Search

Abstract This preliminary study explores how GPT-4 Vision (GPT-4V) technology can be integrated into teacher analytics through observational assessment, aiming to improve reflective teaching practice. Our study develops a Video-based Automatic Assessment System (VidAAS) powered by GPT-4V. This approach uses Generative Artificial Intelligence (GenAI) to provide detailed insights into classroom dynamics. Our study encompasses various methods with multiple steps: a comprehensive literature review, prototype development of the VidAAS, and usability testing with in-service teachers. The study findings reveal that VidAAS demonstrates high accuracy in evaluating skills in the behavioral (psychomotor) domain and offers comprehensive explanations for each assessment. While showing promise in these areas, the system also indicates potential for further enhancement in processing speed and refinement in assessing cognitive and affective domains. We discuss how VidAAS supports teachers’ reflection-in-action and reflection-on-action, emphasizing the need to balance AI-driven insights and human judgment. Our study findings also guide future research avenues for VidAAS design, implementation, and integration in teacher analytics, underscoring GPT-4V’s potential for real-time, scalable feedback and a deeper classroom understanding.

Despite the development of various AI systems to support learning in various domains, AI assistance for art appreciation education has not been extensively explored. Art appreciation, often perceived as an unfamiliar and challenging endeavor for most students, can be more accessible with a generative AI enabled conversation partner that provides tailored questions and encourages the audience to deeply appreciate artwork. This study explores the application of multimodal large language models (MLLMs) in art appreciation education, with a focus on developing LLaVA-Docent, a model designed to serve as a personal tutor for art appreciation. Our approach involved design and development research, focusing on iterative enhancement to design and develop the application to produce a functional MLLM-enabled chatbot along with a data design framework for art appreciation education. To that end, we established a virtual dialogue dataset that was generated by GPT-4, which was instrumental in training our MLLM, LLaVA-Docent. The performance of LLaVA-Docent was evaluated by benchmarking it against alternative settings and revealed its distinct strengths and weaknesses. Our findings highlight the efficacy of the MMLM-based personalized art appreciation chatbot and demonstrate its applicability for a novel approach in which art appreciation is taught and experienced.

Debate is a universally acknowledged competency for its vital role in fostering essential skills such as analytical reasoning, eloquent communication, and persuasive argument construction. This is relevant in both formal educational settings like classrooms and informal venues such as after-school clubs. Traditional debate training methods often face challenges in facilitating personalized learning, primarily due to the difficulty of securing a debate partner. However, the emerging field of generative AI offers a promising alternative. This study evaluates the effectiveness of DEBO, a debate education application that utilizes ChatGPT, an advanced language model by OpenAI. Focusing on university students actively involved in a debate club, six volunteers tested DEBO and participated in interviews to provide insights. The findings reveal that DEBO notably enhances divergent thinking and features a comprehensive analytics dashboard for in-depth performance analysis. However, there are few limitations including occasional shortcomings in debate quality, minor inaccuracies, and the need for faster response times. The study underscores the transformative potential of advanced language models like ChatGPT in reshaping debate education, while also identifying areas for further improvement.

Given the importance of digital technology in daily life, computational thinking (CT) has become a necessary skill for everyone, not just for computer scientists. For CT development, students need to receive appropriate social learning support. However, instructors find it difficult to provide such support to many students in online courses. This study aimed to examine the effectiveness of e-mentoring via social network services (SNS) in developing students’ CT during large-scale online courses. A total of 327 undergraduate students volunteered to participate in this study, which included 16 weeks of lectures aimed at developing CT. The effects of SNS-based e-mentoring on CT development, the influences of each e-mentoring activity on CT development, and gender differences were analyzed using data on participants’ CT assistance scores and their utilization of e-mentoring activities. The findings indicated that SNS-based e-mentoring was effective in developing the CT of undergraduate students’ engagement in a large-scale online course. The most influential e-mentoring activities for students’ CT development were informational and technical support in a group and informational support in a private environment. Female students benefited more from SNS-based e-mentoring than male students, and they also engaged in more types of e-mentoring activities than male students. Participation in SNS-based e-mentoring was found to lower the gap in CT between students with and without prior learning experience. Our study findings can be used by educational institutions and instructors when designing courses for students’ CT development in large-scale online courses or when developing strategies to close the gender gap in CT ability.

Abstract Transferrin receptor (TFRC) is the major mediator for iron entry into a cell. Under excessive iron conditions, TFRC is expected to be reduced to lower iron uptake and toxicity. However, the mechanism whereby TFRC expression is maintained at high levels in iron‐enriched cancer cells and the contribution of TFRC to cancer development are enigmatic. Here the work shows TFRC is induced by adenomatous polyposis coli (APC) gene loss‐driven β‐catenin activation in colorectal cancer, whereas TFRC‐mediated intratumoral iron accumulation potentiates β‐catenin signaling by directly enhancing the activity of tankyrase. Disruption of TFRC leads to a reduction of colonic iron levels and iron‐dependent tankyrase activity, which caused stabilization of axis inhibition protein 2 (AXIN2) and subsequent repression of the β‐catenin/c‐Myc/E2F Transcription Factor 1/DNA polymerase delta1 (POLD1) axis. POLD1 knockdown, iron chelation, and TFRC disruption increase DNA replication stress, DNA damage response, apoptosis, and reduce colon tumor growth. Importantly, a combination of iron chelators and DNA damaging agents increases DNA damage response and reduces colon tumor cell growth. TFRC‐mediated iron import is at the center of a novel feed‐forward loop that facilitates colonic epithelial cell survival. This discovery may provide novel strategies for colorectal cancer therapy.

Artificial neural networks evolve into deep learning recently and perform well in various fields, such as image and speech recognition and translation. However, there is a problem that it is difficult for a person to understand what exactly the trained knowledge of an artificial neural network. As one of the methods for solving the problem of the artificial neural network, rule extraction methods had been devised. In this study, rules are extracted from artificial neural networks using ordered-attribute search (OAS) algorithm, which is one of the methods of extracting rules from trained neural networks, and the rules are analyzed to improve the extracted rules. As a result, we found that when the output value of the hidden layer has an intermediate value that is not close to 0 or 1 after passing through the sigmoid function, the problem of rule uncertainty occurs and affects the accuracy of the rules. In order to solve the uncertainty problem of the rules, we applied the hidden unit clarification method and suggested that it is possible to extract the efficient rule by binarizing the hidden layer output value. In addition, we extracted CDRPs (critical data routing paths) from the trained neural networks and used CDRPs to prune the extracted rules, which showed that the uncertainty problem of rules can be improved.

In recent years, the development of adaptive models to tailor instructional content to learners by measuring their cognitive load has become a topic of active research. Brain fog, also known as confusion, is a common cause of poor performance, and real-time detection of confusion is a challenging and important task for applications in online education and driver fatigue detection. In this study, we propose a deep learning method for cognitive load recognition based on electroencephalography (EEG) signals using a long short-term memory network (LSTM) with an attention mechanism. We obtained EEG signal data from a database of brainwave information and associated data on mental load. We evaluated the performance of the proposed LSTM technique in comparison with random forest, Adaptive Boosting (AdaBoost), support vector machine, eXtreme Gradient Boosting (XGBoost), and artificial neural network models. The experimental results demonstrated that the proposed approach had the highest accuracy of 87.1% compared to those of other algorithms, including random forest (64%), AdaBoost (64.31%), support vector machine (60.9%), XGBoost (67.3%), and artificial neural network models (71.4%). The results of this study support the development of a personalized adaptive learning system designed to measure and actively respond to learners’ cognitive load in real time using wireless portable EEG systems.

The rapid advancements in artificial intelligence (AI) have transformed various domains, including education. Generative AI models have garnered significant attention for their potential in educational settings, but image-generative AI models need to be more utilized. This study explores the potential of integrating generative AI, specifically Stable Diffusion, into art-focused STEAM classes. To this end, we combined STEAM education with image-generative AI. We designed and implemented a learning activity in which students used image-generative AI to generate creative images and wrote imaginative diaries inspired by AI-generated pictures. Using a mixed-methods approach, we collected and analyzed data from surveys, interviews, and prompt datasets to gain insights into students' perceptions, motivations, and behaviors concerning the integration of generative AI in art education. Our study found that incorporating generative AI into art-focused STEAM classes may reduce the gap between male and female students' interest in art practices. We also found that the number of subjects in students' prompt datasets is linked to their divergent and convergent thinking. Finally, we discovered a meaningful correlation between the number of unique words in students' diaries and the image scores that researchers assessed. Our findings shed light on the relationship between students' divergent and convergent behaviors, offering valuable insights into their creative thinking processes. This study underscores the potential of integrating image-generative AI models in STEAM education.

Through design and development research (DDR), we aimed to create a validated automatic question generation (AQG) system using large language models (LLMs) like ChatGPT, enhanced by prompting engineering techniques. While AQG has become increasingly integral to online learning for its efficiency in generating questions, issues such as inconsistent question quality and the absence of transparent and validated evaluation methods persist. Our research focused on creating a prompt engineering protocol tailored for AQG. This protocol underwent several iterations of refinement and validation to improve its performance. By gathering validation scores and qualitative feedback on the produced questions and the system's framework, we examined the effectiveness of the system. The study findings indicate that our combined use of LLMs and prompt engineering in AQG produces questions with statistically significant validity. Our research further illuminates academic and design considerations for AQG design in English education: (a) certain question types might not be optimal for generation via ChatGPT, (b) ChatGPT sheds light on the potential for collaborative AI-teacher efforts in question generation, especially within English education.

Intelligent Personal Assistants (IPAs) are becoming more prevalent in daily and educational contexts, increasing the possibility of using them as learning partners that can provide more personalized and learner-centric learning opportunities. However, research has primarily focused on educational advantages that IPAs may provide, overlooking investigation on students' acceptance of IPAs. To address this gap, this study empirically explored factors influencing students' intention to use IPAs for learning. A conceptual research model was designed based on the technology acceptance model. Data were collected from 1044 South Korean students, who were experienced in using IPAs. Results were analyzed using structural equation modeling. The findings indicate that the proposed model sufficiently explains students' IPAs acceptance for their learning. In addition, students' computer self-efficacy and self-management of learning ability were prominent factors influencing their intention. Moreover, their trust toward IPA and feeling of social presence on IPAs were found to be significant factors to be considered as well. Based on these findings, future recommendations for fostering students' active use of IPAs for learning are discussed. Additionally, both theoretical and practical implications are highlighted.