Abstract
In recent years, MOOCs has enjoyed great popularity due to its convenience and openness. However, with the development of MOOCs, the high dropout rate has aroused extensive attention. By analyzing the data of students' behavior and then predicting whether students are at risk of dropout, it can improve the course completion rate. Most of the existing methods relying on feature engineering and the sequential characteristic of data is not effectively utilized. In this paper, we propose a time series model named CNN-LSTM-ATT, which focuses more on local valid information and temporal information of the data. Through extensive experiments on a public dataset, it shows that the proposed model can effectively predict students' dropout behavior.
Export citation and abstract BibTeX RIS
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.