EKG Foundation Preps Industry Consultation, Touts Benefits of Knowledge Graphs for Capital Markets

The Enterprise Knowledge Graph (EKG) Foundation, a new industry association co-founded by data management industry veteran Mike Atkin to promote using Enterprise Knowledge Graphs for data management, is finalizing the maturity model that outlines knowledge graphs. It plans to open it up to the market for further development before undertaking a benchmarking study on the state of industry adoption.

EKGs link individual data points by defining data using web standards and creating an ontology that identifies and maps data. EKGs can do this without needing traditional column- and row-based databases, enabling firms to incorporate a broader range of data into their operations and to connect that data in a way that allows them to make greater use of it.

The EKG Maturity Model sets out the expected evolution of EKG usage, starting in 2019 and extending to 2027 and beyond. It identifies four key focus areas—business, data, technology, and organization—that govern every aspect of what must be achieved to adopt and gain value from using EKGs for enterprise-wide data management. Fourteen strategic components feed into those four areas and a further 45 specific functional capabilities sit beneath the components.

Atkin, who serves as director of the EKG Foundation, says he will give away the work he has completed on the model so far to encourage involvement and contributions from interested parties in the financial services industry and beyond. This will help build a more comprehensive consensus model and provide a means by which firms can measure their progress. After that, the Foundation will conduct a benchmarking study that measures how peers are progressing, how quickly plans are developing, how to define a center of excellence, what are inhibiting factors, and how to evaluate the technologies and platforms used.

Atkin says there are already mixed levels of experimentation underway with EKGs among financial markets participants. He estimates that roughly one-third have not yet started, one-third are experimenting, and one-third have reached the second of five defined stages. The latter are leveraging EKGs to re-use data across business lines with a small percentage having gone fully operational enterprise-wide.

“Many firms recognize the liability associated with data siloes … where everything is structured in such a way that it doesn’t allow you to link data together to take advantage of upselling opportunities, or to perform research or analytics,” he says. “When they recognize there is a better way of doing that, and how standards can help solve the problem, you can see an awakening take place: people need to be able to envisage a different world. We can fix this problem, it’s not expensive—you don’t have to rip-and-replace your existing infrastructure; you just have to map it—and you can get operational savings of perhaps 30%, and your analysts will be able to understand all the complexities hidden in the data.”

That estimated 30% saving comes from eliminating full-time employees currently manually validating data against structured columns and rows. However, Atkin acknowledges that there will be some initial investment involved in adopting EKGs.

“It’s new, so there is a skills gap … and people have built fiefdoms around existing technologies, so there will be resistance,” he says. “But the value of the Foundation will be helping people through that.”

However, that cost saving is the tip of the iceberg in terms of the benefits EKGs can bring, Atkin says.

“For risk analysis, cybersecurity, connected inventory, and flexible analysis … all the things we want to do with data become possible and affordable,” he says. “For example, take the Internet of Things—we’re collecting all this data from IoT devices, and we’re only using 1% of it. We’re not using all that data because we don’t know how to link and integrate it … but now, all of a sudden, we’ll be able to use all of that data.”

EKGs will also help reduce the burden of presenting similar datasets in distinct ways to serve different use cases, such as sharing data within the enterprise to identify business opportunities versus sharing the same underlying data with regulators to meet compliance requirements. Instead of being completely separate data-mining exercises, in a semantics-governed world, this could be achieved with just a business rule change, Atkin adds.

“The first thing we see financial institutions using this for is ‘connected inventory’—building a repository of your core assets, so your technology assets, your people assets, etc., and being able to see relationships between them, so you can see activity and identify redundant systems, for example,” he says. “But it’s things like transaction-product linkage that enables you to perform Anti-Money Laundering and Know Your Customer tasks, and upselling—that’s what everyone has their eye on. That’s the goal of this.”