- 學位論文
學習歷程探勘與思考風格於適性化補救學習之研究
Applying Learning Portfolios and Thinking Styles to Adaptive Remedial Learning
摘要
線上學習不同於傳統課堂上的教學方式，在虛擬的網路中，學生與教師並無實際的接觸，因此教師很難確切掌握學生的學習情況。本研究藉由分析學生線上學習狀況，進一步掌握學生學習活動，發出適性化警訊來對症下藥。 線上學習歷程為龐大複雜的量化數據，若以傳統人力觀察的方式，需花費相當大的時間。因此本研究利用資料探勘相關技術，處理龐大的資料數據，整理有效的資訊提供給教師使用。在數據的處理上，我們採用集群分析，來歸類學習方式相仿的學生。在資訊提供方面，我們觀察各分群學生的學習歷程，歸納同群組學習的相依性，不同群組間的差異性。同時我們針對個別學生的學習活動類型發送適性化警訊，驗證警訊對學習成就的影響與關係。 本研究採用ANOVA檢定來分析學生在線上學習的活動表現、學生的學習成就及適性化警訊，三者間的相互關係。並且進一步觀察學生思考風格、線上行為及適性化警訊之間交互影響所產生的學習模式。 本研究透過上述之技術與方法，實作學習歷程分析系統。針對學生的線上各項活動進行探勘，提供了外部匯入的功能，增加線上學習外的學習因素。藉由本系統，教師可從中觀察學生的學習活動與演變，而利用觀察學習活動與學習成就的關係，視學生學習狀況，採取不同的輔導機制。
並列摘要
Online learning is different from the instruction provided in traditional classroom lecturing where the teachers and students cannot have actual contact. Thus, the teachers can have little control over the students’ learning situations. However, the teachers can observe the students’ learning activities by analyzing their online learning portfolios, understand students’ learning activities further, and then provide remedial instruction to students who have bad learning situations. Online learning portfolios are complex row data, it will cost much time for observation. So, this study uses data mining algorithm to analyze students’ online learning portfolios,and provide effective information to teachers. This study uses cluster analysis to categorized students with similar learning behaviors into the same group. This study aims at sending adaptive warning messages according to different students’ learning activities, demonstrates the relation between warning message and learning performance. This study used ANOVA evaluation to analyze the relation of students’ learning activities、learning performance and adaptive warning message. Obtaining students’ learning model by observing student’s learning activities, thinking styles and adaptive warning message. The correlations between these factors are also considered. This study implements learning portfolio diagnostic system to handle complex online learning portfolios.Thus, the students’ learning situations can be revealed by the aids of the online learning portfolio diagnostic system. Therefore, the system provides teachers interface to import additional students and curricular information, such as curricular schedule roll call data. By using this system, teachers can observe students’ learning activities and variations, and further find the relations of learning activities and learning performance .