6 February 2025
Annette Dortmund, Product Manager, OCLC (ORCID ID)
Looking back there’s one thing that hasn’t changed since the days of the Library of Alexandria—the core mission of libraries. The mission to collect, preserve, and share knowledge, whilst also helping to create new knowledge. That’s something that remains constant giving us the confidence to embrace the changes ahead.
Metadata work has always been part of this. Not just good old cataloging of books, but all types of data and metadata work done by skilled information professionals. But what will the future of metadata hold? What is to come? And how will we prepare, and evolve with these changes? Which metadata future do we want to create?
Pairing stability with innovation
Conversations about replacing MARC with linked data have been going on for over a decade. And while we are seeing linked data being integrated with mainstream library service platforms, we’re still also very much tied to MARC.
As we see more metadata being created in linked data, it will be in addition to—and not as a replacement for—existing metadata formats. It is important that we embrace a variety of approaches because we need models that liberate library data from their silos. We need easier compatibility with data in museums and archives, with scholarly communications. And we really need to use persistent identifiers (PIDs) and create data models that are understood by machines, including artificial intelligence (AI).
The key takeaway is that flexibility will be crucial. We expect more models to evolve and with no one size fits all approach, there should be plenty of interoperability and hopefully fewer barriers.
AI: hype versus reality
AI will, of course, be a big help in data driven decision-making, such as resource sharing and collection management. It’s already helping with metadata quality management. At OCLC, we’ve been using machine learning for years to improve WorldCat, identifying errors, enriching records, and de-duplicating metadata. This will only accelerate over the coming years.
But will AI be creating our metadata for libraries? Not entirely! While AI will greatly support formal cataloging, the more complex work around rare books, special collections, and nuanced subject indexing will still need human expertise. AI will also help with descriptions of images in digital libraries and tackle backlogs of uncataloged materials—something we are already seeing. But regardless, there will still be a place and need for human expertise in the loop.
Context, not just description
By 2030, metadata work will shift to even more context-driven work. The growing information overload, multiplied by AI, will require selective, context-aware metadata creation. It won’t just be about describing objects or concepts; it’ll also be about understanding their place in the larger knowledge ecosystem.
Think about the work some museums do—minimal formal description when something new enters the collection, instead focusing on how it connects to other objects, themes, or people. Over time, more formal descriptions are added based on the object’s relevance and use.
Similarly, metadata work will focus more strongly on the context and relationships around information, including at scale. Bringing those to the larger knowledge graph, new relationships and contextual cues are unlocked. Instead of browsing shelves, we’ll browse the graph, finding connections and discovering diverse perspectives that weren’t immediately obvious.
It’s not just about efficiency, but also about equity, allowing different communities to add their own perspectives, creating a more inclusive view. It’s not about one single ‘truth’, but about embracing the richness of different viewpoints.
We want it all
As we look to the future of metadata, there are four key goals that remain essential while being bones for contention—openness, accessibility, sustainability, and persistence.
For libraries, making data open and accessible is critical, but we must also ensure that metadata creation is sustainable. Inclusivity is key. How do we involve marginalized communities while ensuring long-term viability? This requires continuous adjustments, fundings, and honest conversations.
There’s also the economic reality that open and accessible metadata comes at a cost, even if it’s not always obvious. The free model is lovely, but someone’s paying for it. Sustainability isn’t just a matter of funding, it’s about governance, long-term planning, and making sure we can preserve what we’ve built for future generations. Transparency about the funding needed to support these infrastructures is a must. We may not want big commercial entities taking control of our data, but we can’t ignore the reality that some level of profitability—or at least financial viability—is required to keep things running.
On top of this, we need metadata to be persistent. This will require ongoing investment in the systems and policies that guarantee, for example, the persistent identifiers (PIDs) and URIs in our linked data remain resolvable and usable for future generations. Persistence goes beyond technical solutions. It’s about governance, commitment, and ensuring that the infrastructure is sustainable. Libraries will need to think carefully about who finances and supports these infrastructures, and who they want to finance and support them, who they trust to do it in the ways that align with their mission and values.
Hopefully we will find a balance—between global and local, centralization and decentralization—that supports both sustainability and accessibility. However, it seems likely that this will remain a continuous challenge for some time to come.
Conclusion
Metadata work is evolving rapidly, and we must advocate for its importance. Rather than passively waiting for change, we need to take proactive steps to futureproof our work.
Futureproofing starts with revisiting our mission—what value do we create? It’s crucial that we communicate this value clearly. While the library’s role in the past was well understood, recognized and valued by library stakeholders, this may not always be true today. We can no longer rely on people knowing what metadata work involves. So as metadata teams greatly diversify their portfolio and transfer core metadata management skills far outside traditional realms of activity, this changed and complex value proposition—what we offer and why it matters—needs to be communicated effectively, both internally and externally. We also need to make sure that libraries are visible and engaged with partners on campus as well, therefore, extending their reach and influence.
By futureproofing our work and advocating for its value, we can ensure that metadata professionals remain relevant and impactful well into the future.