Design And Implementation Of Data-driven Scalable E-learning Platform

In recent years,the popularity of online education has increased,more and more learners choose to learn what they need through e-learning platforms anytime and anywhere,at the same time,more and more teachers also tend to conduct teaching activity with online and offline methods combined.Although there are various e-learning platforms on the market,there are still many problems:the orientation of these platforms is not clear,the offline learning resources are copied stiffly to the online,and there is very little information that is really useful for learners;the platform’s functions and businesses are redundant,bringing out poor scalability and bad online learning experience;focusing only on business but ignore data,and failing to collect and fully utilize the large amount of valuable learning behavior data accumulated by users on the e-learning platform;too many e-learning resources,learners often need to spend a lot of time and energy to find the e-learning resources they need,so that their learning efficiency and leaning effect are not good.In view of the problems above,the main work and contributions of this paper are as follows:·A design and implementation method of scalable e-learning platform is proposed.Firstly,it is clear that the research object of this paper is the full stack scalable e-learning platform for computer subject,then the core functional and non functional requirements for this kind of e-learning platform are analyzed,and finally the implementation methods of these core requirements and solution of the key problems are explained.For e-learning platform,its scalability can be understood in two aspects:first,the scalability of business system.The scalability of this aspect mainly depends on the implementation of micro service architecture in single system and single sign on in multi subsystem,besides,the design concept that the business of each subsystem constitutes an organic whole of a course in the form of e-learning resources also counts;second,the scalability of physical resources.The scalability of this aspect mainly depends on the complete container deployment of applications and the elastic scaling mechanism of kubernetes cluster.·An optimized online learning behavior data collection and storage method is proposed.Different from the traditional e-learning platform that only focuses on business and ignores data,this paper proposes an optimized learning behavior data collection and storage method based on xAPI technology.Compared with the traditional xAPI data collection method,this paper innovatively adds a data service layer between the data collection layer and the data storage layer,and explains the design and implementation details of this optimized data collection and storage method.The data service layer improves the stability and high availability of data collection,experiments show that this optimized learning behavior data collection and storage method can still maintain high availability and low interface delay in high concurrency scenarios,and its comprehensive performance is better than the traditional xAPI data collection and storage methods.·A data-driven and efficient e-learning resources recommendation method is proposed.Based on the collected and stored learning behavior data,especially the interaction data between students and e-learning resources,this paper constructs a heterogeneous information network about students,e-learning resources and teachers,and creatively proposes an efficient neighborhood interaction enhanced e-learning resources recommendation method which called NIRec-for-ELR to improve learning efficiency and experience of students.Experiments show that this method can effectively represent the neighborhood interaction information in heterogeneous information networks,and its performance is better than many existing methodsIn addition,taking Shuishan Online,an e-learning platform developed by the school of data science and engineering of East China Normal University,as the practical application scenario,the feasibility of the above design and implementation method of scalable e-learning platform is verified,and the optimized online learning behavior data collection and storage method has innovation and generality to some extent,the e-learning resources recommendation method called NIRec-for-ELR has excellent performance and can effectively help learners improve learning efficiency and experience.