• December 8, 2021
  • By Michael E. D. Koenig Professor and former and founding dean of the College of Information and Computer Science at Long Island University
  • ViewPoints

KMWorld Connect 2021 recap: Knowledge graphs were front and center

Article Featured Image

At KMWorld Connect 2021, the dominant aspect of the presentations was the proportion of content devoted to Knowledge Graphs. The 2019 KMWorld event 2 years ago saw the emergence of substantial interest in, and attention to, knowledge graphs, a topic not much apparent at previous KMWorld conferences. In 2020, it could be proclaimed that KM had just entered stage four of KM development, and that the appropriate name for this new stage was obviously “the knowledge graph stage.” In 2021, knowledge graphs weren’t just blossoming—they were blooming—and it was clear that we are now in the knowledge graph stage of KM development.

Key characteristics of the fourth stage

What are some of the characteristics of this stage? The most obvious is the intent of trying to pull everything together into one system.  In earlier years, it was commonplace to think in terms of discrete systems—for example, a “lessons-learned” information system, an expertise-locator system, or a system to capture the expertise of persons leaving the organization or about to retire.  The focus of stage four is knowledge integration.

The enthusiasm and the world view that comes with the knowledge graph mindset is that everything is connected and from that web of RDF triples, and the ontology that it enables, you can address the questions that those lessons-learned databases and expertise-locator systems are intended to answer.

To date, the discussion about knowledge graphs has been almost entirely in the context of getting a better handle on the organization’s knowledge. Remember the mantra of “if only Texas Instruments knew what Texas Instruments knew?” However, as KMers become expert in developing and using knowledge graphs, the logical extension will be to attach those knowledge graphs to data and knowledge sources beyond the organization to in fact build a semantic web. This will be the stage or the point where KM really takes off.

But what’s in a name?

It should probably be acknowledged that “ontology stage” might well conceptually be a better name than knowledge graph stage. Knowledge graphs are powerful tools for arriving at a full ontological understanding of your domain. 

It may at this point be useful, or perhaps just nostalgic, to remember that when KM first appeared, strong arguments emerged, that other tags such as knowledge sharing would be more descriptive and preferable. Indeed, there were arguments made in some quarters that KM was so inadequate a name that it was just nonsense. But KM was the name chosen by early participants and it stuck. In this case, the name is much less of an issue; it is used merely for a stage of KM, and knowledge graphs are clearly front and center, and the most observable aspect of this new phase.

However, on the topic of KM and what it should be called, I was very much taken with the term “knowledge mobilization” used in the session, “Mobilizing Knowledge in the U. S. Army,” presented by Michael Prevou, deputy CKO, U.S. Army Training & Doctrine Command (TRADOC). Knowledge mobilization is broader, more energetic, and less passive than the term KM, and it avoids the debate about whether management is a concept that makes sense to apply to something as amorphous as knowledge. It’s probably too late now to make a change though.

Other hot topics

Another enthusiasm that surfaced at this year’s KMWorld conference is “UX design,” which is design based on or centered around user experience. UX design is not so much about any particular technique or methodology, as it is simply the well-established theme that one must design and implement KM systems not just with the user in mind but with the user actually on board. UX design is then, in a sense, simply the re-emphasis of the understanding that led to the second stage of KM—that “if you build it, they will come” is not an effective strategy for KM. The concept may not be new, but it is still vital, and it is surely salutary to keep reminding ourselves of that.

AI also continued to be a topic of some interest, but conspicuous by its absence this year was mention of the language transformer GPT-3, and its impressive accomplishments. Last year, it was mentioned at a number of sessions, but was typically presented in the context of “keep your eyes and ears open for further developments.” This year, there was no mention of any developments.

Save the Date! KMWorld 2022 will be held November 7–10, 2022 at the JW Marriott Washington DC.

KMWorld Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues
Companies and Suppliers Mentioned