【摘要】 本文提出并实验了中断学习风险预测模型，该模型通过特定的学习者活动特征，可以根据学习者在线开放课程学习过程的数据，寻找可能导致学习者终止学习或长时间中断学习的迹象。实验结果表明，对于处于学习状态的学习者，该模型的正确预测率在40%到50%之间。对于中断学习14天或更长时间的学习者，正确预测率可以提高到90%以上，该预测模型具有一定的实用性和有效性，将有助于降低辍学率，促进在线开放课程的可持续发展，使课程价值最大化。更多还原

【Abstract】 This paper proposes and experiments a risk prediction model for interrupting learning. This model can look for signs that may cause learners to stop learning or interrupt learning for a long time according to the data of the learning process of learners’ online open courses through specific learners’ activity characteristics. The experimental results show that the correct prediction rate of the model is between 40% and 50% for learners in learning state. For learners who have interrupted learning for 14 days or more, the correct prediction rate can be increased to more than 90%. This prediction model has certain practicality and effectiveness, which will help reduce the dropout rate, promote the sustainable development of online open courses, and maximize the value of the courses.更多还原