Font Size: a A A

Research On Early Learning Warning Model And Assistance Strategy Based On Big Data

Posted on:2022-11-26Degree:MasterType:Thesis
Country:ChinaCandidate:J MengFull Text:PDF
GTID:2517306779475844Subject:Computer Software and Application of Computer
Abstract/Summary:PDF Full Text Request
Driven by modern information technology,the big data of education generated by hybrid teaching can better meet the personalized needs of learners.In the process of blended teaching,how to find students' learning crisis and realize effective warning and intervention services,so as to ensure the quality of students' learning,has become the development trend of the future education field.At present,the scale of data involved in the study of early warning and intervention is generally small,and the performance is more empirical than theoretical,resulting in less research results and explanations.In addition,the coarse granularity of early warning makes students unable to obtain enough feedback information through early warning to improve their learning process.In order to realize students' dynamic learning early warning,a data-driven early warning model was constructed,and the early warning mechanism was further improved based on classification of different objectives,so as to achieve effective intervention and assistance and improve learning effect and quality.The main contents of the study are as follows:(a)Build dynamic learning early warning model.This research introduces the design process of constructing students' dynamic learning early warning model in detail,including the construction of logical model,process model,algorithm selection and model optimization.Firstly,the logical model is constructed according to the target requirements,including the purpose,content,way and result of early warning.Secondly,the process model is built according to the specific analysis process of learning early warning.Finally,the complex correlation between performance and behavioral attributes was analyzed by integrating various learning indexes.Finally,the random forest algorithm was selected to optimize the learning early warning model,and the prediction accuracy and feasibility of the model were verified according to the evaluation indexes.(b)Study early warning mechanism research and early warning system design.The first is the cognitive level of the early warning mechanism design,around the course "College Computer Fundamentals" analysis of the correlation between knowledge points,determine the degree of influence between knowledge content,complete the difference processing of the course learning data,verify the scientific analysis of the construction of students' knowledge points and related application;Secondly,the characteristics of all kinds of learning activities are analyzed from the perspective of learning behavior,and learner portraits are constructed,and customized learning paths are generated based on portraits,cognitive and behavioral characteristics and attributes,which are recommended to students.Then,the learner emotion model is constructed from the perspective of emotion and emotion to analyze the relationship between students' emotion and learning under different conditions.Finally,according to the support services of different dimensions,the prototype of learning early warning system is designed for students and teachers.(c)Implement learning interventions and develop help strategies.Based on the advance safety system,this research puts forward the strategy of helping the students to realize the process interference of the students in the process of fine-grained,adaptive and dynamic learning.First,further refine the warning information particle degree,promote the students' learning and teacher's interpretation teaching from the perspective of cognition and behavior,and support the recommendation service based on cognition,behavior and emotion.Secondly,the individualized teaching intervention of the students in real time is designed to design different forms and recommendations around the object,environment and learning process,thus achieving the internal intervention mechanism design of the system learning process.Finally,the personalized learning recommendation service feedback is given to students to complete the strategy of fine-granulation of three kinds of early warning dimensions,and the implementation of the system based learning assistance service guarantee.
Keywords/Search Tags:Big data, Blended teaching mode, Early learning warning, Learning intervention, Assist
PDF Full Text Request
Related items