The increasing proliferation of artificial intelligence (AI) in educational settings has drawn significant attention recently. The rapid commercialization of AI technologies from 2014 onwards has led to a burgeoning body of academic literature intersecting the interdisciplinary fields of AI and education. This paper aims to disseminate part of the findings of our bibliometric study exploring this contemporary academic discourse: 239 relevant publications were identified from the Web of Science (WOSCC) database from 2014 to 2023* (until September). A keyword co-occurrence analysis was conducted to elucidate prominent research trends; 71 co-occurring keywords were clustered into 4 clusters. An inductive-interpretive analysis was then employed to analyze the articles corresponding to each cluster, highlighting eleven overarching research trends. In this paper, we will predominantly discuss the trends regarding the proliferation of AI in the context of K-12 STEM education and teacher education. The findings provide key insights for researchers aiming to prepare K-12 STEM teachers to be AI-ready.

In recent years, educational research has emphasized the significance of computational thinking (CT) and mathematics knowledge acquisition for students' learning in engineering subjects. Pre-college education plays a crucial role in preparing students for future engineering learning. While the relationship between CT and mathematics has been extensively studied at the K-12 level, limited attention has been given to early childhood education (ECE) contexts. Our research aims to investigate the potential development of early CT through mathematics learning in a potentially reciprocal manner using developmentally appropriate educational tools and instructional approaches. To commence this exploration, our primary focus is to conduct a systematic review of existing empirical studies, synthesizing evidence to illuminate the interaction between CT and mathematics during early childhood. Using the seven powerful ideas from early CT proposed by Bers as the theoretical framework, our initial findings indicate that algorithms, hardware/software, and representation are the most studied CT competencies intersecting with early mathematics skills, including numerical and spatial knowledge, measurements, basic algebraical thinking, and data processing. The complex relationship between early CT and mathematics, in general, is also discussed. Overall, the findings of this study suggest that existing mathematics education in early childhood may foster the development of CT in young children. Although this study is preliminary, it serves as a valuable foundation for future research endeavors in uncovering the interdisciplinary nature of early CT within mathematics-integrated learning contexts, thus contributing to the advancement of this field.

This systematic review aims to explore how students' CS learning experience influences their perception of computational thinking. We identified 21 publications in SCOPUS and coded them for comparison and synthesis. Through a meta-analysis, we found a medium effect size (g =0.656; p<; 0.001) supporting the influence of CS learning on students' change of perception of CT. Through qualitative synthesis, results show that CS learning experience influences perception of CT in terms of self-efficacy, behavioral & emotional engagement, and awareness of careers. Findings provide insights to facilitate CT education strategies and promote educational equity in the STEM field.

The promotion of STEM (Science, Technology, Engineering, Mathematics) education aims to provide students with chances of solving real-world problems based on their disciplinary knowledge and skills. One teaching approach in STEM is through integration, which may enable students to understand the interconnection among disciplines and formulate a solution with consideration of multiple disciples. To implement STEM education in primary schools, an attention has been paid to teachers' professional development (PD), and one particular interest is to develop an effective way to elevate the teachers' knowledge in engineering to become capable in teaching in the integrated approach. In order to guide policymakers to design an effective PD workshop/course, our work aims to first gain an understanding of teachers' expectations toward the PD and then develop a conceptual PD model for integrated STEM. This paper conducts a qualitative case study to deeper analyze primary school teachers' perception of PD and STEM education, as well as their potential challenges. The data came from the individual interview with eight in-service primary teachers (n = 8), the result shows that teachers' challenges in attending PD lie in the intrinsic challenge of a lack of knowledge in 1). how to apply knowledge learned into practical teaching, 2). differentiation of needs and skills, and 3). their values while extrinsic challenges are reflected in aspects of 4). a lack of support, 5). vague school decision and 6). a lack of guidance of direction. Regarding their expectation of STEM PD, suggestions could be made on a focus of content & pedagogical knowledge of STEM subjects, practical activities and contextualisation.

In recent years, we have seen an increasing interest in bringing programming back to K-12 education. Many educators begin to recognize the necessity of helping children develop computational thinking (CT) as an essential skill to address 21st century challenges. In this poster, we report the preliminary findings from the first year of a 3-year longitudinal study based on a coding curriculum for Grade 4 to Grade 6. This 3-year study aims to gain an understanding of how coding activities helps children at upper primary school ages develop CT for problem solving, and if this problem-solving mindset is transferrable to other contexts, for example, daily life scenario and mathematical field. A part of the research focuses on a comparative study between students' intrinsic motivation of coding (i.e. "plugged") activities and "unplugged" activities (i.e. learning concepts from computer science through paper-based games without programming), which helps us understand the advantages and disadvantages of different learning approaches to CT. At the end of the first year, around 600 fourth graders" were assessed with their CT competence and affective tendency through: (1) a pre/post-test on the CT skill and mathematical ability; (2) a questionnaire on intrinsic motivation in plugged and unplugged activities. The preliminary findings indicate that students have gained a good programming-related knowledge and improved problem-solving skills through our curriculum. They were intrinsically motivated to participated in both types of activities. However, students did not show the ability to transfer computational thinking to mathematical context.

Using coding education to promote computational thinking and nurture problem-solving skills in children has become an emerging global trend. However, how different input and output modalities in coding tools affect coding as a problem-solving process remains unclear. Of interest are the advantages and disadvantages of graphical and tangible interfaces for teaching coding to children. We conducted four kids coding workshops to study how different input and output methods in coding affected the problem-solving process and class dynamics. Results revealed that graphical input could keep children focused on problem solving better than tangible input, but it was less provocative for class discussion. Tangible output supported better schema construction and casual reasoning and promoted more active class engagement than graphical output but offered less affordance for analogical comparison among problems. We also derived insights for designing new tools and teaching methods for kids coding.

In this paper, we present our preliminary results of identifying serendipitous findings from discussion forums of students by using a text-mining analytical tool to predict their academic performances. The analytical results were visualized by constructing KeyGraphs so that teachers can assess the effectiveness of teaching and innovation of learning respectively through the visualization of hidden patterns in the online learning environment. Our results show that the serendipitous findings have shown a traceable pattern, which is statistically significant to predict the academic performance of students. The research findings can lead to adaptive pedagogical designs for teaching and learning by finding hidden patterns and linkages among the students' serendipitous learning. The identified results are expected to support both teachers and students on how to improve teaching and learning with feedbacks from this new tool. Ultimately, this creates a new approach for transformative learning and teaching in education by using the advanced mining technology to assess the students' knowledge discovery process.

Computational Thinking (CT) has been widely introduced and investigated in recent years, particularly in the U.S. since the born of visual, block-based, drag-drop programming environments such as Kodu, Scratch, Minecraft and App Inventor. Although the user interface is mainly in English, the characteristics of these easy-to-use, game-based, and interactive tools attract many teachers and researchers in the world to pay much attention to the possibilities and opportunities of introducing these tools to students. Recently, some primary school teachers in Hong Kong begin to independently introduce some of these programming tools to students at age 7 - 11 as a part of learning activities in their computer lessons. Their motives are similar but not the same, such as making a fun learning and teaching experience, motivating students for active and collaborative participation, and introducing CT concepts to develop generic skills (e.g. problem solving skills, creativity, and critical thinking). However, there is an absence of well-developed and planned curriculum for "coding education" to introduce computational thinking systematically to students in the local context with expected learning outcomes. Due to the uniqueness of K-12 curriculum in Hong Kong, the existing curriculum model in the U.S. may need to be customized and redesigned to become suitable for integrating into the curriculum in Hong Kong. In this poster, it describes the first proposed coding education curriculum in Hong Kong primary education (Primary 4 to Primary 6) with relevant objectives, structures, contents, and learning outcomes. A new pedagogical design framework for CT is introduced in this poster, which could be generalizable and yet to be evaluated. This new curriculum will serve as the curriculum guide to local teachers, and is the first research initiative of a three-year longitudinal study investigating the impact of CT activities to students particularly in Hong Kong. The experience of this curriculum development for CT concepts in K-12 education can inspire teachers and researchers in other parts of the world when adopting and internationalizing CT activities based on the curriculum model developed under the U.S. education.

The role of computer and information literacy in the contemporary digital society is as important as language proficiency. This is widely exemplified in the graduate requirements in many tertiary institutions: students must pass the compulsory IT subject(s), often as a mandatory General Education (GE) course, before graduation irrespective of majors. Challenges arise to the computer science faculty as students are from different academic backgrounds with diversified level of IT background, and are often less motivated than taking a course in the major program. In this paper, we attempt to explore the reasons fostering students to devote efforts (i.e., persistence) in the GE IT course in tradition confucian examination-orientated educational settings. Student behavior and perceptions in two major groups of ethnically Chinese students (Mainland China and Hong Kong) are analyzed, revealing that mastery is the main reason leading to persistence in the mandatory IT course. Theoretically, we address the paucity on GE IT related research on ethnically Chinese students through validation and comparison of findings from the literature, leading to new directions of research. Practically, we generalize our experience from practitioner's view and propose different suggestions to effectively deliver GE IT courses under the confucian culture.

Digital technologies have been introduced for the purpose of enhancing the learning and teaching effectiveness, and many tools are designed specifically for mathematics education. However, whether pre-service teachers will actually intend to adopt the software or not in their future workspace is an important question. Even if they do, it is important to study the reasons behind their decision in this particular local context, and how the training curriculum can assist them in fulfilling their teaching goals. In this paper, our aim is to study the technology acceptance of our pre-service teachers in primary mathematics education, and investigate the incentives behind their decision of adoption in our local context based on their technology-enhanced learning experience in our Institute. Using multiple-regression analysis, we examine the factors influencing the technology acceptance of the pre-service teachers in their future teaching career. Our goal is to provide preliminary insights into how technology is perceived by these pre-service teachers in mathematics education training and career. This insight will help us build a better analytical model for a more formal analysis in our future study that fits both local and global contexts.

Abstract To prolong the battery lifetime, it is important to continue designing a better energy efficient mechanism for different mobile technologies. Most of the existing works on the IEEE 802.16e sleep mode operation focus on the decision making before a mobile station switching to sleep mode state. The correlation of the decision is mainly on when and how to sleep based on the traffic demands. After the mobile station is switched to sleep mode, the deactivation of it mainly depends on new incoming traffic regardless of the actual amount. Truly, frequent switching can increase the energy cost on the mobile station, which can significantly reduce the battery lifetime. To minimize the switching frequency, we propose a novel approach to resolve this issue by making a heuristic decision during the listening interval. With this aim, we propose a real‐time heuristic algorithm, WAKSLP_DECISION, to accommodate our target. Three main decision criteria are analyzed and designed, namely the probability of buffer overflow, expected delay violation, and battery lifetime expiry, to achieve our goal. We verify the energy consumption performance with simulation experiments to validate our proposed scheme. The result shows that our scheme performs 25–30% better compared with the original standard in terms of energy consumption. We believe this algorithm is practical and implementable without changing the original standard, which can contribute both in the research community and industrial development. Copyright © 2009 John Wiley & Sons, Ltd.

The IEEE 802.16e mobile WiMAX technology aims to provide with an energy efficient communication platform for various mobile applications. In the standard, the sleep mode feature with three Power Saving Classes (PSCs) is designed to compensate for the power saving as the target. Prior work has shown that the proposed Markov Decision Process (MDP) approach can achieve the optimal performance in terms of energy consumption and packet delay through the optimal PSC selection. In this paper, we further consider the design of when to sleep and how to sleep in the system to achieve better tradeoffs between the switching frequency during idle mode and mean power consumption in overall. First, we present our optimal timeout scheme designed based on the previous proposed MDP approach, which can help the system determine when to sleep as well as how to sleep optimally with less switching frequency intuitively. To guarantee the optimality, we show the equivalency of the MDP approach and our optimal timeout scheme that both can achieve the optimal performance in terms of energy consumption level. Also, we demonstrate the performance of our proposed scheme through numerical analysis and validate it with simulation experiments using ns-2 compared with the MDP approach. Finally, we evaluate the impact of Poisson and non-Poisson traffic on our proposed optimal scheme.

To prolong the battery lifetime, it is important to continue designing a better energy efficient mechanism for different mobile technologies. Most of the existing works on the IEEE 802.16e sleep mode operation focus on the decision making before a mobile station switching to sleep mode state. The correlation of the decision is mainly on when and how to sleep based on the traffic demands. After the mobile station is switched to sleep mode, the deactivation of it mainly depends on new incoming traffic regardless of the actual amount. Truly, frequent switching can increase the energy cost on the mobile station, which can significantly reduce the battery lifetime. To minimize the switching frequency, we propose a novel approach to resolve this issue by making a heuristic decision during the listening interval. With this aim, we propose a real-time heuristic algorithm, WAKSLP_DECISION, to accommodate our target. Three main decision criteria are analyzed and designed, namely the probability of buffer overflow, expected delay violation, and battery lifetime expiry, to achieve our goal. We verify the energy consumption performance with simulation experiments to validate our proposed scheme. The result shows that our scheme performs 25% to 30% better compared with the original standard in terms of energy consumption. We believe this algorithm is practical and implementable without changing the original standard, which can contribute both in the research community and industrial development.

Energy efficiency in mobile devices remains a critical issue in the design of wireless communication system. In order to provide a high quality of service (QoS), the IEEE 802.16e-2005 standard supports three unique power saving classes (PSCs) which aim to reduce the power consumption of mobile devices based on the different types of traffic. Our goal is to design an optimal sleep mode selection scheme so as to maximize the energy efficiency in a mobile WiMAX system while providing a certain QoS guarantee. In this paper, we propose a theoretical framework based on the semi-Markov decision process along with a performance evaluation on the sleep mode operation. Based on this model, we formulate a stochastic optimization problem and solve it through our novel policy optimization algorithm, which is capable of finding the optimal policy for the selection of different PSCs. The numerical and simulation results demonstrate the validity of our proposed scheme under different QoS requirements such as the packet delay and energy consumption level. We believe that our scheme has achieved a high energy efficiency and is best fit to the IEEE 802.16e mobile system.