Research On The Application Of Mining Emotion In Evaluation Of Teaching Quality By Students Based On Aspect-Level Sentiment Analysis Technology

With the rapid development of the education field,its direction has also shifted from the initial speed-based scale expansion to the quality-based connotative development.The demand generated by this change has also made the education data mining field become a hot spot of social concern.In addition,student evaluation of teaching is a very important part of the current teaching quality monitoring system.Students can express their feelings and emotions in the teaching process through the teaching evaluation process,therefore,the data generated in the teaching evaluation process contains very rich emotional information.And by mining and displaying this information,it is possible to better inspect the completion of teaching goals and evaluate the teaching ability of teachers,so as to achieve the purpose of improving teaching quality.The purpose of this thesis is to fully utilize the emotional information in the teaching evaluation data,focus on the mining and analysis of student emotions in teaching evaluation data,research on relevant methods in the field of education data mining and sentiment analysis,and according to the characteristics of teaching evaluation data,design a classification model for teaching evaluation,which combines students’emotional and emotional characteristics and rating preferences.The main work contents are as follows:(1)According to the characteristics of evaluation text data,designed rule-based evaluation collocation extraction rules,and through the evaluation object’s evaluation factor matching and evaluation words’ emotional polarity calculation,extract the aspect-level emotions of students in the teaching evaluation data for various teaching evaluation factor.(2)In response to the subjectivity and difference of students’ rating,combine the emotional information and characteristics extracted from the teaching evaluation text with the students’rating preferences,construct a classification model based on the XGBoost algorithm for teaching evaluation rating,classification teaching evaluation categories.(3)Based on the results of the model,use the association rule mining method to analyze the impact of student emotions on the results of teaching evaluation.and visualize the students’emotions and feedback in the teaching evaluation data in the form of teacher teaching effect portraits.Finally,verifies the method proposed above by labeling and experimenting with real teaching evaluation data,the experimental results show that the teaching evaluation factor,evaluation word extraction algorithm and teaching evaluation scoring classification model are effective.Through the above methods,students’ emotions in teaching evaluation data can be efficiently mined and analyzed,and provide more intuitive and reliable data support for teachers and academic staff to improve teaching quality.