The Research Of MOOC Learners’ Learning Behavior And Push Application

With MOOC swept the world,the development of online education ushered in another golden period,tens of millions of learners flock to the Coursera,edX,Udacity as the representative of the MOOC learning platform.Learning,analysis,data mining,depth learning,machine learning,artificial intelligence and so on,and how to use these techniques to conduct MOOC learner behavior analysis,for learners to provide a better learning experience for vocational education,lifelong learning building blocks,Has become one of the most concerned issues in the field of education.This paper introduces the current development of the well-known MOOC platform at home and abroad,analyzes the current research situation and existing problems of learning behavior analysis at home and abroad;the edX public data set published by Harvard University and MIT and the dream course of National Defense Science and Technology University Platform data sets for the experimental data set,from the platform,curriculum,learner type three aspects of the comparative analysis of the MOOC learners learning behavior,to explore the Chinese learners and MOOC birthplace learners behavior differences;This paper introduces MOOC integral formula and improved k-means clustering method,and proposes a classification method of learner type based on MOOC integral clustering by qualitative and quantitative analysis,and tests the learner type partitioning method,and divides the learner type and the learner type Electronic badge combination of MOOC based on the system of continuous learning motivation incentive mechanism;to dream platform platform from October 2013 to October 2016 accumulated more than 100 million learning behavior data,based on correlation analysis from the curriculum 11 behavioral data which are most relevant to the MOOC learners ’academic performance are found out in the 50 behavioral data of three dimensions of learners’ own factors and other personnel factors,and the behavioral data of multi-linear regression and neural network algorithm are compared respectively.And finally the average accuracy rate of MOOC dropout forecast is more than 90%.Based on the learning behavior analysis and research conclusion,MOOC data-based push platform framework of dream lesson is designed,and the "curriculum push module" "Push module" for the experiment object,through the flow chart design,JAVA language programming and push the curriculum push and friends to achieve the function,the ultimate success of the officers and soldiers for the service platform for the dream.MOOC data based learner behavior analysis can use learners in the MOOC platform to leave massive learning behavior data,which translate into effectivelearning behavior information to help learners get a better learning experience for teaching teachers to provide teaching assistance Decision-making,platform development platform for the development of the views and suggestions will be to promote MOOC continue to move forward the inexhaustible motive force.