Sovereignty, Trust and the Open Layer: Takeaways from the 3rd Nordic Bibliometrics Infrastructure Spring Meeting

An invitation from Laura Himanen, a long-standing collaborator through CoARA's OI4RRA Working Group (Open Infrastructures for Responsible Research Assessment) and the GraspOS project, brought OpenAIRE to the 3rd Nordic Bibliometric Infrastructure Spring Meeting. What came as a genuine surprise was the discovery of a thriving and eagerly growing community working on research intelligence with a level of seriousness rare to find. The meeting crystallised something OpenAIRE has long argued: research intelligence is no longer peripheral. It is becoming structural, and the infrastructure beneath it cannot remain a black box.

The Nordic shift: research analytics as a pillar in the making: Across the Nordic region, there seems to be a growing recognition that research analytics should not sit downstream of decisions as a reporting layer. Although in its infancy, this is firmly in the thinking. Erja Heikkinen, Director at the Finnish Ministry of Education and Culture, set the tone in her opening remarks. Research intelligence, in her framing, is moving towards becoming a critical component of policy development, investment strategy and institutional governance, rather than a siloed administrative task.

This direction of travel matters because it raises the bar for everyone working on the supporting infrastructures. If research intelligence is going to shape policy and funding, then the systems producing that intelligence must be transparent, auditable and trustworthy. Anything less means letting opaque processes shape public decisions about science. 

The paradox we cannot keep ignoring: There is a contradiction at the heart of how research is currently evaluated, and the Nordic discussions made it impossible to look away from.

Public money funds research as a public good. Governments, taxpayers and citizens collectively invest billions in scientific work that is meant to serve society. Yet the intelligence layer used to evaluate that research, the data, the graphs, the analytics, remains largely dependent on closed, commercial systems.

This dependency creates real problems. The community cannot fully audit the data or the methodologies behind it. Institutions rely on external actors to decide what is visible and what counts as impact. Strategic control over the representation of public research sits outside the research community itself.

Is OpenAIRE comfortable with that arrangement? No. And increasingly, neither are the institutions and ministries shaping the future of European research policy.

Open infrastructures are not optional: This is why OpenAIRE keeps insisting, in every forum, that open infrastructures are a prerequisite for responsible research assessment, not an optional add-on.

Without them, evaluation systems are built on top of structures no one can inspect. Open infrastructures allow methodologies to be transparent and open to scrutiny. They let the community retain control over how research is represented. They align research intelligence with public values. And they build trust through provenance and traceability.

This is the core argument of OI4RRA, the Open Infrastructure for Responsible Research Assessment initiative within CoARA, and the Barcelona Declaration. It is not just about producing better data. It is about governance, accountability and long-term sustainability of the systems that shape how science is judged.

A careful transition, not a disruptive one: What stood out in the Nordic conversations was the realism. Nobody in that room was suggesting we abandon existing systems overnight.

Metrics are deeply embedded in funding decisions, evaluation frameworks and policy structures. Destabilising them risks undermining the very trust we are trying to build. The question is not whether we move to open. It is how to manage that transition responsibly, protecting institutional stability while shifting the foundations underneath.

This is where mapping the open landscape becomes essential. The work led by Cameron Neylon and others in the ORION DBS community on registries of open sources is doing exactly that, giving institutions the practical guidance they need to navigate the transition with clarity. Cameron's vision of an open research information collective points to a different model altogether. One where infrastructure is built and maintained collectively, where we all put effort to maintain the openess and transparency, where we open up for scrutiny to the experts so as to improve. To be better than the existing closed systems.This is not about sharing the burden. It is about building resilience, so that no single point of failure can collapse the systems that represent our collective knowledge.

Open graphs have grown up: Perhaps the most important shift observed at the meeting was the gap between how open scholarly graphs are still perceived and what they have actually become.

There is a lingering assumption that open means incomplete, lower quality, or experimental. That assumption is increasingly out of date. Yrjö Leino from CSC and others presented detailed work showing how open data can support rigorous national-level analysis in Finland with OpenAlex, though typically after considerable effort spent on cleaning and filtering. And ABACUS, the Greek bibliometric infrastructure, demonstrates something the community needs to internalise: the OpenAIRE Graph is no longer an experimental project. It is a production-ready backbone for national research intelligence.

Cameron Neylon also presented evidence from a Swedish study, where he combined the OpenAlex and OpenAIRE Graph datasets to achieve full coverage of the national research output, and suggested this complementary approach to the audience as a practical way forward. What surprised several participants was how strong the OpenAIRE Graph's coverage of Swedish research turned out to be. That result is not accidental. It reflects the work done on the ground with SciLifeLab, where the OpenAIRE Graph powers the SciLifeLab Open Science Monitor through deep integration with national repositories, curation by SciLifeLab teams, and full-text mining to link publications, datasets and software. The lesson is straightforward: open infrastructure improves dramatically when it is built in partnership with the communities it serves.

Are we there yet? Yes, we are on the last mile. What needs to change now is perception.

The question underneath all of this 

If everything heard in those days had to be distilled into a single question, it would be this: who controls research intelligence?

Who decides what counts, what is visible and how impact is measured? If the answer is not the research community itself, governed transparently and accountably, then we have a structural problem that no amount of better metrics can fix.

This is what digital sovereignty means in the context of science. It is not a slogan. It is the practical capacity of the research community to steer the infrastructures that shape how knowledge is represented and evaluated. Without that capacity, public research autonomy is incomplete.

A path that takes longer but holds together: The Nordic region is showing that this transition is possible. Not through disruption, but through deliberate, coordinated action grounded in institutional repositories, open access publishers and community-governed infrastructure. The path forward is not about swapping one closed system for another. It is about building a trusted, open and community-governed layer for research intelligence that the entire ecosystem can rely on.

It may take longer. It is certainly more complex. But it is the only path that keeps the values of science aligned with the methods used to evaluate it.

Warm thanks to Laura Himanen and the CSC for the invitation, and to the entire Nordic community for the seriousness, depth and honesty of the conversations. The work ahead is substantial, but the direction is clear.