Exploring the influence of video viewing action on learning performance by using sequence mining.

• Main Authors: Po-Wen Shih, 施博文
• Other Authors: Chen-Hua Yang
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/8zu4d9

碩士 === 國立中央大學 === 資訊工程學系在職專班 === 107 === From Lag-sequential Analysis, by analyzing the differences in learning behavior between high achievement and low achievement, it is helpful to find out which learning behavior is helpful for improvement. ABSTRACT With advances in technology, Internet has gained become more popular. Online education opportunities have been growing in recent years. Unlike traditional studies, online learning offers great flexibilities and convenience. To begin with, you don't have to waste time and money traveling to the campus. You can also select a program that entirely fits with your interests and needs because you are not restricted to the classes that are offered in your area. Because students can use many functions in Massive Open Online Courses (MOOCs), such as watching movies, watching documents, quizzing exams, fill in questionnaire, etc. These behaviors are recorded by the MOOCs system. If they are only recorded but not analyzed, they are too wasteful. The most commonly used behavior is watching a learning video. Therefore, this method observes the operation of the students watching the learning video on the MOOCs and try to analyze whether the student's operational learning model will affect the student's test scores. This research uses Correlation coefficient, Multiple Factor analysis, Exploratory factor analysis s and Lag-sequential Analysis to find out the key viewing learning modes that have an impact on students' learning performance. Among them, Lag Sequential Analysis is used to find out the difference between the learning behavior between high and low achievement students and the influence of difference learning performance is found through by Correlation coefficient. Use Exploratory factor analysis to reduce the number of patterns found and find out the learning mode. Use Multiple correspondence analysis to find the most influential viewing mode patterns for overall student performance. Use the analysis results provided by the research, teachers can find out the students who have poor learning patterns and help them to catch up with the learning schedule. Keywords: learning analysis, MOOCs, Video learning behavior, Correlation Coefficient analysis, ANCOVA, Lag Sequential Analysis, Exploratory Factor Analysis, Multi Factor Analysis