Research On MOOC Evaluation And Improvement Based On Learners' Academic Emotions

The spread of the Internet has allowed massive open online courses(moocs)to become a new way of learning in education.With the rapid expansion of MOOC,the evaluation research on MOOC is in full swing.On the MOOC platform,during or after each course,learners will write course comments in the course evaluation area,which is an important feedback information for learners to MOOC managers and teachers,as well as an important support for comprehensive evaluation of MOOC course quality and teaching effect.However,in the actual study of MOOC evaluation,due to the huge amount of data of comment information and the complex content noise,it has not been effectively utilized.The emergence of text mining technology accelerates the processing speed of big data,especially text data.The application of this technology to the MOOC course comment information processing can deeply understand the feelings and needs of MOOC learners,and can serve as an important supplement and support for MOOC evaluation research at the present stage.After summarizing the evaluation of MOOC in foreign countries,this study established a MOOC evaluation method with learners as the evaluation subject and text mining as the technical support.Firstly,the review texts of MOOC platforms in Chinese universities were collected as data sources,and the concept of academic emotion was used to establish the classification criteria.Learners were divided into four categories: positive high arousal,positive low arousal,negative high arousal and negative low arousal.Secondly,deep learning algorithm was used to establish the automatic classification model,and the model accuracy reached about 80%.Using this model,it is found that the positive state of learners Posting comments is far more than the negative state,the high arousal level is far more than the low arousal level,and there is a significant difference between the students' emotional state of science and technology courses and humanities courses.On this basis,the text sentence level is classified to obtain the subjects of learners in the five dimensions of teaching team,teaching content,teaching resources,teaching effect and teaching technology.It is found that the most mentioned subjects are teachers in the positive state,while the most concerned subjects are teaching content in the negative state.Therefore,MOOC course units are evaluated and improvement strategies are put forward to provide references for MOOC construction and reform.