The RKD Research website, launched by the RKD-Netherlands Institute for Art History in 2023, makes art-historical data accessible and interoperable, supporting increased collaboration across the Dutch heritage sector.
Founded in 1932, the RKD-Netherlands Institute for Art History has for over 90 years played an important role as a knowledge center for art-historical research on Dutch art in an international context. One of the Institute’s core collections, the photo and reproduction collection, which comprises approximately 4.5 million images, has long enabled researchers to study art objects that are dispersed around the world.
Reproduction collections such as the RKD’s allow art historians to carry out stylistic comparisons and iconographic research, among other types of study. Until the end of the last century, art historians relied on physical photos and reproductions for their research. In recent decades, however, digital images have not only surpassed physical and printed materials in quality but have also become broadly accessible online. It was therefore a logical step for the RKD—where Astrid is acting managing director and Reiner is information manager—to continue acquiring reproductions digitally and, over time, to digitize the physical collection. We also wanted stop storing online-available image material locally, and instead to digitally link it to images that are made persistent and open available by other institutes.
To this end, in 2018, we launched the RKD Research project, under the motto “from analog to digital.” This transition fundamentally changes how art-historical research can be conducted. By moving from isolated physical collections to interconnected digital infrastructures, we have enabled broader access, faster research, and the ability to link data across institutions.
In the fall of 2018, we sent the first boxes of reproductions of artworks from the collection to a scanning company for digitization. But digitization alone is not sufficient to support the researcher. We also needed to create a search system to allow optimal access to the material and provide tools for comparison and annotation to make the found materials easy to study. Users had previously been able to access digitized materials in the collection through a platform called RKDexplore. But we needed to create a true digital research environment that integrated images, metadata, and tools for scholarly analysis. This combination of digitized image material and an interface with search and analysis capabilities was realized in the RKD Research website, which went live in December 2023.
Compared to RKDexplore, the RKD Research interface offers several new functionalities. For example, users can now search the RKD’s entire image collection using an image search tool developed by the Visual Geometry Group of Oxford University; researchers upload an image of an artwork, and the tool identifies identical reproductions in the RKD collection. Additionally, thanks to conversion through optical character recognition (OCR) and handwritten text recognition (HTR) technology, users will soon be able to search the texts that accompany the images. Researchers can also load found materials into a personal digital desktop where they can compare and manipulate images, for example, by zooming in on details, adjusting resolution, or aligning two works side by side for comparative study. Furthermore, the RKD is working toward full International Image Interoperability Framework (IIIF) compliance, which will enable researchers to continue their investigations in other tools, such as Mirador, an advanced IIIF-based image viewer. We plan to add more digital research functionalities to the platform in the coming years.
Developing an innovative digital research platform like RKD Research requires significant investment. In 2018, the RKD took out a loan to support the development of the platform, which meant that we not only faced the substantive challenges of creating the platform but also had to demonstrate financial feasibility and our ability to recoup our investment.
We intend to keep the data and search functionalities of RKD Research freely available to all. To cover the costs of continued development, maintenance, and loan repayment, in 2024, we introduced a subscription model for specific features, such as visual search and downloading high-resolution images. We are currently evaluating this subscription model for its effectiveness, accessibility, and sustainability, which includes exploring questions of user acceptance of the model and the commitment required by employees to further develop and manage the model.
The shift from analog to digital also requires new financial insights and approaches—a learning process the RKD is still undergoing. So far, we’ve learned that clear agreements, solid project management, and transparent communication with external partners are essential. We are applying insights gained during this transition, for example, in improving cross-departmental collaboration and clarifying ownership of key processes, to comparable projects within the organization. One of the challenges we’re facing is aligning different interpretations of strategic goals across teams. We’re experimenting with structured feedback loops and short-cycle evaluations to identify gaps and adjust our approach. These methods are helping us learn how to better translate strategic ambitions into operational changes.
The Dutch heritage sector strives to make digital heritage information openly accessible online. As expressed in the National Digital Heritage Strategy, the Ministry of Education, Culture and Science supports this vision and encourages the sector to realize it collaboratively. To that end, the Ministry has established The Digital Heritage Reference Architecture (DERA) Council and the Digital Heritage Network (Netwerk Digitaal Erfgoed). The DERA Council, under the guidance of the Ministry and in collaboration with professionals from the field, develops the principles and service landscape needed to make heritage data openly accessible and interconnectable. The Digital Heritage Network coordinates the implementation of digital heritage services.
Ultimately, the Ministry’s goal is to digitally connect all Dutch museums, archives, and libraries in a large knowledge graph on which various services, such as thematic portals, can be built from different perspectives. This unique approach aligns closely with the RKD’s vision and mission. The RKD has aligned the architecture of its information infrastructure with DERA by, for example, making data persistent and openly available as linked open data (LOD) and using standardized vocabularies and named entity files, and expects this to enhance its service to its target audience. While the RKD still stores a lot of art-historical documents itself—albeit now digitally—in future, we expect that we could instead connect to material held by other institutions. This will become possible as heritage institutions increasingly make their data available according to DERA guidelines: using persistent identifiers, adhering to open standards, and being available without restrictions.
This interconnectedness allows for a division of responsibility. Museums create descriptions of artworks, which the RKD can either retrieve or link to. We provide RKDartists, a LOD dataset of biographical data on artists and other relevant figures in art history, which relieves museums of the need to maintain biographical data themselves. The RKD, in turn, can link back to artworks described by museums that are published as LOD.
As the amount of art-historical data being published according to LOD standards grows, we’ll face the challenge of continuing to develop the platform so we can make that data available to researchers.
Ultimately, the RKD aims to make art-historical information more accessible, interoperable, and future-proof. By investing in digital infrastructure, we’re not only preserving knowledge but also enabling new ways of working and collaborating across the heritage sector. In a time of financial pressure, it’s crucial to make the case for the long-term value of this type of infrastructure—not just for researchers, but for everyone who benefits from well-supported, publicly available cultural data.
10.1146/katina-090525-1
