Temporal Models for Personalized Grade Prediction in Massive Open Online Courses
With the booming of massive open online courses (MOOCs) in recent years,online educational pattern,both inside and outside of the campus community,have being become more popular.However,high attrition rate and low completion rate remain major impediments.How to reduce the dropout rate and improve the learning effect,with providing the personalized feedback to teaching and learning by forecasting the personal grade,has become the focus of the research of MOOCs.In this paper,a personalized grade and learning trend prediction method is proposed.A supervised classification method is applied to analyze the online learning behavior.According to the characteristics of online learning behavior changing with time,a long short-term memory recurrent neural network is employed to construct a learning trend prediction model,which predicts whether the learner will be online in the subsequent weeks (the possibility of returning class) and the final grade.The evaluation of the method was based on the data from five multi-disciplinary courses on icourse163 platform.The test results showed that the average AUC of the dropout prediction was 0.863 and the average AUC for predicting the final grade was 0.748 respectively.Compared with the baseline methods (Logistic Regression and SVM),this presented method increases the average AUC of the dropout prediction by 3.2% and 3.5% as well as the average AUC of the final grade prediction by 6.1% and 7.7%.Early warning and timely feedback would be valuable on MOOC,which is a promising way to optimize the design and organization of a course and to improve the learning effectiveness.
Supervised learning Grade prediction Neural Network MOOC
Qi Qi Yuexia Liu Fan Wu Xiangguo Yan Ning Wu
School of Electronic and Information Engineering Xian Jiao Tong University Xian, China The Third Research Institute of the Ministry of Public Security Shang Hai, China School of Life Science and Technology Xian Jiao Tong University Xian, China
国际会议
2018中国图灵大会(ACM Turing Celebration conference-China 2018)
上海
英文
67-72
2018-05-19（万方平台首次上网日期，不代表论文的发表时间）