Educational Information Publish-Ing And Management Platform Based On Microservices Archite-Cture

In the context of Internet plus education,more and more online education service products have flooded the market,but traditional education service products have great shortcomings in terms of business and technology.In terms of business,the traditional education service platform has a single function and does not form an independent education service ecosystem.Therefore,there is no one-stop education service platform in the vertical education field.In terms of technology,due to the single function,companies generally adopt a relatively simple management monolithic architecture for implementation.In the later stage,with the expansion of the business,the iteration of functions,and the growth of team personnel,the difficulty of project development and management will also become index Level growth.In view of the above issues,this article provides one-stop education services in business to create an independent and perfect education ecosystem.Technically,the architecture design is based on the design concept of "big,middle,and small front desks",and the front end uses React Native A hybrid mobile application development framework with a back-end design based on the Spring Cloud microservices architecture,so that the system has features such as grayscale release,high cohesion,modularity,high availability,scalability,and flat management.Based on the concept of CAAS(Container as a Service),this article combines container technology to achieve automatic system operation and maintenance,and cloud computing technology to achieve system elastic cloud expansion.Based on this,an enterprise-level online education service cloud platform is implemented.First of all,this paper makes a comprehensive analysis of the system requirements,and combines the characteristics of microservices to complete the granular separation of system service modules.Researched the key distributed technologies involved in the system and optimized them in accordance with the actual needs of the system,including constructing a set of distributed global unique serial number generators based on the business characteristics of the system,optimized distributed locks,and The single sign-on implementation scheme,a set of MySQL and ElasticSearch data synchronization schemes,and improved load balancing strategies based on a consistent HASH algorithm.Then,for the next optimization work of the system,a collaborative filtering recommendation algorithm based on knowledge graph is proposed.The simulation verification results show that the fusion knowledge map can to some extent make up for the lack of semantic information of the traditional collaborative filtering algorithm.The algorithm has better performance in three aspects of accuracy,recall and F value.Alleviate data looseness and cold start problems.Finally,the implementation and containerization of the functions of each service module of the system were completed,and the functions were verified.From the results,the function of the system has achieved the expected results.