Academic Emotion Analysis Method And Service System Development For Student Feedback Text

Emotion is closely related to cognition,decision-making,communication and so on.It is of great significance to mine and analyze emotions.In the field of education,researchers pay more attention to academic emotions.A large number of studies have shown that academic emotions are closely related to academic achievements.Understanding students' academic emotions is conducive to better teaching intervention and management.At present,in the large-scale online open learning platform,feedback texts such as course comments published by students become an important carrier of academic emotion expression.It is of great theoretical and practical significance to analyze students' academic emotions from these texts.Emotional analysis based on artificial intelligence technology has been applied to the fields of public praise analysis and public opinion analysis.However,the application of emotion analysis has domain uniqueness,it is difficult to obtain better results by directly applying emotional analysis methods in other domain.Therefore,this paper constructs the academic emotion analysis corpus and lexicon,proposed a method of academic emotion analysis based on deep learning network model,and designs academic emotion analysis service system based on academic emotion analysis method.The research process of this paper includes four stages: In the first stage,this study firstly designed the distributed crawler program to obtain student feedback text data in the online learning platform,and standardized the dimensions of academic emotions.Then,based on the dimension of academic emotions,to design a text annotation system for student feedback.The results of data annotation include 8213 topic category data and 8712 academic emotion category data.In the second stage,based on the corpus of academic emotion,the general emotion lexicon is used as the seed word,and the emotional word expansion method is proposed,and the academic emotion words classification algorithm is designed to construct emotional lexiocn automatically,contained vocabulary or expression 860.In the third stage,this study incorporates the feature of academic emotion lexicon,and proposes an academic emotion analysis method based on the deep learning network model,including subject classification methods and academic emotion classification methods.Through experiments,the method proposed in this study is better than the general machine learning classification effect,in which the accuracy rate of the topic classification model is 88.62%,and the accuracy rate of the academic emotion classification model is 71.12%.In the fourth stage,based on the academic emotion analysis method,this paper designs and develops the student feedback text academic emotion analysis service system,which is convenient for students,teachers and teaching administrators to quickly analyze textual emotion.The contributions of this paper are as follows:(1)This paper constructs a set of academic emotion analysis resources in the domain of education,including academic emotion analysis corpus and academic emotion lexicon.(2)This paper proposes an academic emotion analysis method for student feedback texts.(3)This paper develops academic emotion analysis service system for student feedback texts.