| 摘要: | 本研究旨在設計並實作一套整合多模態資料分析與人工智慧助理的閱讀支
持系統，期望促進學生閱讀習慣的建立，並強化家長在家庭閱讀中的參與及引
導角色。相較於傳統閱讀紀錄大多僅呈現書籍數量與簡要文字回饋，難以深入
理解學生實際的閱讀行為與情境，本研究整合學生每月的閱讀書籍登記資料、
家長與教師上傳之閱讀照片與說書影片，以及親師回饋表單，建構出一套具多
面向觀測能力的閱讀資料分析架構。系統生成涵蓋閱讀行為面、情意面與認知
面的 AI 分析報表，並依據學生表現進行個別分群與趨勢比較，協助教師與家長
掌握學生閱讀發展樣貌。
為提升家長參與的便利性與回應品質，本研究開發了「AI 個人助理」模
組，依據每月生成的分析報表與親師回饋表單內容，提供家長個別化的閱讀建
議與學生閱讀狀況說明。系統經歷了兩階段版本測試：第一階段為家長主動提
問型，第二階段加入了 AI 主動引導功能，以提升親子對話的頻率與深度。研究
對象為參與「明日閱讀習慣暨家校閱讀博覽會」之國小二至五年級學生之家
長，共 19 名。研究採用設計研究法，並針對有意願接受訪談的家長進行兩輪半
結構式訪談，第一階段訪談 10 人，第二階段追蹤 6 人，深入探討系統實際使用
後的體驗與家庭閱讀互動的改變。
研究結果顯示，多數家長肯定 AI 閱讀分析報表對理解孩子閱讀狀況的幫
助，部分家長將其視為家校溝通與選書建議的重要依據；AI 個人助理因具備互
動性與個別化回應能力，在部分家庭中有效促進了閱讀活動與學生閱讀行為的
正向轉變。部分家長根據 AI 的建議進行調整後，發現學生的閱讀頻率顯著增
加，且對書籍的接受度也有所提升。本研究初步驗證了 AI 技術應用於家庭閱讀
支持系統的可行性，並指出未來可進一步優化使用介面與個別化回應策略，以
強化其在學生閱讀習慣養成中的實質效益。;This study aims to design and implement a reading support system that integrates
multimodal data analysis with an artificial intelligence assistant, with the goal of
promoting the development of students′ reading habits and enhancing parental
involvement and guidance in family reading. Compared to traditional reading records,
which mostly present the number of books read and brief written feedback, making it
difficult to deeply understand students′ actual reading behaviors and contexts, this study
integrates students′ monthly reading book registration data, reading photos and
storytelling videos uploaded by parents and teachers, and parent-teacher feedback
forms to construct a reading data analysis framework with multi-dimensional
observation capabilities. The system generates AI analysis reports covering reading
behaviors, affective aspects, and cognitive aspects, and groups students based on their
performance, conducting trend comparisons to help teachers and parents track students′
reading development.
To improve parental engagement, this study developed the "AI Personal Assistant"
module, which offers personalized reading suggestions and explains students′ reading
status based on monthly analysis reports and feedback. The system was tested in two
stages: the first with a parent-initiated inquiry model, and the second incorporating AI
driven guidance to increase parent-child conversations. The study involved 19 parents
of second- to fifth-grade students from the "Tomorrow′s Reading Habits and Home
School Reading Expo." Two rounds of semi-structured interviews were conducted, the
first with 10 parents and the second with 6, to explore their experiences and changes in
family reading interactions.
The results show that most parents recognize the AI reading analysis report′s
contribution to understanding their children′s reading status, with some parents
considering it an important basis for school-home communication and book
recommendation. The AI Personal Assistant, due to its interactivity and individualized
response capabilities, effectively promoted positive changes in reading activities and
students′ reading behaviors in some families. After making adjustments based on AI
suggestions, some parents observed a significant increase in students′ reading frequency
and improved acceptance of books. This study verifies the feasibility of applying AI
technology in a family reading support system and suggests that future improvements
could be made in the user interface and individualized response strategies to strengthen
its practical benefits in fostering students′ reading habits. |
