Epistemic neural network based evaluation of online teaching status during epidemic period

During the epidemic, online teaching became the mainstream. Online teaching evaluation aims to systematically test teachers' teaching process according to certain teaching objectives and standards, and evaluate its value, advantages and disadvantages, so as to improve the quality of teaching. It is not only an important part of the teaching process, but also the basis of all effective and successful teaching. In this paper, we propose an online teaching evaluation method based on Epistemic Neural Network (ENN), which is an evolutionary intelligence method. In terms of uncertainty modeling, ENN's design innovation provides the improvement effect of geometric progression in terms of statistical quality and calculation cost. Therefore, it is very suitable for teaching evaluation, which is an evaluation process guided by a variety of uncertain factors. Specifically, this paper considers the content and grade standards of online teaching evaluation from five aspects. (1) Teachers' syllabus, teaching progress, teaching plan, courseware and other teaching documents and teaching materials; (2) Abide by teaching discipline, the implementation of teaching plan and the completion of teaching tasks; (3) Teaching attitude, teaching investment, teaching and educating people, and the comprehensive quality of teachers; (4) Whether the concepts taught in the course are accurate, the expression is clear, whether the key points are prominent and whether the difficulties are clearly explained; (5) The depth, breadth and frontier of teaching content, and the amount of classroom information. According to the above five evaluation indexes which involves the big data analysis, we train ENN to get an evaluation score that can evaluate the teacher's online teaching process. In addition, we also test the average evaluation time to verify the effectiveness.