教育大数据下基于离群检测的教学预警模型

针对在线教学过程中教学方案缺乏针对性、教学效果评估和反馈缺乏时效性的问题,在量化数据与相似度排序的基础上提出基于学习过程和学习成绩的分析算法,通过规则检测与离群检测建立了交互式教学系统中的预警模型,实现了基于学习行为分析的自动干预和主动干预。预警信息可视化结果表明,参与交互的各方均能够直观、高效地获取预警信息,对学习效果和学习成功率的提升起到了促进作用。

Early Warning Model Based on Outlier Detection under Background of Big Data on Education

In view of the lack of pertinence in teaching plan and timeliness in teaching effect valuation,an early warning model of interactive teaching system is proposed based on rule detection and outlier detection． Analytical algorithms based on the learning process and academic performances are designed for automatic intervention and active intervention based on learning behavior analysis． After information visualization,the result shows that all parties involved can acquire early warning information intuitively and efficiently,which can promote the improvement of learning effect and learning success rate．