運用教育機器人與基于物聯網的實體物件建構沉浸式中文學習環境: 以學習量詞與把字句為例

Project Aim ‧ Develop immersive robot + IoT tangible (R&T) learning environments for Chinese as a Second/Foreign Language (CSL/CFL). ‧ Target two difficult areas: 1. Chinese classifiers (measure words) 2. The ba-construction sentence structure Yearly Outputs Year 1 (2022–2023) ‧ Paper 1: Learning Chinese Classifier with Robots and IoT-based Tangible Objects o Adult CSL learners (Taiwan), R&T vs. human teacher o Large learning gains in classifier recognition and production (d?1.33) o Low cognitive load, but tone accuracy still difficult ‧ Paper 2: Designing for Adult Language Learning in a RALL Context o Pilot of ba-construction “Making Dumplings” task using robot + IoT objects o Introduced Context–Problem–Solution–Diversification design-for-learning pattern Year 2 (2023–2024) ‧ Paper: Robot-Assisted Language Learning: Interdisciplinary Collaboration Case Study o Applied TPACK + Activity Theory for RALL design o Clarified knowledge integration and team workflows ‧ Patent: (Smart Storytelling Robot) o Captures user request → text → generative AI → robot narration o Applications in early childhood and language learning Year 3 (2024–2025) ‧ Paper 1: Pedagogical AI Agents as Digital Twins for R&T Learning o Embedded GPT-4.1-based Pedagogical AI Agents (PAIA) o 5-week trial (n=5) showed significant gains in listening, reading, speaking o Usability SUS?85 (Grade A); latency & ASR strictness noted ‧ Paper 2: Designing PAIAs for Teacher Training o Three 2-day training workshops, total 78 participants (HE/secondary/elementary) o TPACK survey gains + positive adoption feedback o Highlighted teachers as designers of AI pedagogy ‧ Outreach: 3 national PAIA design training & promotion events Key Contributions ‧ Pedagogical Impact: Robots + IoT objects enable immersive, embodied learning of classifiers and grammar; outcomes comparable to human teaching. ‧ Tone Learning Challenge: Both R&T and teacher groups struggled; future work to integrate pitch-curve and gesture training. ‧ Collaboration Model: Established interdisciplinary workflow using TPACK/AT frameworks. ‧ Innovation: Patent for AI-driven storytelling robot; introduction of Pedagogical AI Agents as digital twins. ‧ Teacher Development: Promoted teacher-as-designer identity through structured training and workshops. ? Deliverables Across 3 Years: ‧ 10 academic papers ‧ 1 patent ‧ 3 teacher training workshops/outreach events