Research On Personalized Online Learning Effect Evaluation Model Based On User Portrait

With the vigorous development of Internet education,the abundance of online education resources,the freedom of space and time,and the low threshold for learners have attracted a large number of students to come to study.Internet data cache technology left a lot of online learning behavior records,along with the development of network technology,the education platform and the learners are no longer satisfied with the traditional teaching style,started to think according to the diversity and uniqueness of different learners,implement different teaching methods for each of the students according to their aptitude.Therefore,online education platform has become the main stage where personalized education shines.Personalized evaluation of students' learning effect can give timely feedback to online platforms and students,thus promoting the development of personalized education.In the field of personalized evaluation of online learning effect in China,first of all,researchers generally combine online learning data with offline learning data to evaluate students.However,the offline sample data is often not large enough and lacks universality.Moreover,when the number of online learners increases,it is difficult to obtain offline data.Secondly,there is a lack of a comprehensive index system for the personalized evaluation of online education in China.In response to the above problems,this article establishes an online personalized learning effect evaluation model based on user portraits.The following results have been achieved:(1)Using literature analysis and data mining techniques,an index system based on learning input,learning persistence and media preference is proposed,which has good universality.(2)Combining user portrait technology to portray students from multiple feature dimensions to obtain diverse student online learning evaluation results.(3)The accuracy of the student online learning effect evaluation model is as high as 87%,and it is reliable.This article collected 1.5 million student online learning behavior data sets of the online learning platform.After the process of cleaning,screening,statistical analysis and visualization of the data,the user portrait label architecture was established,through the data cleaning,filtering and statistical analysis and visualization of the process,user portrait tags system structure is established,based on the study into three dimensions,learning persistence and media preference to undertake to the student,the basic characteristics as student's identifier.Among them,learning involvement is divided into time involvement and behavior involvement,and the results are described as high involvement,medium involvement and low involvement.The results of learning persistence portrait are described as high persistence,medium persistence and low persistence.Media preference portrait results for text,video and picture three types.The results of the portrait show that only 8.79%of the students will invest more than 100 hours in online learning,about 85%of the students have low behavioral investment,87%of the students have low learning persistence,more than 80%of the students have low learning persistence and low learning engagement,44.7%of the students prefer video media,and 45%of the students prefer text media.Finally,the three dimensional feature clusters were used to evaluate and detect the learning effect of students.The accuracy rate of the model was as high as 87%,which has certain credibility.Finally,the evaluation of students' online learning effect is generated according to the result of the portrait.According to the analysis of the results of the user portrait,due to the lack of constraint and competitiveness of online education,most learners have low learning engagement and learning persistence in online learning,and it is difficult for students to finish a course continuously online and invest a lot of time and energy in online learning.Finally,the personalized evaluation report is generated according to the students' portrait results,and Suggestions for improving the online learning platform and students'teaching are put forward,so as to improve the students' online learning effect.