In the digital age, where media proliferation challenges traditional reading habits, this study investigated the impact of digital platforms on critical thinking (CT) and reading practices. Some conventional e-books may not sufficiently encourage reflective thinking or foster CT skills due to their linear nature and lack of engaging elements. Employing the Practical Inquiry Model (PIM) within the Community of Inquiry (CoI) framework, this study highlights the integration of multimedia elements and the use of students' selfies to boost engagement and social presence in a digital learning environment. This study introduced selfie role-playing reading, along with group discussions from various perspectives as an innovative pedagogy to enhance cognitive presence within the CoI, ultimately enhancing learners’ CT performance. Using a quasi-experimental design, 63 participants were grouped into single-perspective or multiple-perspective categories by narrative perspective exposure, and then engaged in online story discussions with worksheets. Reading performance was measured through a detailed examination of their CT patterns via behavioral sequential analysis. Results revealed that students in the multi-perspective groups demonstrated superior CT and reflective thinking levels, underscoring the potential of innovative e-book designs in enhancing CT. This study not only validates the importance of diverse narratives and role-playing in CT development, but also pioneers a method for analyzing such skills through behavioral sequence analysis. It signifies a leap in applying the PIM to multimedia e-book reading, offering insights for future educational strategies and CT assessment in the digital era.

The transformative wave of generative AI is reshaping the creative thinking processes of learners, posing a significant challenge to education and industry in cultivating technological literacy and creativity. This study delves into the exploration of how learners can effectively tackle new challenges by deconstructing fragments from a macro perspective and generating innovative methods or concepts. In the Scratch visual programming environment, learners in the self-regulated learning mode observed entire functioning projects, facilitating easy disassembly and learning, namely by using Code Decomposed by Learner (CDBL). A total of 104 fifth-grade students were divided into two learning scaffoldings: (1) domain-general, learning from the top down (CDBL-TD), and (2) domain-specific, learning from the bottom up (CDBL-BU). Students in the CDBL-TD group exhibited a high degree of completion, strong exploration abilities, and the willingness to experiment with unlearned functions. Although there was no significant difference in originality between the two groups, students in the CDBL-TD group showcased greater uniqueness in designing characters or items within the game. This study introduces a novel programming learning scaffolding, offering instructors a tool to guide students’ creativity and enhance their programming capabilities.

In the realm of Virtual Reality (VR), the exploration of players' modes of free movement has persistently been a pivotal research focus. This study introduces a novel locomotion approach, specifically employing gaze-directed instantaneous destination selection, aiming to enhance the current methods of movement in VR games and elevate the overall user experience. Through the creation of corresponding game scenarios, we assess the instantaneous movement performance in terms of efficiency, precision, and comfort for both gaze-directed and controller-directed destination selection. Statistical analyses reveal that, in terms of efficiency, gaze-directed instantaneous destination selection outperforms controller-directed movement. However, in the evaluation of comfort, controller-directed movement surpasses gaze-directed instantaneous destination. The findings of this research not only present a viable alternative in movement methodology but also underscore the necessity of striking a balance between efficiency and comfort in VR game design to deliver an enhanced gaming experience.

Past research on sign language recognition has mostly been based on physical information obtained via wearable devices or depth cameras. However, both types of devices are costly and inconvenient to carry, making it difficult to gain widespread acceptance by potential users. This research aims to use sophisticated and recently developed deep learning technology to build a recognition model for a Taiwanese version of sign language, with a limited focus on RGB images for training and recognition. It is hoped that this research, which makes use of lightweight devices such as mobile phones and webcams, will make a significant contribution to the communication needs of deaf and hard-of-hearing (DHH) individuals.

Like any other skill, social-emotional learning (SEL) must be practiced to be automatic. To transform the knowledge of SEL into a practical and deliverable lesson, we propose an emotion-oriented journaling system with comics for students. The system transforms the cognitive-behavioral therapy into manageable steps using comics-based storytelling and uses a method combining the term frequency-inverse document frequency (TF-IDF) method and generative adversarial network (GAN) algorithm to implement a text-to-graphics conversion. The system was tested by 36 fifth-grade students for four weeks. The experimental result shows that 41.7% of students improved their abilities of emotional recognition and self-awareness after one month of journaling on the system. Consequently, the system provides a creative and effective way for social-emotional education while speeding up the generation of visual journaling content in less time.

The purpose of designers developing games is to create experiences. However, with the development of technology, people have a new choice of playing games - watching Game live Streaming. The live broadcast industry has developed rapidly in recent years. It combines game screens, chat rooms, and live host video viewing methods, allowing viewers to interact with live host and other audiences while enjoying the game process. Therefore, this study wants to explore what makes live viewers choose to watch others play games instead of playing games themselves, and how different types of viewers have different viewing experiences in the live viewing process. This research focuses on the audience, adopts an experimental method to record the process of subjects watching the live action with an eye tracker, and cooperates with the viewing motivation scale and the game experience questionnaire. Through follow-up interviews, explore the viewing behavior of different types of audiences, trying to understand what elements in the live game catch the audience's attention. The results show that the motivation for social interaction can clearly distinguish the audience's attention allocation on the game screen and the chat room, showing that the chat room is very important to the audience with this motivation; The motivation to acquire new knowledge will allow the audience to pay more attention to the game screen to receive more new information. Finally, this study also presents suggestions based on the results as a reference for follow-up research and live game management methods.

Education has always been a crucial and concerning issue in human civilization. With the increasing popularity of games, various instructional theories based on game-based learning have been proposed. However, whether in gaming or learning, scaffolding plays a significant supportive role. Therefore, this study adopts the four dimensions of FSLSM as the classification of learning styles and designs four different types of scaffolding for players to use. By analyzing the behavior of players with different learning styles in using scaffolding during the game, we aim to tailor the most suitable scaffolding for players with different learning styles. The research findings indicate that when the educational content in game-based learning relates to logic and engineering, providing more visualized and detailed scaffolding will enhance learners’ motivation to actively use scaffolding and improve the overall learning quality.

After the Taiwanese government launched the Blueprint for Developing Taiwan into a Bilingual Nation by 2030, the Implementation Project of Bilingual Instruction in several domains of primary and junior high school education was promoted by the Taiwan Ministry of Education. Content and language integrated learning (CLIL) is a dual-objective strategy in which students simultaneously acquire language skills and subject knowledge. CLIL has been widely implemented and proven to be successful in European countries. This strategy will become the primary method of bilingual education for instructors in grades K-12. Other non-European countries, such as Latin America and Indonesia, however, found difficulties applying CLIL in class. The issues need to be identified to encourage researchers and practitioners to find solutions. The purpose of this paper is to identify the barriers to implementing CLIL from the perspective of K-12 teachers. We surveyed 102 K-12 teachers in Keelung using a questionnaire. The findings indicate that (1) 49.0% of teachers feel concerned if they are required to teach half of their content in English, (2) 64.8% of teachers are willing to improve their English proficiency in their spare time for implementing CLIL in class, and (3) only 36.3% of teachers are willing to adopt CLIL in class, while 50.0% of teachers are willing to implement CLIL if a teaching assistant system supports them with English course materials. Teaching assistant systems that provide English course materials might influence teachers’ willingness to implement CLIL in class.

Computational thinking skills are increasingly required for working with information technology products and are considered core learning objectives in science and technology curriculums across all grades. However, there is yet to be a curriculum model for computational thinking, and many teachers are still figuring out this issue and designing courses to cultivate these skills in students. We planned 8-course periods for 108 curriculums, using the Bebras International Computational Thinking Challenge and programming learning motivation scale to evaluate game-based lessons from Code.org. The grade-3 and -4 students were randomly divided into self-regulation and guided-learning groups, and 153 valid data were analyzed using paired t tests and ANCOVA. As a result, we found the learning behaviors of the two groups of students to be worthy of further exploration in terms of time management and help-seeking learning strategies. Code.org’s game-based lessons effectively engage students to complete most of the course, addressing the usual course completion issues when self-paced. The self-regulation group spent more time in peer discussions and had better learning outcomes than the guided-learning group. To this end, we provide detailed curriculum information as a teaching model for the self-regulated learning of computational thinking in primary schools.

Flow, a profound psychological state associated with optimal experiences, serves as a pivotal gauge of engagement across various contexts. Traditionally evaluated through methods like surveys, interviews, and sampling surveys for qualitative analysis, these approaches aimed to decipher participants' experiences and perceptions. However, these methods are time-intensive and prone to memory and expressive limitations. In response, contemporary research increasingly leans towards physiological indicators for assessing flow. Utilizing physiological signals to detect flow reduces narrative inaccuracies and minimizes sensitivity to participants' subjective awareness, albeit demanding significant time, human resources, and specialized equipment. To surmount these challenges, this study introduces an innovative methodology for efficiently predicting flow states. utilizing gameplay and interaction data as inputs for machine learning models. By employing a real-time strategy game as the experimental environment, participants' gameplay recordings and interaction records are collected and paired with the Flow Short Scale questionnaire to establish a predictive model for flow state. The results demonstrate the success of this approach, achieving a significant prediction accuracy ( $MAE =0.0623$ ) and highlighting a strong correlation between objective gameplay records and subjective flow experiences. This streamlined methodology offers a promising avenue for quantifying and predicting flow state, contributing to a deeper understanding of engagement dynamics in digital environments.

With the advancements in hardware, deep learning, and the invention of generative adversarial networks (GANs), the integration of games and artificial intelligence (AI) has mostly focused on assisting game development, such as creating maps, skills, monsters, NPCs, and levels. In this chapter, a different approach is proposed, which utilizes artificial intelligence in games to allow players to customize game content. Based on this concept, a method for automatically generating 3D game character models using images is presented, specifically the Parallel PIFu (Pixel-Aligned Implicit Function) model. This method leverages the characteristics of the PIFu model and combines the features of the generated 3D models from the front and back views of the person captured in the images. By merging these features, the method produces 3D character models that preserve the details from both images without any missing body parts. This approach builds upon existing techniques for automatic generation of character models and further enhances them. It enables users to simply use their mobile phones to capture images of their desired characters, which can then be automatically transformed into corresponding 3D models. These models are compatible with most existing games on the market, allowing players to easily create personalized appearances for their in-game characters and enhance the overall gaming experience.

Real-time strategy (RTS) games simulate battlefield leadership and tactical and strategic operations. Most overemphasize the number of actions per minute (APM), which encourages players to click rapidly and constantly rather than apply deliberate and finely tuned strategies or tactics. New RTS games featuring resource dispersion game mechanics aimed at reducing APM demand and promoting strategic planning are being released. We created three versions of a single RTS game, recruited players, recorded their game control data, and asked them to complete a simple after-game questionnaire. Data were used to analyze tactical and strategic applications. Player actions were observed and player opinions analyzed in an attempt to identify an optimal game structure as measured by strategic and tactical play.

Past research on sign language recognition has mostly been based on physical information obtained via wearable devices or depth cameras. However, both types of devices are costly and inconvenient to carry, making it difficult to gain widespread acceptance by potential users. The goal of this research is to use sophisticated and recently developed deep learning technology to build a recognition model for a Taiwanese version of sign language, with a limited focus on RGB images for training and recognition. It is hoped that this research, which makes use of lightweight devices such as mobile phones and webcams, will make a significant contribution to the communication needs of deaf and hard-of-hearing (DHH) individuals.

The authors recruited 267 Taiwanese elementary school students to play digital games in order to investigate the influence of prior game play experience on problem-solving performance, with specific focuses on self-regulated learning, problem-solving processes, problem contexts, and problem characteristics. Data were acquired using a self-regulated learning scale, a game experience questionnaire, a problem-solving process evaluation instrument based on the 2012 Program for International Student Assessment, a computer-based survey focused on seven problem-solving scenarios, and four digital games. Our results indicate a positive and significant effect on problem-solving performance among study participants with seven or more years of game play experience, and positive feelings toward game learning as reflected in negatively skewed self-regulated learning scores. We also found that the “representation and formulation” stage served as a significant performance predictor, the “planning and execution” and “monitoring and reflection” stages were significant predictors of dynamic/technical problems, and the “exploring and understanding” stage was a predictor for static/non-technical problems. Our data indicate significant differences in problem-solving cognitive processes across ranges of contexts and characteristics. We believe our findings will be useful for researchers studying the potential use of digital game frameworks to measure specific learning mechanisms.

C/C++ is one of the most common programming languages in introductory computer science courses. For students with varying levels of digital literacy, traditional teaching strategies that begin with C/C++ syntax and concepts appear inappropriate. Moreover, different programming languages are required for specific industrial purposes. To complement traditional teaching methods, we design a learner-centered constrictive strategy. Other than C/C++, students begin with two programming languages. Rather than conventional lectures, students learn to program by modifying, decomposing, and reassembling example codes. The midterm assesses their learning effectiveness. After the midterm, they can choose to take the collegiate programming examination (CPE) or not. If they pass CPE, they can pick between 1) regular class and final exam, or 2) no class after midterm, find a topic for a final project, finish it and present it in the last class. We invited 21 students from a class of 90. Eight students had no prior coding experience. Even though these eight students obtained lower scores in the midterm, they all had significant improvements in the final exams. Among sixteen students, six of whom had no prior coding experience, acknowledged that they learned useful skills in the class. Our results show that students with varying coding backgrounds can benefit from our proposed teaching strategy.

We study the problem of controllable citation text generation by introducing a new concept to generate citation texts. Citation text generation, as an assistive writing approach, has drawn a number of researchers’ attention. However, current research related to citation text generation rarely addresses how to generate the citation texts that satisfy the specified citation intents by the paper’s authors, especially at the beginning of paper writing. We propose a controllable citation text generation model that extends a pre-trained sequence to sequence models, namely, BART and T5, by using the citation intent as the control code to generate the citation text, meeting the paper authors’ citation intent. Experimental results demonstrate that our model can generate citation texts semantically similar to the reference citation texts and satisfy the given citation intent. Additionally, the results from human evaluation also indicate that incorporating the citation intent may enable the models to generate relevant citation texts almost as scientific paper authors do, even when only a little information from the citing paper is available.

When controversial products are introduced, effective promotion efforts and eventual public acceptance require consideration of multiple factors such as existing social network structures, numbers of pioneers and their locations, and appropriate methods for product information diffusion. These factors have been the focus of marketing, investment and other studies for many years, most recently in computer information science. Researchers are especially motivated to understand diffusion processes for new technologies and controversial products within and across social networks. While many product diffusion simulation models have been proposed, most suffer from assumptions of unchanging internal agent attitudes toward products, no opinion exchanges between agents, and non-significant relationships between agent internal opinion attitudes and diffusion thresholds. In this paper we propose an opinion dynamics model that assumes both agent interaction and changes in agent attitudes over time. Social psychology theory is used to explain interactions between opinion and diffusion dynamics, with changing agent attitudes and behaviors affected by interpersonal relationship factors. Simulations were used to study dynamic diffusion processes involving controversial products (e.g., vaccines and genetically modified foods) in different social networks. Results indicate that the proposed model accurately reflects several kinds of social phenomena, including pioneer influences, rural marketing strategies, and the influence of social network structure. This effort to identify instances of product diffusion under various social conditions is offered in support of research in communication dynamics and social media-centered marketing strategies.

Lung sounds remain vital in clinical diagnosis as they reveal associations with pulmonary pathologies. With COVID-19 spreading across the world, it has become more pressing for medical professionals to better leverage artificial intelligence for faster and more accurate lung auscultation. This research aims to propose a feature engineering process that extracts the dedicated features for the depthwise separable convolution neural network (DS-CNN) to classify lung sounds accurately and efficiently. We extracted a total of three features for the shrunk DS-CNN model: the short-time Fourier-transformed (STFT) feature, the Mel-frequency cepstrum coefficient (MFCC) feature, and the fused features of these two. We observed that while DS-CNN models trained on either the STFT or the MFCC feature achieved an accuracy of 82.27% and 73.02%, respectively, fusing both features led to a higher accuracy of 85.74%. In addition, our method achieved 16 times higher inference speed on an edge device and only 0.45% less accuracy than RespireNet. This finding indicates that the fusion of the STFT and MFCC features and DS-CNN would be a model design for lightweight edge devices to achieve accurate AI-aided detection of lung diseases.