics of land cover changes during the 2016–2020 period and predict development trends for 2028 and 2040. The methodology integrated Cellular Automata and Artificial Neural Network (CA-ANN) models utilizing the MOLUSCE plugin within Geographic Information System software. Several driving factors were incorporated into the modeling process, including distance from road networks, distance from rivers, slope, elevation, and population density. The analysis revealed a significant transition from natural vegetated areas, such as peat swamp forests and shrublands, into anthropogenic land uses, specifically oil palm plantations and built-up areas. Model validation was performed using the Kappa coefficient test, which yielded a high level of accuracy, thereby confirming the reliability of the model for spatial projection purposes. The prediction results for 2028 and 2040 provided critical spatial insights regarding the potential continuous expansion of built-up areas. These findings were intended to serve as a crucial reference for local governments in formulating sustainable spatial planning policies to mitigate future environmental degradation.

Blockchain Technology in Donation-Based Crowdfunding Systems

2025-09-02T08:26:51+00:00

This study successfully designed and developed a blockchain-based crowdfunding system using Base, implementing secure authentication through wallet addresses as the primary user identity. The system supports key features such as campaign creation, donation, and fund withdrawal. Campaign creation allows users to set fundraising targets and deadlines, donations are recorded directly on the blockchain to ensure transparency, and fund withdrawals can only be performed by the campaign owner once specific conditions are met. The implementation of smart contracts played a vital role in ensuring that all transactions were transparent, secure, and accountable. Furthermore, the system effectively eliminated the need for intermediaries, as typically seen in conventional platforms, by replacing the role of fund custodians with smart contracts. As a result, it no longer required high administrative fees, instead relying on a fixed and significantly lower gas fee. This approach enhanced public trust, reduced operational costs, and strengthened accountability in the management of donations through a decentralized and tamper-resistant system.

The Effectiveness of Digital Flipbook Media in Vocational Engineering Drawing Education

2026-03-07T08:37:16+00:00

This research is motivated by the low learning outcomes of students in the Technical Drawing subject at Vocational Schools due to the dominance of conventional methods that are less able to visualize objects comprehensively. The purpose of this study is to analyze the effect of using digital flipbook learning media on the learning outcomes of class X students of Welding Engineering Expertise Concentration at SMK Negeri 10 Malang. The research method used is quantitative with a One Group Pretest-Posttest design on 70 students selected through a total sampling technique. The research instrument is a validated multiple-choice objective test, and the data are analyzed using the Paired Sample T-Test and N-Gain tests. The results of the study showed a significant increase in the average value from 79.91 to 90.89 with a significance value of 0.000 (p < 0.05). The effectiveness of media use is in the "Medium" category with an average N-Gain of 0.5611. The conclusion of this study states that digital flipbook media has a real influence in improving students' conceptual understanding and cognitive achievement in technical drawing competencies. This media is recommended as a strategic tool to support innovative digital learning transformation in the Independent Curriculum era.

Design and Development of a Web-Based Educational Chatbot Using Natural Language Processing for Public Information Services at SMK Negeri 2 Padang

2025-09-02T08:32:19+00:00

This study addresses the limited efficiency of public information services in vocational schools, which often results in delayed responses and repetitive administrative workload. This research aims to design and develop a web-based educational chatbot using Natural Language Processing (NLP) to improve information accessibility at SMK Negeri 2 Padang. The system was developed using the Waterfall model and implements text preprocessing, TF-IDF vectorization, and cosine similarity for intent recognition. System evaluation was conducted through Black Box Testing and accuracy measurement based on user queries. The results show that the system achieved a 100% success rate in functional testing and 91% accuracy in intent classification, indicating its effectiveness in providing relevant and real-time information. This study contributes by offering a practical, scalable, and user-friendly NLP-based solution to enhance public information services in educational institutions.

Comparative Performance of Indonesian Stemming: PL/SQL Implementation of Nazief and Adriani Algorithm versus Sastrawi Library

2026-02-23T06:09:35+00:00

Text preprocessing in Indonesian applications commonly relied on external libraries such as Sastrawi. However, performing this task outside the database layer often introduced significant latency due to data communication overhead between the application and the server. This study proposed and evaluated a native stemming mechanism utilizing the Nazief and Adriani algorithm implemented directly within an Oracle PL/SQL environment. The primary objective was to determine whether in-database processing could offer better performance than the standard application-layer approach. The assessment compared the PL/SQL implementation against the Python-based Sastrawi library using a comprehensive dataset of 54,715 words sourced from the Kamus Besar Bahasa Indonesia (KBBI). Performance metrics focused on stemming accuracy and total execution time. The empirical results revealed that the proposed PL/SQL method achieved an accuracy of 96.82%, which proved slightly superior to the 96.58% accuracy obtained by Sastrawi. Furthermore, the stored procedure implementation demonstrated significant efficiency, completing the process in 602.22 seconds, whereas the baseline method required 1,259.28 seconds. It was concluded that migrating the stemming logic into the database layer effectively reduced execution time by approximately 52.18% while maintaining high precision. These findings suggested that native database implementation provided a more robust solution for systems requiring high-performance text processing.

Hyperparameter Tuning of YOLOv8n for Real-Time Material Truck Detection

2026-04-16T06:11:23+00:00

The increasing number of material trucks on arterial roads has posed challenges for traffic surveillance and regulatory compliance. Traditional monitoring techniques that rely on manual observation are often ineffective and susceptible to irregularities, highlighting the need for automated real-time monitoring systems. This study proposes a lightweight object detection approach using YOLOv8n to improve real-time truck detection performance in traffic monitoring applications. A quantitative experimental methodology was employed by performing hyperparameter tuning through adjustments to the number of epochs, batch size, optimizer, and learning rate. The dataset was collected from real traffic environments using smartphone cameras and CCTV (TP-Link Tapo C320WS). A total of 36 experimental configurations were evaluated using Precision, Recall, F1-score, mAP@50, and mAP@50–95 metrics. Experimental results showed that the optimal configuration, consisting of 100 epochs, a batch size of 16, the Adam optimizer, and a learning rate of 0.001, achieved a mean Average Precision (mAP)@50 of 0.9302 and mAP@50–95 of 0.7226. Although the performance improvement over the baseline YOLOv8n model was relatively modest, repeated experiments demonstrated improved model stability and consistency after hyperparameter optimization. Real-time deployment on a local GPU achieved a stable processing speed of 14–23 Frames Per Second, with an average of 19 FPS, enabling real-time monitoring performance aligned with the camera input rate. The integrated system successfully combines object detection, tracking, and license plate recognition for practical traffic monitoring applications. However, smaller objects such as license plates remained more challenging to detect due to localization limitations under occlusion and low-light conditions.

Hyperparameter Tuning Strategy for YOLOv8n in Real-Time Post-Accident Traffic Monitoring

2026-04-15T05:13:57+00:00

Traffic accidents continue to provide a considerable difficulty in contemporary transportation systems, frequently leading to vehicle damage and heightened risks for pedestrians on streets. Precise and instantaneous identification of post-accident scenarios is thus crucial for facilitating swift response and sophisticated traffic management. This research introduces a streamlined object detection methodology utilizing YOLOv8n to recognize six essential traffic-related categories: bus, automobile, damaged vehicle, motorbike, pedestrian, and truck. The main aim is to examine the impact of hyperparameter modification on detection efficacy, specifically in recognizing damaged automobiles as signs of post-accident situations. Twelve model configurations were created by systematically altering three hyperparameters: learning rate (0.01, 0.001, and 0.0001), batch size (32 and 64), and optimizer type (Adam and MuSGD). All models underwent training for 200 epochs with a dataset derived from actual traffic situations, augmented by techniques such as grayscale transformation, blurring, and rotation. The performance evaluation utilized precision, recall, F1-score, mAP50, and mAP50:95. The findings indicate that hyperparameter selection substantially influences convergence stability and detection accuracy. The optimal model attained a mAP50 of 0.905 and a mAP50:95 of 0.751, utilizing a learning rate of 0.01, a batch size of 64, and the Adam optimizer. Moreover, substantial items like cars, buses, and trucks were identified with high precision, whereas damaged vehicles and pedestrians necessitated more meticulous calibration due to increased visual variability.The findings indicate that optimized lightweight models can attain competitive performance, rendering them appropriate for real-time intelligent traffic monitoring applications.

Analysis of Provocative Speech During the 2025 DPR Demonstration on X Using the IndoBERTweet Method

2026-04-15T06:56:24+00:00

Social media platforms have become important channels for public discussion during political events. During the DPR demonstrations in August 2025, online discussions on X (formerly Twitter) contained various forms of expressions, including provocative speech that may influence public opinion and collective behavior. Detecting such content automatically is challenging due to the informal language, slang, and contextual nuances commonly found in social media texts. This study aims to analyze provocative speech on the social media platform X using text classification techniques and transformer-based models. A total of 8,899 Indonesian tweets related to the demonstration period from August 25 to August 31, 2025 was collected using the Tweet Harvest crawling tool. The dataset was manually labeled into two categories, namely provocative and non-provocative, using a majority voting approach by three annotators. Several preprocessing steps were applied, including cleaning, normalization, stemming, tokenization, and stopword removal. Several models were evaluated, including Multinomial Naïve Bayes, Linear Support Vector Machine, BiLSTM, IndoBERT, and IndoBERTweet. Experimental results show that transformer-based models outperform traditional machine learning approaches. The best performance was achieved by the IndoBERTweet model with a learning rate of 3×10⁻⁵, achieving an accuracy of 93.07% and an F1-score of 91.56%. These findings indicate that domain-specific language models are effective for detecting provocative speech in Indonesian social media discussions related to political events.