Reference Models are abstract frameworks for understanding significant relationships among the entities of some environment [SOA-RM]. They support communication, comparison and the creation of standards.
“Socio-economic Sciences and the Humanities contribute to an in-depth, shared understanding of the complex and interrelated socio-economic challenges facing Europe and the rest of the world. Social sciences and humanities deal with knowing (data, statistics, indicators), explaining (tools, definitions, methods), understanding (context, causes, mechanisms), forward looking (modelling, foresight, scenarios) and recommending (from knowledge to policy)” [FP7-SSH]. Social sciences, and to some extent humanities, have used digitally born data as basis for empirical research for a long time and they have also used digital repositories for a long time, the first ones started in the sixties as national resources. They have since then formed domain specific international clusters, later identified by the European Strategy Forum on Research Infrastructures [ESFRI] as important services in the European research domain. Together with two multinational (European) researcher-driven data creators they are now organised in five domain specific European Research Infrastructure Consortia (ERICs). The services of these infrastructures provide are very similar, but the way they are organised and are functioning is very disparate. The different disciplinary domains need to support their own specific research questions and methodology, however, up till this date it remains unclear up to what level their services need generic infrastructures and to what level they need discipline-specific services. This task will not impose a single reference model that all SSH infrastructures are required to follow, but rather create an architectural framework that facilitates both inter-infrastructure access to data as well as maintaining the unique needs, requirements and essence of each community‘s infrastructure – an architecture that offers a roadmap towards harmonization and will assist in widening and strengthening the SSH domain in Europe.
Research infrastructures are referred to as the facilities, resources and related services used by the scientific community to conduct top-level research in their respective fields [EU-RI]. However, they are often confused with the projects or organizations that aim to improve research by providing the infrastructure. Other activities of these projects and organizations include support, training and outreach. To avoid confusion, this document will use the term data infrastructure for all services to support the research lifecycle of data, from creation to reuse.
[ISO 42010] describes architecture as “the fundamental concepts or properties of of a system in its elements, relationships and the principles governing its design and evolution”. It describes how architecture can be described using one or more views, which are each governed by viewpoints that frame specific concerns of stakeholders.
Architecture frameworks provide viewpoints to describe specific types of systems. Enterprise Architecture Frameworks do so for enterprise systems, and typically provide a methodology for defining and implementing the these [TOGAF-ADM][DYA].
Reference architectures model the abstract elements of a domain independent of its technologies, protocols and the products that are used to to implement a solution for the domain [SOA-RAF]. Reference architectures should be derived from existing architectures by taking their best practices from explicitly documented knowledge or implicit knowledge. In turn, they can be used to instantiate system architectures [CLOUTIER]. Governmental reference architectures [NORA][UK-RA] are typical as they put an emphasis on architectural principles for aligning the different types of systems of a government.
Reference Models are abstract frameworks for understanding significant relationships among the entities of some environment [SOA-RM]. They support communication, comparison and the creation of standards.
The documented current state of architecture is assessed on a strategic level and on implementation level.
On a strategic level the ESFRIs have similar or complementary objectives and most of them identify high-level principles. All aim to support high quality research: ESS and SHARE focus on improving the creation of european-wide data, CESSDA focusses on data management and provision of data and DARIAH and CLARIN focus on the complete research lifecycle.
All the ESFRIs aim to provide infrastructure and standards to overcome distribution-issues. CESSDA, DARIAH and CLARIN have defined values or principles to guide the collaborative infrastructures.
The strategic documents identify similar services for achieving the objectives: seamless (persistent) identification, description and discovery of data and seamless authentication and access to data. The analysis of the documented architecture [DASISH-SOA] indicates a technical focus from which no common view could emerge.
Existing work on architecture for research infrastructures exists on individual components and on infrastructure level. One is the OAIS Reference Model [OAIS][ISO 14721] provides a recommended practice for long-term preservation archives. The recently developed ENVRI reference model aims to model an “archetypical” environmental research infrastructure based on [RM-ODP], which allows the community to understand the top-level technical architecture [ENVRI]. Other work includes a reference architecture for long-term preservation [SHAMAM], which focuses on designing and developing solutions for digital preservation in distributed scenarios, and [RIEDEL] which provides a reference model for data infrastructures from a very technical perspective.
[SOA-RM]
[FP7-SSH]
[ESFRI]
[EU-RI]
[TOGAF-ADM]
[DYA]
[SOA-RAF]
[CLOUTIER]
[NORA]
[UK-RA]
[DASISH-SOA]
[OAIS]
[ISO 14721]
[RM-ODP]
[ENVRI]
[SHAMAM]
[RIEDEL]