Abstract
Precision Education (PE), an individualized approach to learning based on learners’ characteristics and needs, has been extensively explored in medical and STEM education. However, how to utilize Artificial Intelligence (AI) to improve PE for language learning, especially for Foreign/Heritage Language Learners (FLL/HLL), remains less understood. These challenges are evident for learners of non-Latin-based scripts, such as Chinese, due to the heterogeneity of HLL learner backgrounds and the mismatch in literacy levels within one-size-fits-all instructional settings. This study aims to address this disparity by adapting the PE principles of learner-centered personalization and targeted intervention to directly address the learning needs of Chinese HLLs through Large Language Models (LLMs), such as ChatGPT. This research employs a Dual-layer Prompt-Engineered Model, consisting of Prompt-Engineered Learning Protocols (PLP) and Prompt Engineering Techniques (PET). The PLP outlines topical, grammatical, and vocabulary requirements for a passage that the LLMs generate, grounded in the existing curriculum. The PET offers structured patterns for students to tailor their interactions with the PLP texts based on individual students’ engagement patterns, identity, prior knowledge, and literacy levels. Using Chinese Foreign Language Learners (CFLLs) as a comparison, this study examines how the PET can customize the LLM’s output to suit CHLL’s cultural needs. Findings demonstrate that AI guided by PET can generate more personalized language content, learning strategies that differentiate between CHL and CFL learners. Functioning as both a technical tool and a pedagogical strategy, this Dual-layer model expands the understanding of the potential for PE in AI-assisted learning for many foreign languages.
Presenters
Xiang ZhangAssistant Professor of Chinese, Modern Languages, The University of Alabama, Tuscaloosa, Alabama, United States
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
KEYWORDS
Foreign/Heritage language; Prompt Engineering; Chinese; Precision Education; AI
