Getting Started with Knowledge Graphs and Machine Learning: Part 1 Q&A with Sebastian Schmidt, CEO of metaphacts

Video produced by Steve Nathans-Kelly

Today, organizations need to make information accessible to all their users, not just a select few. But getting information to the people in an organization who need it, when they need it, continues to be a widespread challenge.

In a recent KMWorld Drill Down Video Interview, Sebastian Schmidt, CEO of metaphacts, explained how to help knowledge workers gain direct access to actionable data for faster, more informed decision making.

Joyce Wells: We are talking about knowledge democratization and end-user empowerment with knowledge graphs. But first, can you share a bit about metaphacts?

Sebastian Schmidt: We are a software company based in Walldorf, Germany. We were founded in 2014 and serve customers worldwide. At metaphacts, we help our customers unlock the value of their data assets to drive digital transformation, and we do this by focusing on empowering business users and enabling knowledge democratization using knowledge graphs.

JW: Could you explain what knowledge democratization refers to for anyone who is not familiar with it?

SS: When we talk about knowledge democratization, what we are referring to is making actual actionable knowledge easily accessible to everyone in the enterprise.

Gartner calls this "citizen access." So, if you think about it, knowledge democratization is, obviously, largely a behavioral trend. It's something that requires cultural transformation across the entire organization. It especially requires giving access to and control over data to domain experts and business users, something that is mostly embedded into the IT team today. What we need are systems and tools also that allow these users to find, access, share, and re-use information on demand without having to involve an IT specialist.

The data democratization systems and tools that are available in the market today are mostly made for data engineers or IT professionals with an extensive knowledge of data management practices and processes. Business users and domain experts are really outside of the process, even though they are the ones who mostly need to consume this information and need it for their daily work. What many companies are trying to do today to address this issue is, they are working on data catalogs—which is a very important part to get structure into the data chaos—but, data catalogs alone, to my understanding, will not enable knowledge democratization because what knowledge democratization requires is not just this inventory of data, but the integration of the data itself. And, most importantly, it requires a structured, ideally semantic description of the data so everyone can understand it and contribute to it.

JW: How can an organization enable knowledge democratization?

SS: The first ones who thought about this and were really pushing the boundaries were knowledge workers in the research community. They pinpointed this challenge and saw that a lot of the research data produced started to go beyond manageable proportions. And a lot of that data was not in a form that a researcher could easily access. It was mostly published in the form of documents, and the underlying data and derived knowledge could only be retrieved by re-extracting the information back from the document, which is a very error-prone and tedious process, because you really need to go back to trying to understand how this information was collected and compiled and how knowledge was derived from that.