Research On Service Oriented Architecture Of Scientific Data-Sharing Platform

Scientific data is the data and information produced during the process of science and technology practice and scientific exploration of human society. Over the years, our scientists spend great effort doing research, exploration and observation in various scientific fields and accumulate rich resources of scientific data. Scientific data-sharing platform is a software platform aimed at providing unified management, distribution and sharing of scientific data.First this paper discusses the problems that traditional scientific data-sharing platform has to face about system architecture and design, and analyze the existing deficiencies it has in the aspect of business agility, scalability, platform interoperability and automation. Then, based on the introduction of basic concepts and features of Services Oriented Architecture (SOA), it conducts a comprehensive study and practice of scientific data-sharing platform using serviced oriented approach.Service-Oriented analysis and modeling of scientific data-sharing platform is the key to the successful implementation of SOA. In the process of service-oriented analysis, it uses the integration of three analysis methods, i.e., business process analysis, entity-relationship analysis and use case-driven analysis to conduct a comprehensive analysis of scientific data sharing platform; meanwhile, as for the existing legacy systems in the traditional scientific data-sharing platform, it discusses in detail the application of services oriented integration of legacy systems in the scientific data-sharing platform.In the service-oriented modeling part, first, this paper discusses the service-oriented hierarchical model, and based on this model, divides the services in the scientific data-sharing platform into three layers: SOA application layer, SOA business layer and SOA orchestration layer; Then...