近年來人工智慧迅速發展，將人工智慧應用在教育領域幫助學習已成為熱門討論的研究話題。過去許多研究已經表明人工智慧能夠在許多方面幫助學習，但仍然缺乏學習者可以直接向人工智慧互動進而影響學習的相關研究。在這項研究中，開發了一款手機應用程式稱為 Smart 3D-UG，此系統可以幫助小學生利用手機測量和計算周遭環境中的真實物體，透過在真實情境中的學習認識立體幾何物體並應用在生活中，並且在系統中設計了智慧回饋機制，能提供學習者一對一的實時引導和互動。本研究的目的是探討學習者在智慧回饋的輔助下進行真實情境式學習對立體幾何能力的能力影響，進一步分析哪些學習行為會影響學習成績。本研究將桃園市某國小共四十七名六年級學生分為兩組進行為期八週的教學實驗，實驗結合幾何內容設計了四種學習主題活動和兩種學習任務來幫助學習。相關研究結果表明，實驗組學習者使用具有智慧回饋機制的 Smart 3D-UG 系統學習後在 3D 幾何學習成效上優於使用不具有智慧回饋機制的 Smart 3D-UG 系統的控制組學習者，同時實驗組學習者也在計算活動的答案正確率上優於控制組學習者。並且學習者使用智慧回饋的次數以及向智慧回饋提問的次數越多，和幾何能力呈現高度相關。這意味著結合智慧回饋的真實情境學習能有效幫助學習者的立體幾何學習。最後根據學習者的飯潰，多數學習者認為智慧回饋機制和真實情境中學習能夠幫助他們學習立體幾何概念並實際應用，同時能有效降低學習焦慮。;In recent years, the rapid development of artificial intelligence has made applying AI in the educational field to assist learning a popular research topic. Numerous studies in the past have shown that AI can help learning in many ways, but there is still a lack of research on learners directly interacting with AI to influence their learning, especially in geometry learning. In this study, a mobile application called Smart 3D-UG was developed. This system helps elementary school students measure and calculate real objects in their surroundings using their phones. Through learning in authentic contexts, students can understand 3D geometry objects and apply this knowledge in daily life. The system also includes a smart feedback mechanism that provides real-time, one-on-one guidance and interaction for learners. The purpose of this study is to investigate the impact of authentic contextual learning assisted by smart feedback on learners′ 3D geometry abilities and further analyze which learning behaviors affect learning outcomes. In this study, 47 sixth-grade students from an elementary school in Taoyuan City were divided into two groups for an eight-week teaching experiment. The experiment combined geometric content with four learning themes activities and two learning tasks to aid learning. The results of the related research indicate that the experimental group using Smart 3D-UG with smart feedback outperformed learners using Smart 3D-UG without smart feedback in terms of 3D geometry learning effectiveness. The experimental group also had a higher accuracy rate in answering geometry calculations than the control group. Moreover, the frequency of learners using smart feedback and asking questions to smart feedback showed a high correlation with their geometry abilities. This implies that authentic contextual learning combined with smart feedback can effectively assist learners in their 3D geometry learning. Finally, according to learners′ feedback, most learners believed that the smart feedback mechanism and learning in real contexts could help them understand and apply 3D geometry concepts and effectively reduce learning anxiety.
