Data-Driven Artificial Intelligence in Education: A Comprehensive Review

IEEE Transactions on Learning Technologies, Journal Year: 2023, Volume and Issue: 17, P. 12 - 31

Published: Sept. 12, 2023

As Education constitutes an essential development standard for individuals and societies, researchers have been exploring the use of Artificial Intelligence (AI) in this domain embedded technology within it through a myriad applications. In order to provide detailed overview efforts, article pays particular attention these developments by highlighting key application areas data-driven AI Education; also analyzes existing tools, research trends, as well limitations role can play Education. particular, reviews various applications including student grading assessments, retention drop-out predictions, sentiment analysis, intelligent tutoring, classroom monitoring, recommender systems. The provides bibliometric analysis highlight salient trends over nine years (2014–2022) further description tools platforms developed outcome efforts For articles from several top venues are analyzed explore domain. shows sufficient contribution different parts world with clear lead United States. Moreover, students' evaluation observed most widely explored application. Despite significant success, we aspects education where alone has not contributed much. We believe such is expected baseline future

Language: Английский

---

A pilot study of measuring emotional response and perception of LLM-generated questionnaire and human-generated questionnaires

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Feb. 2, 2024

Abstract The advent of ChatGPT has sparked a heated debate surrounding natural language processing technology and AI-powered chatbots, leading to extensive research applications across various disciplines. This pilot study aims investigate the impact on users' experiences by administering two distinct questionnaires, one generated humans other ChatGPT, along with an Emotion Detecting Model. A total 14 participants (7 female 7 male) aged between 18 35 years were recruited, resulting in collection 8672 ChatGPT-associated data points 8797 human-associated points. Data analysis was conducted using Analysis Variance (ANOVA). results indicate that utilization enhances participants' happiness levels reduces their sadness levels. While no significant gender influences observed, variations found about specific emotions. It is important note limited sample size, narrow age range, potential cultural impacts restrict generalizability findings broader population. Future directions should explore incorporating additional models or chatbots user emotions, particularly among groups such as older individuals teenagers. As pioneering works evaluating human perception text communication, it noteworthy received positive evaluations demonstrated effectiveness generating questionnaires.

Language: Английский

---

Student Recognition and Activity Monitoring in E-Classes Using Deep Learning in Higher Education

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 66110 - 66128

Published: Jan. 1, 2024

Monitoring student activity manually constantly is a laborious endeavor. Over the past few years, there has been rapid expansion in usage of cameras and automatic identification odd surveillance behavior. Different computer vision algorithms have used to observe monitor real-world activities. Most educational institutions are already offering online programs lessen impact this epidemic on education industry. However, ensuring that students correctly identified, their behaviors monitored crucial make these learning sessions dynamic equivalent conventional offline classroom. In study, we introduced brand-new deep learning-based continuously track student's mood, including rage, contempt, happiness, sorrow, fear, surprise. The effectiveness monitoring classrooms was also studied using CNN model reaches 99% accuracy. Our approach superior because its many convolutional layers, dropout regularization, batch normalization. It caught properties decreased overfitting. By identifying them more frequently, techniques can enhance engagement outcomes e-learning situations, according research. With techniques, educators instructors may support effectively by better comprehending behavior specialized individualized support, improving academic performance evaluation.

Language: Английский

---

Measuring student attention based on EEG brain signals using deep reinforcement learning

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 126426 - 126426

Published: Jan. 1, 2025

Language: Английский

---

Cultivando la Inteligencia Emocional en la Era de la Inteligencia Artificial: Promoviendo la Educación Centrada en el Ser Humano

Estudios y Perspectivas Revista Científica y Académica, Journal Year: 2024, Volume and Issue: 4(2), P. 16 - 30

Published: May 16, 2024

Este artículo aborda la profundización de Inteligencia Emocional en Era los avances tecnológicos relación con promoción Educación Centrada el Ser Humano Colombia IA. El objetivo este se centra ahora comprensión floración dominio las TIC relacionadas emociones y uso posterior investigación. La recolección datos llevó a cabo través del análisis documental mapeo IA implicaciones desarrollo modelos pensamiento complejo dentro ajustes significativos campo educación. estudio engloba una investigación dirigida esclarecer estrechas relaciones entre conexiones emocionales preocupaciones inteligencia artificial. Los resultados iniciales revelaron que aprendizaje emocional podría abordarse mediante inclusión pedagógica tecnología. En conclusión, aumento tecnológico emergente fomenta reconocimiento seres humanos partir interacciones asertivas.

---

The analysis of educational informatization management learning model under the internet of things and artificial intelligence

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Aug. 1, 2024

This study explores the influence of Internet Things (IoT) and Artificial Intelligence (AI)-enhanced learning models on student management in educational informatization management. A game-theoretic enhanced model is proposed to achieve this objective, incorporating resource scheduling strategies under fog computing a system that integrates IoT AI technologies. model's performance are then tested. The results indicate computing-based hierarchical Q-learning (Q) achieves faster convergence than single Q model, reaching after 80 training rounds, ten rounds earlier comparative algorithm. exhibits lower average workload delay 0.5 ms node below 1 ms, showcasing significant advantages terms overall cost-effectiveness, thus minimizing service costs. has 3000 concurrent user connections, static page request times ranging from 0 25 s, login response time predominantly at 60 capacity process up 20 parallel tasks per second with zero errors. functionalities fully realized, meeting usage demands effectively achieving highest functional score 9.03 for online interaction functionality. demonstrates efficacy environment positive impact technologies better caters individual needs, enhancing outcomes experiences. study's innovation lies integration technology AI-enhanced models, coupled introduction strategies, enabling intelligently identify requirements, allocate resources, dynamically optimize process, ultimately improving outcomes. holds implications education quality promoting personalized development.

Language: Английский

---

The Integration of Advanced AI-Enabled Emotion Detection and Adaptive Learning Systems for Improved Emotional Regulation

Journal of Educational Computing Research, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 7, 2024

This study explores the integration of advanced AI technologies, including emotion detection and adaptive learning systems, to enhance second language acquisition among 274 English as a Foreign Language (EFL) learners. Utilizing pretest-posttest randomized control trial, research evaluates effects AI-enhanced interventions on emotional self-regulation linguistic proficiency compared traditional teaching methods. The results indicate significant improvements in retention regulation for learners using tools. Qualitative feedback from interviews surveys corroborates these findings, underscoring positive impact educational experiences. highlights potential deepen engagement tailor experiences, recommending incorporation technologies into programs boost competencies enrich outcomes.

Language: Английский

---

Navigating the YOLO Landscape: A Comparative Study of Object Detection Models for Emotion Recognition

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 109427 - 109442

Published: Jan. 1, 2024

The You Only Look Once (YOLO) series, renowned for its efficiency and versatility in object detection, has become a fundamental component diverse fields ranging from autonomous vehicles to robotics video surveillance. Despite widespread application, notable gap exists the literature concerning selecting YOLO models specific tasks. Current trends often lean towards latest models, potentially overlooking crucial factors such as computational complexity, speed, accuracy, model size, adaptability, generalization. This approach may not always yield optimal choice given application. Therefore, this paper aims provide an exhaustive comparative analysis of various focusing on emotion recognition. We trained tested YOLOv5, YOLOv7, YOLOv8, YOLOv9 along with their respective variants, using subset AffectNet dataset, which consists facial images annotated one five emotions, namely angry, happy, sad, neutral, surprise. study evaluates based several key parameters: accuracy metrics like mean Average Precision (mAP), inference time, FPS, adaptability altered datasets, generalization capability. Comprehensive results are presented, highlighting strengths limitations each variant across these parameters. Insights provided guide researchers most suitable architecture recognition requirements, considering constraints, real-time performance needs, importance vs tradeoffs. reveals exceptional performances certain YOLOv9e high YOLOv8n balancing speed accuracy. Overall, work fills by offering detailed facilitate informed decision-making when deploying

Language: Английский

---

Student Classroom Behavior Detection Based on YOLOv7+BRA and Multi-model Fusion

Lecture notes in computer science, Journal Year: 2023, Volume and Issue: unknown, P. 41 - 52

Published: Jan. 1, 2023

Language: Английский

---

Classroom Behavior Recognition Using Computer Vision: A Systematic Review

Sensors, Journal Year: 2025, Volume and Issue: 25(2), P. 373 - 373

Published: Jan. 10, 2025

Behavioral computing based on visual cues has become increasingly important, as it can capture and annotate teachers' students' classroom states a large scale in real time. However, there is lack of consensus the research status future trends computer vision-based behavior recognition. The present study conducted systematic literature review 80 peer-reviewed journal articles following Preferred Reporting Items for Systematic Assessment Meta-Analysis (PRISMA) guidelines. Three questions were addressed concerning goal orientation, recognition techniques, challenges. Results showed that: (1) vision-supported focused four categories: physical action, learning engagement, attention, emotion. Physical actions engagement have been primary targets; (2) behavioral categorizations defined various ways connections to instructional content events; (3) existing studies college students, especially natural classical classroom; (4) deep was main method, YOLO series applicable multiple purposes; (5) moreover, we identified challenges experimental design, methods, practical applications, pedagogical vision. This will not only inform application vision but also provide insights research.

Language: Английский

---