生涯探探-自主學習專案引導系統
講者/Speaker： 江彧璇, Jiang, Yu-Syuan
本設計以「生涯探探－自主學習專案引導系統」為題，針對中學階段學生在自主學習與生涯探索中常見的選題困難、行動起點不足、執行鷹架缺乏與個別化引導不足等問題，運用生成式 AI 建構一套以生涯輔導理論為基礎，並結合專案學習歷程的學習引導系統。系統以國中生涯發展教育為起點，並可延伸支持高中階段之自主學習與學習歷程檔案整理，協助學生從自我覺察、職業探索、資訊查核、專案構思、資源盤點、計畫拆解到歷程記錄，逐步完成具個人意義與生涯發展連結的自主學習專案。
本設計運用 TPACK 架構，並從納入 AI 知識與人機互動設計的 AIPACK 教學應用觀點出發，讓 AI 透過分階段提問、任務生成與回饋，引導學生思考自己的選擇、查核資訊、修正方向並落實行動。其中，「真理之秤」關卡特別強調 AI 時代的媒體識讀與批判思考，避免學生直接接受生成內容，並培養查證、比較與重新詮釋的能力。
課程實施資料顯示，隨著課程結構化程度與 AI 導入深度增加，學生專案成果提交率呈現提升趨勢，由第 1 屆的 64% 提升至第 4 屆完整 AI 系統導入後的 94%。學生作品亦由早期的片段任務，逐步發展為具完整歷程、成果作品與公開展示的專案學習成果。教師亦能藉由 AI 所提供的個別化鷹架，降低重複性基礎指導負荷，進而更精準地掌握學生差異並提供關鍵引導。
關鍵字： 生成式 AI、自主學習、生涯探索、TPACK、AIPACK、學習歷程、專案學習
This project, titled Career Explorer: An AI-Supported Guidance System for Self-Directed Learning Projects, addresses common challenges faced by secondary school students in self-directed learning and career exploration, including challenges in formulating project topics, lack of action-oriented starting points, insufficient implementation scaffolding, and limited individualized guidance. By using generative AI, this design constructs a learning guidance system grounded in career guidance theory and integrated with project-based learning processes. Initially implemented in junior high school career development education, the system can also be extended to support self-directed learning and learning portfolios in senior high school. It guides students through self-awareness, career exploration, information verification, project ideation, resource planning, task decomposition, and process documentation, supporting them in completing self-directed learning projects that are personally meaningful and relevant to students’ career development.
Grounded in the TPACK framework, this design extends it through an AIPACK perspective by incorporating AI knowledge and human-AI interaction into instructional design. Through staged questioning, task generation, and feedback, AI guides students to reflect on their choices, verify information, revise directions, and take action. In particular, the “Truth Scale” stage requires students to verify, compare, and reinterpret AI-generated information. This stage emphasizes media literacy and critical thinking in the age of AI, helping students avoid directly accepting generated content.
Implementation data show that as the course became more structured and AI integration deepened, the student project submission rate showed an upward trend, increasing from 64% in the first cohort to 94% following the implementation of the full AI-supported system in the fourth cohort. Student project outcomes also evolved from fragmented tasks into project-based learning outcomes with complete processes, final products, and public presentations. In addition, the individualized scaffolding provided by AI reduced the burden of repetitive basic guidance, enabling teachers to better identify individual student needs and provide targeted guidance at key learning moments.
Keywords: generative AI, self-directed learning, career exploration, TPACK, AIPACK, learning portfolios, project-based learning
- 論文全文：生涯探探-自主學習引導系統
- 影像： 相關影片