DiSSCo Prepare Deliverable D6.2 - Implementation and construction plan of the DiSSCo core architecture

Abstract

Work Package Deliverable 6.2 aims to propose an architectural overview of the core components of the DiSSCo data infrastructure. This overview should provide clarity to stakeholders and guidance for the development team. The architecture is based on the earlier work done in the EC funded ICEDIG project in work package 6, discussions in the DiSSCo Prepare Work Package 5 and 6 meetings, DiSSCo Prepare end-user services development sessions and DiSSCo Prepare Technical Team meetings. The data infrastructure aims to support mass digitisation, to provide a FAIR digital representation of specimen data based on DiSSCo’s Digital Specimen concept. This approach provides rich provenance information through the full data lifecycle and enables enrichment through annotations and linking related data objects, with particular emphasis on machine actionability and support for AI. In total 23 different components were identified, grouped in 9 containers. Each component is an independent unit built for its specific purpose. The components have been grouped in containers, each aimed at providing a specific functionality. The communication between the different components is based on events using queues as a way to decouple the components. In this document the function of each component is described as well as its role within the broader container. We work through the architecture starting with the data providers, where the data is generated. The provided data are collected by the translators, which convert it into the generic openDS data format, based on community standards like Darwin Core. After translation, the enrichment services are a vital part of the architecture as they provide an automated way to enrich the data with additional information. Data is validated and stored with the data processing services in the data storage layer. Data exposure goes through a wide array of DiSSCo end-user services providing a range of functionalities to use and annotate the data. Last but not least are several services aimed at orchestration, providing new and updated data models or providing globally unique persistent identifiers. This wide collection of services working together in unison will provide a stable but extensible architecture for the DiSSCo Research Infrastructure.