[1K5-GS-3c] Knowledge utilization and sharing

6:55 PM - 7:10 PM JST(9:55 AM - 10:10 AM UTC)

[1K5-GS-3c-06] Proposal of AI-Assisted Collaborative Abduction for Facilitating Somatic Knowledge Emergence

〇TAKUICHI NISHIMURA¹, Reina Hagiwara¹, Naoki Mitsugi¹, Tudai Wada¹, Kazuhisa Nakano¹, Yuto Kano¹, KoKi Ijuin¹, Ageha Kanazawa³, Saki Kanazawa¹, Yasuyuki Yoshida² (1. JAIST, 2. Health Promotion Support Office for the Elderly, Tokyo Metropolitan Institute for Geriatrics and Gerontology, 3. Saitama Municipal Kishicho Elementary School)

Keywords:

Wild Knowledge,Collaborative Abduction,Somatic Knowledge Emergence,Knowledge Structuring,Individualized Adaptive Learning

The transfer of somatic and tacit knowledge poses significant challenges in nursing care, medical practice, sports coaching, and other domains due to difficulties in formalization. This study proposes a novel system for facilitating the emergence of "Wild Knowledge" through triadic collaborative abduction among learners, instructors, and AI. The system integrates motion capture-based movement data with discourse analysis-based linguistic data, enabling AI to structure and present novel patterns and body-type-specific optimal teaching strategies that cannot be explained by existing instructional knowledge. We are planning empirical experiments targeting Rumba walk instruction in ballroom dancing, aiming to achieve improvements in center of gravity stability and discover unpredictable new somatic knowledge (e.g., emergent metaphors such as "sensing the temperature of the ground through the soles of the feet"). We construct a framework that enables individually optimized learning where learners formulate hypotheses based on their bodily resources at hand and integrate multiple perspectives from AI and instructors. This model has potential applications in nursing care, medical practice, and traditional skill transmission.