eJournal Komunitas Dosen Indonesia
Not a member yet
1406 research outputs found
Sort by
Pengembangan Chatbot Pada Platform Telegram Sebagai Media Informasi Seputar Handphone
Get PDFDi tengah kompleksitas pasar teknologi yang terus berkembang, generasi muda sering menghadapi tantangan dalam menemukan informasi yang relevan, terutama terkait perangkat seluler. Untuk mengatasi masalah ini, penelitian ini bertujuan mengembangkan Chatbot Tanya Phone, sebuah solusi interaktif yang dirancang untuk memberikan informasi spesifikasi, harga, dan ulasan produk kepada pengguna Telegram. Proses pengembangan chatbot ini mencakup analisis menyeluruh terhadap kebutuhan pengguna, perancangan alur percakapan yang intuitif, serta pengembangan berbasis API Telegram untuk memastikan integrasi yang efisien dan responsif.Implementasi sistem diharapkan dapat memberikan respons yang cepat dan akurat, membantu pengguna dalam memahami informasi penting terkait perkembangan teknologi di pasar handphone saat ini. Metode pengujian yang digunakan dalam penelitian ini adalah black box testing, yang bertujuan untuk memastikan bahwa semua fitur chatbot berfungsi sesuai dengan ekspektasi dan memenuhi kebutuhan pengguna. Selain itu, proses pengujian juga mengidentifikasi beberapa aspek yang memerlukan penyempurnaan guna meningkatkan kinerja chatbot secara keseluruhan. Hasil penelitian menunjukkan bahwa Chatbot Tanya Phone tidak hanya mampu memberikan umpan balik secara real-time, tetapi juga meningkatkan pemahaman pengguna terkait teknologi, memudahkan pencarian informasi, serta memberikan kontribusi positif bagi generasi muda dalam menghadapi perkembangan teknologi yang semakin dinamis di era digital saat ini, serta membantu mereka dalam membuat keputusan yang lebih baik dan memperkuat keterampilan literasi digital mereka untuk beradaptasi dengan perubahan teknologi yang cepat.Di tengah kompleksitas pasar teknologi yang terus berkembang, generasi muda sering menghadapi tantangan dalam menemukan informasi yang relevan, terutama terkait perangkat seluler. Untuk mengatasi masalah ini, penelitian ini bertujuan mengembangkan Chatbot Tanya Phone, sebuah solusi interaktif yang dirancang untuk memberikan informasi spesifikasi, harga, dan ulasan produk kepada pengguna Telegram. Proses pengembangan chatbot ini mencakup analisis menyeluruh terhadap kebutuhan pengguna, perancangan alur percakapan yang intuitif, serta pengembangan berbasis API Telegram untuk memastikan integrasi yang efisien dan responsif.Implementasi sistem diharapkan dapat memberikan respons yang cepat dan akurat, membantu pengguna dalam memahami informasi penting terkait perkembangan teknologi di pasar handphone saat ini. Metode pengujian yang digunakan dalam penelitian ini adalah black box testing, yang bertujuan untuk memastikan bahwa semua fitur chatbot berfungsi sesuai dengan ekspektasi dan memenuhi kebutuhan pengguna. Selain itu, proses pengujian juga mengidentifikasi beberapa aspek yang memerlukan penyempurnaan guna meningkatkan kinerja chatbot secara keseluruhan. Hasil penelitian menunjukkan bahwa Chatbot Tanya Phone tidak hanya mampu memberikan umpan balik secara real-time, tetapi juga meningkatkan pemahaman pengguna terkait teknologi, memudahkan pencarian informasi, serta memberikan kontribusi positif bagi generasi muda dalam menghadapi perkembangan teknologi yang semakin dinamis di era digital saat ini, serta membantu mereka dalam membuat keputusan yang lebih baik dan memperkuat keterampilan literasi digital mereka untuk beradaptasi dengan perubahan teknologi yang cepat
The Concept of Justice in AI-Driven Legal Decision Making
Get PDFThe integration of Artificial Intelligence (AI) into legal decision-making processes has introduced significant advancements in efficiency and predictive capability. However, its implications for justice—particularly fairness, impartiality, transparency, and due process—remain critically debated. This study employs a Systematic Literature Review (SLR) methodology to examine how AI-driven legal decision-making aligns with classical and contemporary philosophical concepts of justice. Drawing on 48 peer-reviewed articles, policy documents, and case studies published between 2015 and 2024, the research identifies four core thematic issues: the persistence of algorithmic bias, the lack of transparency in AI systems, inconsistencies in global regulatory frameworks, and the misalignment of AI logic with moral reasoning. While AI offers promising tools for streamlining judicial processes, its application often risks reinforcing existing inequities and undermining legal principles such as corrective justice and procedural fairness. The study concludes with targeted recommendations for the development of transparent, accountable, and ethically governed AI systems that support—rather than supplant—human judicial discretion. This research contributes to the growing discourse on legal AI by highlighting the necessity of embedding justice-oriented values at the core of technological innovation in the legal sector. This research has several limitations: not based on empirical findings and no validations from experts both in AI and in legal theories. Future research should address these limitations
UTAUT-Based Analysis of Customer Loyalty in Tokopedia Using SOR Framework
Get PDFThis study investigates customer loyalty in the context of e-commerce by applying an integrated Unified Theory of Acceptance and Use of Technology (UTAUT) model combined with the Stimulus–Organism–Response (SOR) framework. Tokopedia, one of Indonesia’s leading e-commerce platforms, faces increasing difficulty in sustaining customer loyalty due to common service-related issues such as delayed deliveries, system errors during voucher usage, and slow customer service. These challenges negatively impact user experience and trust key elements in fostering long-term loyalty. To address these issues, this research explores the need for an integrated approach by employing the Unified Theory of Acceptance and Use of Technology (UTAUT) and the Stimulus–Organism–Response (SOR) framework. The UTAUT model provides insights into the factors driving user adoption, while the SOR framework helps in understanding the psychological responses that lead to sustained loyalty. This research explores how five UTAUT constructs Performance Expectancy, Effort Expectancy, Social Influence, Facilitating Conditions, and Price Value serve as external stimuli (Stimulus) that influence users’ trust in the Tokopedia application (Organism), which then drives customer loyalty (Response). A quantitative survey of 384 Tokopedia users was analyzed using PLS-SEM. The findings reveal that all UTAUT variables significantly affect trust, and trust serves as a strong predictor of customer loyalty. This study contributes theoretically by bridging UTAUT and SOR models to explain loyalty behavior in e-commerce settings, and offers practical insights for Tokopedia to improve user retention. Notably, future studies may consider longitudinal approaches and explore additional psychological factors beyond trust to deepen understanding of loyalty formation
Performance Comparison of NGINX, Apache, and Lighttpd Using WRK on a Debian
Get PDFA web server plays a crucial role in delivering HTTP-based services that are essential for web applications and information systems. In environments with limited resources, such as Virtual Private Servers (VPS), selecting the most efficient web server becomes a key consideration to ensure system stability and responsiveness. This study addresses the performance comparison of three widely used web servers—Apache, NGINX, and Lighttpd—under various load conditions, with the aim of identifying which server provides the best performance in constrained environments. The experimental method was applied using the WRK benchmarking tool, which simulates concurrent user connections under four scenarios combining different user counts (50 and 100 users) and durations (1 and 3 seconds). The performance metrics evaluated include latency, requests per second (RPS), and transfer rate. Results showed that Apache consistently offered the best overall performance with low latency and high RPS, making it a reliable and stable option. NGINX demonstrated strong performance in handling concurrent connections and showed the highest transfer rate, although it requires more complex configuration. Lighttpd, while lightweight and efficient in resource usage, exhibited weaker performance in high-load scenarios. This research contributes to the current body of literature by offering up-to-date and practical benchmarking data. The findings serve as a useful reference for developers and system administrators in selecting a web server based on specific operational needs, resource availability, and performance priorities in modern VPS-based infrastructures.A web server plays a crucial role in delivering HTTP-based services that are essential for web applications and information systems. In environments with limited resources, such as Virtual Private Servers (VPS), selecting the most efficient web server becomes a key consideration to ensure system stability and responsiveness. This study addresses the performance comparison of three widely used web servers—Apache, NGINX, and Lighttpd—under various load conditions, with the aim of identifying which server provides the best performance in constrained environments. The experimental method was applied using the WRK benchmarking tool, which simulates concurrent user connections under four scenarios combining different user counts (50 and 100 users) and durations (1 and 3 seconds). The performance metrics evaluated include latency, requests per second (RPS), and transfer rate. Results showed that Apache consistently offered the best overall performance with low latency and high RPS, making it a reliable and stable option. NGINX demonstrated strong performance in handling concurrent connections and showed the highest transfer rate, although it requires more complex configuration. Lighttpd, while lightweight and efficient in resource usage, exhibited weaker performance in high-load scenarios. This research contributes to the current body of literature by offering up-to-date and practical benchmarking data. The findings serve as a useful reference for developers and system administrators in selecting a web server based on specific operational needs, resource availability, and performance priorities in modern VPS-based infrastructures
Blended Learning in Higher Education for Informatics Engineering Education: A Bibliometric and Systematic Literature Review
Get PDFThe evolution of digital technology has prompted a significant transformation in higher education, particularly within the Informatics Engineering curriculum. In this landscape, blended learning has surfaced as a vital strategic method, combining traditional face-to-face teaching with online learning to boost flexibility, effectiveness, and student engagement. This research utilizes a Systematic Literature Review (SLR) and bibliometric analysis to map the trends and research focus of blended learning. Data was gathered from Google Scholar for publications from 2021 onward, using specific keywords. An initial pool of 1,850 articles was refined through a PRISMA-based screening process, resulting in 33 highly relevant articles for detailed analysis. The findings show that publication activity was highest in 2021, with major contributions from the United Kingdom, Indonesia, and the United States. Key themes identified in the literature include the use of Learning Management Systems (LMS), the flipped classroom model, and project-based learning. The evidence consistently suggests that blended learning improves educational outcomes, increases student motivation, and fosters essential 21st-century skills. However, challenges such as insufficient infrastructure and the need for enhanced educator competency remain. This paper recommends strengthening institutional policies and providing faculty training to support the successful and sustainable implementation of blended learning, especially in technology and vocational education.The evolution of digital technology has prompted a significant transformation in higher education, particularly within the Informatics Engineering curriculum. In this landscape, blended learning has surfaced as a vital strategic method, combining traditional face-to-face teaching with online learning to boost flexibility, effectiveness, and student engagement. This research utilizes a Systematic Literature Review (SLR) and bibliometric analysis to map the trends and research focus of blended learning. Data was gathered from Google Scholar for publications from 2021 onward, using specific keywords. An initial pool of 1,850 articles was refined through a PRISMA-based screening process, resulting in 33 highly relevant articles for detailed analysis. The findings show that publication activity was highest in 2021, with major contributions from the United Kingdom, Indonesia, and the United States. Key themes identified in the literature include the use of Learning Management Systems (LMS), the flipped classroom model, and project-based learning. The evidence consistently suggests that blended learning improves educational outcomes, increases student motivation, and fosters essential 21st-century skills. However, challenges such as insufficient infrastructure and the need for enhanced educator competency remain. This paper recommends strengthening institutional policies and providing faculty training to support the successful and sustainable implementation of blended learning, especially in technology and vocational education
Development of a Web-Based Interactive E-Learning Platform for a Vocational High School
Get PDFDigital transformation is essential in modern education, particularly at SMK Bina Am Ma’mur, where traditional learning methods, such as printed materials and the WhatsApp application, are still primarily used. These conventional methods limit flexibility, interactivity, and student engagement, which are crucial for effective learning. To address these challenges, this study develops an interactive web-based e-learning platform designed to enhance the learning experience at SMK Bina Am Ma’mur. The research employs a Research & Development (R&D) methodology, guided by the ADDIE model, which includes five phases: Analyze, Design, Development, Implementation, and Evaluation. Data were collected through literature reviews, observations, interviews with teachers, expert validation, and student feedback, followed by descriptive analysis to interpret the results. The validation process revealed that the developed platform is highly feasible for educational use. Media experts rated it 95% suitable, while subject matter experts provided a 91% rating. Additionally, the platform received an 89% approval rating from students, indicating its effectiveness in improving material comprehension and engagement. These findings suggest that the interactive e-learning platform is a highly effective tool for enhancing the learning process at SMK Bina Am Ma’mur, making learning more flexible, accessible, and engaging. This innovative platform has the potential to overcome the limitations of traditional media, fostering greater student motivation, interactivity, and overall learning outcomes, particularly in vocational education
Smart Diagnostic Assistant for Peugeot 406: A Web Expert System Based on Production Rules
Get PDFThis research presents the development of a web-based expert system for diagnosing engine faults in Peugeot 406 vehicles, addressing the challenge of limited access to model-specific diagnostic tools particularly in remote or underserved regions where authorized service centers are scarce and the resulting difficulty in early identification of mechanical issues. The system employs a forward-chaining inference method coupled with a depth-first search (DFS) algorithm for rule traversal within a structured knowledge base of 25 engine symptoms and 13 fault types, all formulated through expert interviews and literature review. Through IF THEN production rules, users input observable symptoms via an intuitive, non-technical interface to receive preliminary diagnoses. System testing encompassed simulated case studies covering the full spectrum of defined fault types as well as real-world trials with Peugeot 406 owners in workshop settings, demonstrating over 85% consistency with professional mechanic evaluations. These findings underscore the system’s potential to improve diagnostic efficiency, reduce repair costs, and extend vehicle longevity by enabling timely, data-driven interventions. The implementation of this rule-based expert system thus offers an effective solution for empowering vehicle owners in areas with restricted service access, and future enhancements such as integration of real-time OBD and IoT sensor data, multilingual support, and adaptive machine-learning rule updates are expected to further boost accuracy, flexibility, and user reach
Sentiment Classification of Customer Reviews in the Fast-Food Industry Using the Naïve Bayes Algorithm
Get PDFIn the digital era, online reviews have become a significant source of information, influencing consumer perceptions and purchasing decisions, particularly in the fast-food industry. This research focuses on classifying customer sentiment towards A&W restaurants based on online reviews using the Naïve Bayes algorithm. The objective of this study is to analyze customer feedback to understand their perceptions of A&W’s services and products. The research follows the Cross Industry Standard Process for Data Mining (CRISP-DM) methodology, which involves six phases: business understanding, data understanding, data preparation, modeling, evaluation, and deployment. Data was collected from Google Reviews of the A&W Palem Semi branch, consisting of 200 customer reviews, which were preprocessed to remove irrelevant content and prepare the data for analysis. The Naïve Bayes algorithm was applied to classify the sentiments into three categories: positive, negative, and neutral. The model achieved an overall accuracy of 83%. However, the results revealed a significant class imbalance, with most reviews labeled as neutral. While the model performed well in identifying neutral sentiment (precision 0.89, recall 0.97, F1-score 0.93), it failed to classify positive and negative sentiments accurately, as both achieved precision, recall, and F1-scores of 0.00. This demonstrates that the data imbalance severely impacted the model’s ability to detect minority sentiment classes. The research concludes that while Naïve Bayes offers useful insights into customer sentiment, improvements are necessary, including applying data balancing techniques or exploring alternative algorithms such as SVM or Random Forest to enhance classification performance across all sentiment categories
Naïve Bayes Algorithm Analysis For Student Graduation Timeliness Prediction
Get PDFThis study developed a student graduation prediction system using the Naïve Bayes algorithm, using PS1-PS4 scores, PK, and SKS as indicators of academic progress. This model achieved 88.33% accuracy and an ROC value of 0.900, indicating superior predictive ability. These results outperform other common models such as logistic regression and the C4.5 decision tree, which have approximately 85% accuracy in predicting student graduation. These results also outperform previous research in the same field, which had ROC values of approximately 0.85.Graduation predictions were categorized as "ON TIME" and "LATE" with high precision. The Naïve Bayes algorithm has proven effective in predicting student graduation, particularly in identifying factors that influence graduation timeliness, such as poor academic performance, difficulty completing final assignments, poor personal conditions, and lack of motivation and interest.By designing a graduation prediction system using the Naïve Bayes algorithm, this research aims to help educational institutions predict student graduation timeliness and provide appropriate interventions. This system can improve educational quality and reduce dropout rates, making it an important tool for educational institutions to improve graduate quality and achieve their academic goals.This research demonstrates that the Naïve Bayes algorithm can be an effective and accurate graduation prediction method, thus helping educational institutions develop strategies to improve educational quality and reduce dropout rates. Therefore, this research has the potential to significantly impact higher education institutions and assist them in achieving their academic goals
Optimizing Book Genre Classification through AI on a Web Platform
Get PDFIn the rapidly evolving digital era, the exponential growth of online book collections poses challenges in efficiently classifying literature according to genre. Manual classification methods are often time-consuming, subjective, and inconsistent, necessitating the adoption of advanced, automated approaches. This study aims to develop and implement an Artificial Intelligence (AI)-based genre classification system integrated into a web platform to enhance the accuracy, efficiency, and user experience in book discovery. Leveraging Machine Learning (ML) algorithms—particularly Support Vector Machine (SVM), Naïve Bayes, Decision Tree, Random Forest, and Deep Learning—alongside Natural Language Processing (NLP) techniques such as tokenization, stemming, and TF-IDF, the system analyzes book descriptions and synopses to determine the most appropriate genre. The research follows a qualitative and literature study approach, utilizing a dataset sourced from Kaggle, with preprocessing steps to remove noise and convert text into numerical representations. Experimental results demonstrate that the SVM model achieved the highest accuracy, precision, recall, and F1-score compared to other tested algorithms, effectively handling high-dimensional and non-linear data. The developed web application features an interactive dashboard, real-time classification, and a hybrid recommendation system. This work confirms the feasibility and advantages of AI-driven genre classification for large-scale digital libraries and online bookstores. While limitations such as data imbalance and overlapping genre semantics remain, the findings provide a strong foundation for future research employing larger, more diverse datasets and advanced deep learning architectures to further improve classification performance.In the rapidly evolving digital era, the exponential growth of online book collections poses challenges in efficiently classifying literature according to genre. Manual classification methods are often time-consuming, subjective, and inconsistent, necessitating the adoption of advanced, automated approaches. This study aims to develop and implement an Artificial Intelligence (AI)-based genre classification system integrated into a web platform to enhance the accuracy, efficiency, and user experience in book discovery. Leveraging Machine Learning (ML) algorithms—particularly Support Vector Machine (SVM), Naïve Bayes, Decision Tree, Random Forest, and Deep Learning—alongside Natural Language Processing (NLP) techniques such as tokenization, stemming, and TF-IDF, the system analyzes book descriptions and synopses to determine the most appropriate genre. The research follows a qualitative and literature study approach, utilizing a dataset sourced from Kaggle, with preprocessing steps to remove noise and convert text into numerical representations. Experimental results demonstrate that the SVM model achieved the highest accuracy, precision, recall, and F1-score compared to other tested algorithms, effectively handling high-dimensional and non-linear data. The developed web application features an interactive dashboard, real-time classification, and a hybrid recommendation system. This work confirms the feasibility and advantages of AI-driven genre classification for large-scale digital libraries and online bookstores. While limitations such as data imbalance and overlapping genre semantics remain, the findings provide a strong foundation for future research employing larger, more diverse datasets and advanced deep learning architectures to further improve classification performance