E-learning Behavior Analytics in the Curriculum of Big Data Visualization Application

There are two core professional training objectives in the Department of Distribution Management in this study's case university: logistics management and business stores management. To enhance the efficiency of decision makers using complex data, management science and information technology application is vital. As visualization images are more easily understood than complicated statistical analyses in reports, the interactive and visual analytics system, Tableau, is applied in the big data learning and analytics curriculum. Through feedback it has been found that due to the differing skill levels, some students could not follow the sets outlined in face-to-face learning; therefore, the instructor will offer remedy education using blended learning in the new semester. The two facets of blended learning are face-to-face and online teaching. It is necessary to analyze the behavior portfolios of the learners to intelligently modify the remedy education strategy and enhance the quality of e-learning. During February to June 2021, 39 junior and senior undergraduates in the big data class participated in this study. In the case university, the instructor uploaded the teaching materials in the e-learning platform so that students could check the operational videos after in-class demonstrations. According to the participants' background variables and e-learning behavior portfolios, descriptive and inferential statistical analyses were performed. The independent variables are gender, group role, login frequency, class attendance frequency, discussion frequency, reading seconds, reading pages, and skill difficulty. The dependent variable is the final report score of the student groups. Hotspot analysis, basic statistics, ANOVA, correlation, and C5 decision tree analyses were performed in this study. The results show that the e-learning teaching materials were beneficial to the students who needed extra assistance after class. Being female, the frequency of login, class attendance, and discussion were significantly positively correlated with the final report scores of the students. The variables of skill difficulty and discussion were the two main variables identified to predict high scores in the curriculum of big data. These findings can help instructors improve their pedagogy and enhance students' e-learning performance.