Design And Implementation Of Big Data Management Platform For Kindergarten Children Based On Cloud Computing Technology

In recent years,with the rapid development of the domestic economy and the progress of scientific research and exploration in the field of preschool education,China’s emphasis on the education industry has gradually expanded the size of the preschool education market.However,the current management system for the field of preschool education is not only difficult to ensure the security of data,but also difficult to deal with the short-term surge and unstructured data with high value.However,from the perspective of system planning,the development integration process is inefficient because of the tight involvement between system modules or the difficulty in data interaction and connection between subsystems.Therefore,this paper studies and develops a kindergarten children’s data service system based on Hadoop technology support for the field of preschool education,and builds a complete resource management and information interaction platform for the field of preschool education.The main research work of this paper is summarized as follows:(1)Fully comb the characteristics of the current big data technology stack,and realize the deployment of cloud computing platform by rationally integrating computer hardware resources.This system adopts the structural design of combining Hadoop system and basic service platform,and effectively utilizes its stable,efficient and flexible data storage characteristics,and solves the problem of massive unstructured data management.(2)The platform combines Hadoop distributed system with spring cloud based microservice architecture.The system is divided into several modules for independent development and deployment,which realizes the effective decoupling of the major modules.This combination allows the platform to maintain increased flexibility and retain its flexibility and scalability in performance and functionality.(3)Based on the requirement analysis and architecture design,a complete,reliable and efficient data management platform is realized.According to the resource-sharing module,a preschool education resource base based on HDFS is built,and the function of the photo-management module is realized by using the photo distribution system based on face recognition.In addition,the paper analyzes and optimizes the content-based collaborative filtering algorithm,and then realizes the preschool education resource recommendation service through Map Reduce distributed computing framework.(4)In order to improve the stability and robustness of the distributed file system,the construction of the Hadoop cluster adopted the HA design of double Name Node.At the same time,according to the configuration of hardware resources in the system and the actual problems in operation,it has targeted from the perspective of Map Reduce computing performance and HDFS small file processing,effectively improving the utilization of storage space and the efficiency of file access.Finally,a child data resource management platform that is actually deployed and researched is researched and developed.The actual deployment and operation of the system verifies the stability and efficiency of the developed kindergarten data service system.