Analysis Of Students' Behavior On The Online Open Course Platform Based On Big Data

With the continuous development of computer network technology and the popularity of Internet applications,information technology is constantly changing our traditional education and teaching model.The large-scale open online course in 2012-MOOC rises in the United States,quickly set off a worldwide revolution in education technology,which has been generally welcomed by the education community,college teachers and students,and the social audience.The focus of this article is on the learning behavior data of learners in the open online course learning platform in our school.Such as watching the time of learning video,learning effects,participation in learning activities(interactions,discussions,etc.),the completion of the job,the number of practical activities.There are individual differences in the learning behaviors of different learners during the learning process.This paper starts from the study of learners' learning behaviors in MOOC,and mainly carries out the following work:This paper focuses on the use of a variety of research tools and methods of big data to cleanse,process,standardize,and analyze the log data of all course participants generated under the MOOC platform,so as to analyze the learning behavior characteristics of the course learners.The research content of the dissertation mainly includes: cleaning and preprocessing the behavioral log data,then descriptive information from the basic statistics of the course,the learner's personal basic information features,the learner's watching video-related behaviors,the learner's practice related activities,and the learner's discussion area interaction The characteristics of the relevant behaviors were analyzed.Then the correlation analysis and hypothesis testing method were used to extract the factors that have a greater impact on the learner's performance and related behavior characteristics.At the same time,the learner's performance prediction model was established in combination with the machine learning classification prediction algorithm.The extracted sample data is used to train the model and predict the final course certificate obtained by the learner.The accuracy of different algorithms is compared and the research conclusion is draw.Support vector machine(SVM) algorithm has some deviations in the student behavior analysis of our open online course learning platform.Therefore,this paper proposes a Gini model based on the Random Forest(RF)model.The indicator feature-weighted support vector machine(RFG-SVM)method is applied to the student behavior analysis of the online open curriculum platform.The RFG-SVM method uses the Gini index under the random forest model to calculate the degree of influence of each feature variable on the accuracy of RF model classification and identification,so as to compare the importance of each attribute of the data set and set the importance of each attribute feature according to importance.With corresponding weight values,feature attributes with greater influence are obtained with a greater weight ratio than feature attributes with less impact.Then the weight value is applied to the calculation of the kernel function of the support vector machine,so as to obtain a higher correct rate of the student behavior analysis of the online open platform of the support vector machine algorithm,so as to find some universal learning behavior features and habits.And laws to predict students' final learning outcomes.