The recent emergence of Generative Artificial Intelligence (GenAI) tools, such as ChatGPT, has introduced revolutionary capabilities that are predicted to transform numerous facets of society. For students, the advent of GenAI has the potential to profoundly alter studying and learning practices. This paper presents the development of QuizWiz: a web application that provides innovative study tools integrating GenAI. QuizWiz includes two GenAI features, (1) generating study flashcards with questions and answers from uploaded study material and (2) answering study questions with an intelligent chatbot. We detail the technical aspects of developing tools that integrate GenAI, as well as provide recommendations for developers and academics interested in the use of GenAI for students.
Most machine learning (ML) algorithms work best when the samples in each class are almost equal. However, if a dataset has imbalanced samples then the ML model can achieve a high accuracy by just predicting the majority of classes and achieving a high classifier performance. Different resampling methods are used to handle this imbalance issue by creating synthetic data in the minority class or removing samples from the majority class until the classes are balanced. However, using these resampling methods can harm the dataset performance by creating synthetic data samples in a dataset that already suffers from overlap and other data intrinsic issues. Recently, dataset complexity measures, by observing the characteristics of a dataset, are often used in machine learning tasks as intrinsic descriptors to calculate the difficulty of a classification problem. This study investigates how resampling methods can affect the dataset in terms of complexity, overlapping, and classification accuracy. This is achieved by monitoring 22 different measurements related to data complexity measures obtained from 20 public and pre-processed datasets from the KEEL repository. Our empirical findings demonstrate a strong positive correlation between the resampling methods and the dataset's complexity and classification accuracy. We also posit that the main reason for poor classification accuracy on an imbalanced dataset is not due to its imbalanced nature but due to the other intrinsic characteristics of the dataset. Finally, we advocate that pre-processing methods and ML algorithms should be based on the dataset's specific properties, rather than being chosen on an ad hoc basis.
Engineering design course is renowned for its abstraction and complexity, which can have an impact on the quality of classroom instruction. To address this issue, this study has undertaken the design and development of an E-learning software utilizing simulation technology and the C# programming language. With a focus on the challenging topic of torsion resistance coefficient, a 3D model of the gearbox shaft was established, and simulation calculations were used to obtain digital resources such as Mises stress plots. Leveraging these resources, an E-learning software was created, garnering widespread recognition with a high approval rate of 92.86% among students. This software effectively utilizes digital resources to aid students in comprehending abstract concepts and significantly enhances their enthusiasm for learning. The findings of this research contribute to the field of E-learning and provide valuable references for future design and development of engineering design case studies.
The blended teaching for preschool teacher education needs to be predicated on the real educational circumstances in the early childhood settings. Firstly, this study investigates the feasibility of blended teaching practice based on the theory of situated learning. Secondly, in combination with the teaching reform case of the first-class undergraduate course "Mathematics Education for Preschool Children" in Shanghai, the paper explicates the specific approaches of teaching reform: project activity practices based on real educational scenarios, curriculum resource construction based on learning communities, and innovative mixed teaching models based on dual tutors. Finally, the research validates the effect of blended teaching based on the situated learning theory.
Text-form student feedback is an indispensable source of information for all university lecturers. Since manual analysis of such feedback is laborious, it has been suggested that text clustering methods could be used to automate the process. However, the success of text clustering depends heavily on the vector space presentation of documents. In this paper, a comprehensive evaluation of eight vector space models (VSMs) is presented in combination with different linguistic preprocessing techniques in English and Finnish student feedback data. The results show that the choice of VSM has a strong effect on the clustering performance. The models based on short and long character n-grams work best, while word2vec models perform worst. In general, stop word removal has a positive effect, while stemming and lemmatization may be detrimental with many VSMs. The main themes of the data could be well identified from cluster centroids. An alternative approach of describing clusters by frequent word n-grams worked also well for sufficiently large, distinct classes with clear keywords.
Scientific and objective evaluation of education quality is an important demand of the current education industry. Artificial intelligence empowering all walks of life has become an inevitable trend of social development in the future. This study introduces a scheme of artificial intelligence for education quality assessment. The education quality evaluation system combines big data and artificial intelligence technology, uses artificial intelligence model construction, analyzes the collected teaching quality evaluation index data, and generates objective results for education quality evaluation. The teaching quality evaluation of higher education teaching quality is a more complex system engineering. Its remarkable characteristics are complex levels, strong structure, large amount of information processed, and need to consume a lot of manpower, material and financial resources. The traditional manual processing method can not adapt to the requirements of accurate evaluation, fast evaluation and good evaluation. Therefore, this study can help to realize the artificial intelligence of teaching evaluation, combined with computer network technology, so that the evaluation is more suitable for the requirements of higher education teaching quality evaluation.
With technological advancements and evolving educational needs, traditional methods of Chinese reading instruction face challenges. ChatGPT, a robust natural language processing tool, introduces new possibilities for Chinese reading education. This study utilizes standardized reading proficiency test scores from 446 Chinese heritage language (CHL) learners to investigate the impact of character recognition abilities on their overall Chinese reading proficiency through quantitative analysis. Drawing on empirical findings, the research examines ChatGPT's specific contributions in teaching Chinese characters and enhancing reading skills tailored for heritage learners. The study aims to explore ChatGPT's potential as an innovative artificial intelligence language model to enhance efficiency and educational outcomes in Chinese reading instruction, providing recommendations for overseas Chinese heritage language education.
The rise of artificial intelligence is reshaping various fields worldwide, with AI-generated content (AIGC) leading a new wave of content creation. As AIGC technology matures, it is capable of generating richer and more realistic content in various forms, including text, images, audio, and video, thus solving the challenges traditionally requiring human enhancement and data construction. However, how to apply AIGC-generated content and these achievements in education and teaching remains a thought-provoking issue. This paper first introduces the current state of research on AIGC technology. Then, using specific examples of AIGC-generated content, it presents the methods for generating watermarks based on content and form, as well as the development of watermarking technologies in four forms: text, image, audio, and video. Furthermore, this paper explores the transformative potential of AIGC in AI professional education, focusing on the current challenges in AI education development and the innovative applications of AIGC technology. Finally, the paper envisions the development of an innovative AIGC educational system for the future, highlighting potential risks and countermeasures.
This study presents an Emotion Recognition Multi-Attention Model (EmoMA-Net), a novel multimodal neural network aimed at enhancing real-time emotion recognition in educational environments. By leveraging the WESAD dataset, our model combines Convolutional Neural Networks (CNN), a Time Series Memory System (TSMS), and a Multi-Attention Mechanism to analyze diverse physiological signals, such as heart rate variability (HRV) and electroencephalogram (EEG). Unlike traditional emotion recognition methods reliant on subjective self-reports, our model delivers objective and accurate predictions of student stress levels through multimodal physiological data collected from wearable sensors. Achieving accuracy up to 99.66%, it facilitates adaptive educational systems to provide real-time feedback to educators, enabling prompt adjustments to teaching strategies. This advancement represents forward in emotion prediction technology, contributing to more responsive and adaptive educational experiences based on real-time emotional insights.
The ongoing integration of Generative Artificial Intelligence (GenAI) within higher education (HE) signifies a pivotal shift in pedagogical paradigms, demanding comprehensive theoretical and practical considerations. This paper critically examines the multifaceted adoption of GenAI in HE by reviewing interdisciplinary theoretical frameworks from psychology, computer science, and pedagogy. It highlights the insufficiency of traditional technology acceptance models, which predominantly address cognitive and rational decision-making processes, and advocates for the inclusion of emotional and ethical dimensions often overlooked in existing frameworks. By synthesizing research across various disciplines, this review identifies significant gaps and proposes an integrated theoretical model to effectively understand and guide GenAI adoption. The proposed framework emphasizes the need for robust, empirically supported methodologies that accommodate the complex, dynamic nature of GenAI applications. This paper not only contributes to academic discourse by providing a comprehensive review of existing literature but also sets a foundation for future empirical studies aimed at refining GenAI integration strategies in HE, ensuring they are ethically aligned and educationally effective.
With the rapid development of information technology, online learning has become a vital learning method in higher education. This transformation brings both opportunities and challenges. Online learning gives college students more autonomy and selectivity than traditional learning methods. However, it inevitably leads to differences in students’ learning engagement. The attention and support from the student's family environment, as well as the student's emotional perception and regulation, may all be the factors that affect their learning engagement. The primary contribution of this study is an empirical investigation into how college students’ emotional intelligence mediates the association between parenting styles and online learning engagement. Based on a sample of 681 college students, this study indicates that positive parenting is significantly positively correlated with students’ emotional intelligence and online learning engagement; Negative parenting is significantly negatively correlated with students’ emotional intelligence and online learning engagement; Emotional intelligence mediated the relationship between parenting style and college students’ online learning engagement. The results not only emphasize the long-term impact of parenting style in education but also delve into the crucial role of emotional intelligence in online learning environments, providing insights into further supporting students’ learning engagement.
This systematic literature review explored the opportunities and advantages of integrating Artificial Intelligence Generated Content (AIGC) tools like OpenAI's ChatGPT, Copilot, and Codex in programming education. From an initial pool of 1,173 papers, 24 were rigorously selected for detailed analysis. The findings highlighted the dominant use of ChatGPT, particularly versions 3/3.5 and 4, underscoring its effectiveness and accessibility. Python emerged as the most frequently studied language, followed by Java, C, R, and Scala. A notable research gap was identified in block-based programming languages and online/blended learning environments. Key opportunities and advantages identified included enhanced code review, where AIGC tools offer efficient and comprehensive assessments; personalized learning, with ChatGPT providing individualized feedback and improving student comprehension; and increased student engagement and motivation through interactive features. Additionally, AIGC tools significantly improved problem-solving and debugging support, effectively identifying and correcting coding errors. They also supported diverse learning styles by offering varied examples and solutions, facilitated innovative teaching strategies that improved educational outcomes, and reduced teacher workload by automating routine tasks. These insights demonstrated the transformative potential of AIGC tools in revolutionizing programming education.
With the rapid advancement of generative artificial intelligence, its demonstrated intelligence, cognition, and other abilities have created new opportunities for the reform of education. This paper examines the core curriculum development of digital media technology, the situation of emerging engineering education(3E), and the application of AIGC in education. It concludes that the main bottlenecks in the fusion of engineering and design of digital media technology are curriculum development, assessment, and duration of practical training. Consequently, the study explores the application and facilitation of AIGC within digital media technology and how AIGC alleviates the bottlenecks of major development. Building on this, the paper presents a case study centered on the core curriculum of digital media technology that found AIGC can significantly improve course teaching efficiency. It aims to provide insights and recommendations for teaching practices in core digital media technology courses.
To address the need for more engaging and effective English education, this study compared the effects of three English learning approaches for preschool children: (1) traditional English education, (2) an Augmented Reality (AR) tool built using Scratch, and (3) an iPad-based approach. Teachers created innovative and interactive scenarios under three conditions to teach vocabulary and pronunciation. Thirty preschool children were randomly assigned to the traditional, AR, and iPad groups. Our analysis showed that using the AR tool and iPad approach can improve both students' learning motivation and knowledge acquisition. At the same time, the visualized content is easily remembered, which made learning materials easily accepted by preschool children. Furthermore, using an Augmented Reality approach can provide a more relaxed learning environment, which has the potential to stimulate enthusiasm in pre-school children's English learning.
Facilitating financial aid work on college students from economically disadvantaged families is an essential requirement for carrying out the basic task of fostering virtue through education and running universities that are deeply rooted in China. Crucially, financial aid offerings, an important part of college education management and services, concern the vital interests of students in general. As our country's social economy and higher education keep on booming, colleges are working harder to provide better financial support for students in need. In the meantime, there are risks and challenges yet to be tackled, and relevant work needs to be optimized. This research delves into ways of constructing a precise financial aid system in colleges in the era of big data so that the financial aid policies of the Party and the State can benefit every student, therefore maximizing equity in education and mitigating the burden of needy students.
With the acceleration of the globalization of higher education, the number of doctoral students studying abroad is increasing. Based on the survey data of 2459 overseas doctoral students in the 2019 Global PhD Student Survey in Nature Journal, the study found that: First, the academic growth status of Chinese overseas doctoral students is generally higher than the global average, indicating that the doctoral education of Chinese international students has achieved initial results; second, the academic support, mental health support and career support of the training unit have a significant positive impact on the academic growth of international doctoral students, and are indispensable in the academic growth of doctoral students. Third, we found a negative association between the experience of psychological problems and the academic growth of international doctoral students.
In the post-pandemic era, university students face a future replete with uncertainties, engendering significant challenges in both their academic and personal lives. The uncertain environment and continuous online learning may contribute to the prevalence of procrastination among students. Therefore, in this era, it is crucial for college students to have a sense of self-assurance and self-control. Based on self-regulation theory, this study examines the issue of procrastination among college students from a new perspective. A total of 756 university students were surveyed using the Self-Concept Clarity Scale, the Pure Procrastination Scale, the Brief Self-Control Scale, and the Brief Self-Handicapping Scale. Data analysis was conducted using SPSS 26.0 and AMOS 26.0, and a chain mediation model was established (Isolated the influence of control variables). Results indicate that students' self-concept clarity not only directly negatively predicts procrastination but also affects it through self-control, self-handicapping, and their chain-mediated effects, with self-handicapping having the greatest impact. These findings suggest that schools and educators should focus on developing college students' self-concept clarity to alleviate their procrastination behavior, which can be achieved by enhancing self-control and preventing self-handicapping. This study provides guidance for schools and educators on promoting college students' self-awareness and self-management, laying the foundation for future prevention and intervention of procrastination among university students.
The progress of information technology has promoted the transformation of educational forms and learning methods. Online learning have become prevalent in our society. The analysis of online learners' preference characteristics is an important prerequisite for the realization of the precise service of lifelong education, and also an important basis for the implementation of personalized teaching. By capturing the relevant data of Learning Platform of Open University of China(LPOUC), conducting Hierarchical Clustering analysis, and analyzing the characteristics and preferences of online learners , we more deeply portray learners and provide support strategies for further improving online teaching capabilities.
In recent years, online education platforms represented by MOOC platforms are developing rapidly. There are no thresholds such as high school and college entrance examinations in online education, and there are no requirements and disadvantages that you must be in a certain place to study. This allows everyone to equally enjoy the high-quality education of prestigious universities, and learners can freely choose courses that are interesting, popular or praised by everyone according to their interests and needs. However, while providing extremely high convenience for learners, it also cuts off the direct contact between learners and other people in the traditional classroom learning environment. As a result, learners are always accustomed to studying alone, and have no motivation to actively contact or are not used to contacting non-direct contact learning partners on the Internet. Lack of communication with learning partners and a common learning atmosphere lead to problems such as low learning efficiency, low learning persistence, and high course dropout rates. This paper examines the key elements of learning partner construction in traditional pedagogy, combines the user behavior characteristics of online education platforms, and refers to the process of traditional e-commerce recommendation systems, and constructs a set of online education learners collaborative learning social relationship construction methods.
