Research And Implementation Of Homework Mutual Evaluation System Based On Fuzzy Cognitive Diagnosis

In recent years,with the wide application and popularization of Internet technology and mobile devices,online education has been supported and widely developed in the field of education because of its advantages of being free from time and space constraints,speed and timeliness,repeatability,low cost,rich interaction and collaboration,and supporting personalized learning.Many online learning platforms have also been widely developed and used.On the online learning platform,in order to improve students’ mastery of knowledge points,students often need to complete corresponding homework.In deed,some objective questions in the homework can be automatically judged by the system according to the answers.But subjective questions are judged automatically according to the system,and the evaluation effect is not satisfactory.In addition,there are many learners in the platform,which brings huge homework evaluation tasks.Teachers can’t finish them successfully alone.In order to solve the above two problems,using mutual evaluation among students is an effective method.Using the method of mutual evaluation among students to complete the correction of homework.On the one hand,it can reduce the pressure of teachers’ evaluation effectively.On the other hand,students can be encouraged to participate in teachers’ evaluation tasks,deepen their understanding of the question,and strengthen learning communication and feedback.In the current study of homework review,there are two main problems.First,the ability assessment of students is lack of accuracy.Second,the strategy of homework mutual evaluation is relatively single,and other allocation strategies are not considered.Through the development of online homework mutual evaluation system,this paper improves these two problems in allocation strategy.The main work contents are as follows:Using fuzzy cognitive diagnosis technology evaluates students’ ability accurately.Before the distribution of mutual evaluation,it is necessary to diagnose the mastery of students’ knowledge points,and then predict the score of students’ homework to be evaluated accurately as the attribute description of students’ portrait.In the past,the assignment of homework mutual evaluation is mainly to simply predict the mastery of knowledge points by calculating the positive answer rate of the questions contained in each knowledge point.This method is lack of accuracy in the evaluation of students’ ability.This paper uses fuzzy cognitive diagnosis technology to predict the score of evaluation homework,and adds the attribute of students’ answer time to the fuzzy cognitive diagnosis model.According to the students’ answer data in the online learning platform and the answer time of each question,and considering the the high-order latent trait of each student,the discrimination,difficulty,and other attributes of knowledge points,different calculation methods are adopted for objective questions and subjective questions to describe the mastery level of knowledge points.Through using the Q matrix of homework,complete the score prediction of students’ homework.This paper takes the answer time into account in the fuzzy cognitive diagnosis model,which is more accurate than the previous methods in describing the differences of students’ abilities.A variety of allocation strategies are used in different scenarios.Traditional assignment strategies aim to achieve better overall mutual evaluation quality,but in the actual assignment process,there are other needs.We expect that after the overall evaluation quality of an allocation strategy meets certain requirements,the overall evaluation time can be shorter.In order to teach students in accordance with their aptitude,we also hope that students at different ability stages can improve their grades.In order to solve the problem of single evaluation allocation strategy and realize the requirements under the above scenarios,this paper proposes three different allocation strategies.They are according to the comprehensive quality,according to the shortest time and according to the learning ability.In order to calculate the matching results of the three allocation strategies,in the aspect of student portrait,in addition to the predicted score attribute of the homework to be evaluated,this paper also uses the data including students’ score ranking in the group,historical evaluation ability score,students’ evaluation time and homework similarity.According to different strategic requirements,select relevant attributes to build a weighted bipartite graph,and calculate the final result through km algorithm.This paper analyzes the distribution effect of various strategies through three indicators: the proximity to teachers’ scores,RSME value of mutual evaluation score and real score,the average value of overall feedback and the average time of each review,which effectively meets the needs of the above different scenarios and more accurate for students’ ability evaluation.The distribution effect of mutual evaluation is better than other common strategies..The homework mutual evaluation system adopts B / S architecture.The front-end web uses Thymeleaf + HTML and the back-end uses Spring Boot + Mybatis + Apache Shiro framework.The system mainly uses Java language and My SQL database.And the allocation strategy algorithm uses Python language.Teachers issue homework notices to students in corresponding classes and selected courses.Students receive the notice,complete the homework as required and upload it to the system.Teachers choose a certain assignment strategy according to the submission of homework and the characteristics of homework.Students in the same group receive their own homework evaluation tasks according to the results of the assignment strategy and complete the scoring of the question that will be evaluated according to the scoring rules.After students receive their own reviewed homework,they will restore a score to the reviewer according to the evaluation time and accuracy of reviewing.The homework mutual evaluation system developed in this paper alleviates the pressure of teachers’ evaluation,and effectively enhances students’ learning interest and interactive effect.