近年來，線上學習平台日益盛行，線上課程的多元化與豐富程度，學生要學習新知識，並不再侷限於面對面的課程。線上學習平台在國外已經相當的盛行，如Coursera、edX，以及美國知名學校麻省理工學院 (Massachusetts Institute of Technology, MIT) 也都擁有線上學習課程，而國內教育目前也積極地推廣，如Moocs磨課師學習平台，NCUX以及BookRoll線上電子書學習系統等等。 以BookRoll學習平台為例，授課教師將教材轉換成電子書並上傳至平台上，學生上課則會透過平台去閱讀教師準備的教材，來達到學習的目的。BookRoll會記錄每一位學生的學習歷程，但這些資訊並不容易讓授課教師直接掌握每一位學生的學習狀態，來瞭解學生的實際學習狀況，以及時給予協助。 因此本研究分析建立於BookRoll線上學習平台，分析學生學習時，所記錄的歷程與動作。透過方法論，將學習歷程轉換為學習序列，經由統計的方法，分析每位學生的學習行為與動作，來探討其與學習成效之間的關聯性。並藉由Apriori演算法，使用滑動視窗搭配關聯規則分析，尋找高分群學生共同的良好學習樣式，及這些學習樣式對學習成就的影響。;In this few years, online learning, or virtual classes offered is getting more popular, because online learning environments provide a greater degree of flexibility than traditional classroom settings and online platforms can also offer more diverse representations of student populations as learners. In Taiwan, online learning environments are also promoted positively, such as TAIWANMOOC、NCUX and BookRoll, BookRoll is an online eBook learning system, Teachers can convert course contents into online e-books, and BookRoll can collect the reading logs of students. This study is to analyze reading logs on BookRoll, use statistical methods to find out the relationship between reading actions and learning performance, and explore the learning patterns via Apriori algorithm, to use sliding window and association rules to find out which learning patterns are great behaviors on learning.
