Order	Time	ID	Full information
1	16:15-16:30	IT1062	Gamified Teaching for Marginalized Group Engagement: A Case Study in Early Childhood Education Author(s): Shuangqi Li, Li Li Presenter: Shuangqi Li, Shanghai Normal University Tianhua College, China Abstract: This study focuses on the participation dilemmas of marginalized groups in preschool education courses within higher education institutions, with the “Early Childhood Play and Guidance” course as a practical vehicle. It systematically explores the effectiveness of gamified teaching in promoting active participation among these groups. Based on self-determination theory, differentiated instruction theory, flow theory, and inclusive education theory, this research constructs a tripartite gamified teaching model comprising a multi-collaborative participation pathway, a tiered task adaptation system, and a comprehensive incentive evaluation mechanism. Through various methods such as online challenges, offline role-playing tasks, and mixed narrative activities, the model adapts to the diverse needs of marginalized students, including those from different disciplines, introverted individuals, and those with special needs. The practical results indicate that gamified teaching significantly enhances students’ course engagement, mastery of knowledge and skills, and overall competencies, particularly aiding marginalized groups in transitioning from passive observation to active participation. The study recommends strengthening teachers’ gamified teaching capabilities, establishing school-enterprise collaborative practice mechanisms, and further optimizing personalized support and long-term tracking systems. The findings of this research provide innovative practical and theoretical references for the reform of preschool education courses and contribute positively to the international practice of inclusive education concepts.
2	16:30-16:45	IT1063	Building an Inclusive Framework for Chinese Language Learning Supported by Generative AI: A Case-Based Analysis Author(s): Jiexuan Zhu, Yu Zhao, Ana María Pinto-Llorente Presenter: Jiexuan Zhu, University of Salamanca, Spain Abstract: In recent years, the global demand for Chinese as a Second Language (CSL) education has grown rapidly, driven by China’s increasing cultural and economic influence. At the same time, the advancement of Generative Artificial Intelligence (GAI), particularly large language model (LLM)-based tools, has introduced new opportunities for personalized, interactive, and inclusive language learning. Despite the potential of GAI to enhance learner engagement, provide adaptive feedback, and support differentiated instruction, empirical research on its integration in CSL classrooms remains limited. Existing studies often focus on English or general higher education contexts, leaving a gap in understanding how GAI can address learner diversity, promote intercultural competence, and foster equitable learning in CSL settings. This study aims to bridge this gap by analyzing five empirical cases of GAI-assisted CSL teaching and proposing a GAI-Supported Inclusive CSL Framework, which integrates learner characteristics, inclusive pedagogical practices, and AI-enabled technologies. The framework seeks to provide both theoretical guidance and practical reference for enhancing inclusivity, personalization, cultural responsiveness, and ethical implementation in CSL education.
3	16:45-17:00	IT2172	LAVA-Era Learning Method Infrastructure: An AI-Supported Socratic and Spaced Practice System Author(s): Ya Li, Xin Xie Presenter: Ya Li, Sichuan Technology and Business University, China Abstract: University students today face a crisis of "outsourcing thinking" to generative AI, compounded by extended yet volatile careers. We synthesize these environmental pressures as the LAVA conditions: Longevity, AI & Acceleration, Volatility, and Adaptability. Existing digital transformation in education primarily focuses on content infrastructure, merely providing access to knowledge without ensuring cognitive depth. This is insufficient to meet the demands of the AI era. To bridge this gap, this paper proposes a "learning methodology infrastructure": an institutional-level framework that puts into practice methods such as Socratic questioning, spaced retrieval, and knowledge pruning through a learning system embedded with AI technology. Unlike traditional AI tutors, this infrastructure acts as a mechanism that enforces effective learning behaviors as the default study workflow. A design-based, quasi-experimental deployment in a Chinese applied university compared a 2023 baseline cohort with 2024–2025 cohorts using the system. Empirical results indicate a significant shift toward future-oriented capability building: the intervention cohorts earned 62 professional certificates, obtained 23 software copyrights, and achieved 477 competition awards, substantially exceeding the baseline. These findings suggest that shifting investment from content delivery to method infrastructure is a necessary path to foster sustainable learners in the LAVA era.
4	17:00-17:15	IT3240	Co-Design of Personalized Learning Paths with AI, Learners and Teachers: Dynamic Teacher Role Allocation in Intelligent Computer-Assisted Language Learning Systems Author(s): Dong Yu Presenter: Huisi Chen, South China Business College Guangdong University of Foreign Studies, China Abstract: In the rapidly evolving educational landscape of the AI era, this paper focuses on the co-design of personalized learning paths by AI, learners, and teachers, and elucidates the indispensable role of teachers amid the proliferation of intelligent educational technologies. Recognizing the transformative impact of AI on teaching methodologies and learning outcomes, this study redefines the dynamic role of educators in effectively integrating technology into instructional practices, with a specific focus on college English audio-visual-speaking courses. By comparing traditional teaching approaches with deep reinforcement learning (DRL)-driven pedagogy across three key listening stages, the research examines how teachers can adapt to intelligent tools while retaining their core function as learning facilitators and path designers. It emphasizes that AI should serve as a catalyst to enhance student engagement and academic achievement, rather than replacing the humanistic elements of education. The findings propose a pedagogical framework that leverages DRL to optimize personalized learning paths, with teachers playing a leading role in coordinating AI capabilities and learner needs to achieve collaborative design of efficient learning trajectories.
5	17:15-17:30	IT3295	Exploration of optimizing graduate curriculum system based on the concept of personalized learning using particle swarm optimization algorithm Author(s): Tingyu Li Presenter: Tingyu Li, Shenyang University, China Abstract: This study aims to address the inadequacy of traditional postgraduate curriculum systems in meeting personalized learning needs. A multi-objective optimization model was constructed, taking individual demand adaptability, curriculum rele-vance, and resource utilization as core objectives, with constraints on total credits and curriculum type ratio. An improved Particle Swarm Optimization (PSO) al-gorithm was proposed, incorporating dynamic inertia weight and adaptive learn-ing factors, and compared with basic PSO and Genetic Algorithm (GA) via MATLAB simulations. Results show the improved PSO outperforms the other two algorithms in convergence speed 40% faster than basic PSO and accuracy 10.0% higher than GA, with more stable performance. The optimized curriculum system significantly improves key indicators: 18.3% in demand adaptability, 42.1% in curriculum relevance, and 18.1% in resource utilization, while satisfy-ing academic constraints. This model and algorithm provide a scientific basis for personalized postgraduate training, offering practical value for optimizing higher education curriculum systems.
6	17:30-17:45	IT3399	Designing lesson plans for an online English class with an inclusive setting involving college students with and without autisminclusive Author(s): Afifah Muharikah, Rania Chairunnisa Qisti Presenter: Afifah Muharikah, Universitas Islam Internasional Indonesia, Indonesia Abstract: With the increasing number of individuals with autism, many second language (L2) instructors at tertiary institutions are likely to teach inclusive classes in the future. These instructors need to be informed regarding how to teach L2 in inclusive settings. Using an autoethnography approach, this study illustrates how the inclusive learning plans were designed for an online English course in a vocational college in Indonesia during the first author’s doctoral data collection. The study presents the first author’s reflections of the steps she followed when interpreting the general curriculum into inclusive lesson plans, while it also provided an external interpretive analysis of the pedagogical decisions through the lens of the Universal Design for Learning (UDL) framework. To support the analysis, the first author’s observation notes during the online classes and recordings of the sessions were used to examine the students’ responses to the adaptation. In this study, the inclusive lesson plans were designed to cover the specific needs of the autistic students by including detailed scenarios, pre-class assignments, and peer interaction-based class activities. Both autistic and non-autistic students appeared to perceive the lesson plans as beneficial. The captured interactions also indicated that the lesson plan mediated the process of scaffolding and encouraged self-regulation in inclusive peer interactions.
7	17:45-18:00	IT3401	Research on the Construction and Application of an AI-Based Personalized Intervention Model for Early Childhood Games Author(s): Rui Ren Presenter: Rui Ren, Zaozhuang University, China Abstract: This paper addresses issues in traditional early childhood game interventions—such as subjective observation, inefficient recording, and delayed intervention—by constructing an AI-based personalized intervention model. It aims to drive the educational paradigm shift from “experience-driven” to “data and intelligence dual-driven” through human-machine collaboration. Integrating techniques from the Zone of Proximal Development, adaptive learning, multimodal integration, and deep learning-based behavior recognition, the study establishes a comprehensive four-layer framework encompassing data acquisition, intelligent analysis, intervention decision, and application service. This model employs multimodal perception networks to capture real-time behavioral data, utilizes deep learning algorithms for contextual analysis and profiling, and integrates expert rules with reinforcement learning to generate personalized intervention strategies. Through case analyses across four dimensions—environment and materials, interaction and guidance, evaluation and reflection, and home-school collaboration—the study validated the model’s effectiveness in enhancing teacher observation objectivity, intervention precision, and support for children’s personalized development. It provides a viable technical pathway and practical example for fostering personalized development in early childhood education.
8	18:00-18:15	IT4421	GenAI in Pre-Task Planning: A Comparative Study of Oral Fluency, Accuracy, and Complexity in Business English Speaking Author(s): Xiuzhen Xiang, Yi Li Presenter: Xiuzhen Xiang, Wuhan Business University, China Abstract: The importance of pre-task planning(PTP) on students’ output has been acknowledged, so is the positive effect of AI in language learning.However seldom scholars study the different effects between AI and conventional strategy in PTP. To build the gap this study investigates the impacts of AI-assisted and conventional PTP on business English speaking. The research involved 50 sophomore business English major students at a university in China.The experimental group use AI-assisted material while the control class followed traditional textbook- based method during PTP. The findings reveal significant better performance in the experimental group’s speaking fluency and accuracy than the controlled group. But there is no difference in complexity.
9	18:15-18:30	IT4474	Rethinking AI Literacy in L2 Writing: Empirical Evidence from Chinese University Students Author(s): Zhao Fang, Stephen E. Sandelius Presenter: Zhao FANG, East China University of Science and Technology, China Abstract: This study examined the current status of AI literacy in L2 writing among Chinese university students and explored how existing AI writing literacy frameworks and measurement scales could be refined. Adopting a mixed-methods design, the study drew on quantitative data from a questionnaire survey for hierarchical regression analysis and qualitative data from reflective journals for thematic analysis. Results indicated that while students generally reported moderate to high levels of AI writing literacy, only the Understanding and Use dimensions significantly predicted their performance in AI-assisted writing tasks. Evaluation and Ethics, though rated positively, showed limited influence on actual writing outcomes. Thematic analysis further revealed students’ strong reliance on AI for language-level support but a lack of attention to discoursal or ethical aspects. More notably, it also disclosed the underestimation of the nuances of affective components and the overlooking of regulatory and regional constraints within the existing AI literacy frameworks. The study calls for a more student-centered approach to conceptualizing and assessing AI writing literacy, one that better reflects learners’ lived experiences and diverse needs.