Analysis Of The Results Of Online Evaluation Of University Students' Education Based On Data Mining

In university education,the standard of schoolroom teaching is extremely vital,and also the variety of abilities trained by the college,high or low,depends directly on the standard of teaching within the schoolroom,therefore,a student evaluation system is developed,and this system can in turn urge to improve the teaching of teachers.The combination of network and information technology tools applied to the improvement of student evaluation system is the trend of modern society.Nowadays,domestic colleges and universities generally have their own teacher evaluation systems,and college students make evaluations for lecturers through the online platform,and the results of those evaluations will replicate the link between academics and students and therefore the teaching standing of academics.The evaluation content and categories mainly include five aspects: teaching perspective,teacher quality and skill,teaching content,the way of teaching,and teaching effectiveness.However,due to the influence of various extraneous factors,the effect of online evaluation generally fails to reach the expected goal,resulting in the results of online evaluation not being objective and real,thus not reflecting how a teacher his real teaching level.Since there are still some problems with online teaching evaluation,such as haphazard evaluation,malicious poor evaluation,extreme positive evaluation,or just malicious poor evaluation of a teacher and positive evaluation of other teachers,etc.Therefore,if the evaluation data of students are analyzed and classified directly,it may lead to incorrect evaluation results and not play the real role of online teaching evaluation.Therefore,we should first analyze these data submitted by students and filter the data submitted by students.The data sample in this paper is the teaching evaluation result data of students in one of the colleges of J school.This paper focuses on how to make reasonable use of the online teaching evaluation data submitted by students to distinguish teachers' teaching level,so as to provide meaningful reference for college teaching management and teachers' reform and innovation,and to be able to use these evaluations in turn to guide teachers to improve their teaching methods.It is mainly divided into three parts.First,we collect students' online teaching evaluation data and clear the abnormal data values.Since some students' evaluations are not objective and unjust,and have strong personal opinions about a certain course narrated by a certain teacher,their evaluations will have large deviations from those of other students.In this case,we need to remove these abnormal data,and this paper uses the cosine similarity algorithm to remove the abnormal data.Also,because some students have high overall ratings and some students have low overall ratings.In order to remove the influence of these irrelevant factors,a normalization of all sample data should be performed.In this paper,typical normalization and standardization methods: z-score method and min-max method were chosen to standardize all evaluation data and make them comparable.Second,the experimental data are clustered.The algorithm used in this paper is K-means algorithm,which is a traditional clustering algorithm,and can cluster and analyze the evaluation data,we used it to cluster and analyze the evaluation data to obtain a sample of three categories of teachers at the college,which get the sample data for training the classification model.Third,design experiments to classify all teachers in the sample data.In this paper,an artificial neural network BP is used for classification.It also combines a genetic algorithm to optimize the network weights and use the clustering results to train the network,which finally constitutes a network model.From the final experimental results,it can be seen that this network model can classify teachers well,and the classifier model can basically predict the teaching level of teachers accurately.This paper takes the real data of a college as the sample for research,the experimental results show that the scholars faculty of school of faculty L square measure primarily glad with the teaching ability of the lecturers during this college,the overall teaching operation of the college is relatively optimistic,the overall level of teachers is relatively high.