A Reference Software Architecture for Managing, Processing and Serving Multi-field Scientific Data

Scientific data resources typically have the characteristics of rich data types, diverse fields, and close semantic links. Traditional e-scientific technologies cannot simultaneously support the fusion management, processing and service of multi-field scientific data resources simultaneously, leading to the repeated development of scientific data software. This article proposes a reference software architecture for the managing, processing and serving of multi-field scientific data. This architecture consists of function modules like hybrid cloud management, integrated elastic deployment, heterogeneous data fusion store, data pipeline, and data middle platform. It can provide cross-domain heterogeneous infrastructure integration, heterogeneous data fusion management, data pipeline processing, and public master data services. This can improve the technical level of multi-field scientific data governance, service and utilization.