Evolving Semantic Modeling: Balancing Flexibility and Governance

·Apr 29, 2025 03:59 AM

Hi everyone, I’ve been thinking a lot about how semantic modeling might evolve — drawing lessons from how software engineering matured (C from assembler, Git from CVS, etc.). I wrote up some reflections here Curious whether others here feel the same tension:

We want flexible graphs, but we also need governance, versioning, and structured evolution if we’re going to scale.

(For context: I’ve been working on these challenges for a while, including building a platform that’s now being used in initiatives like UN/CEFACT’s Transparency Protocol. But this post isn’t about promoting tools — it’s about whether the structural challenges and opportunities I’m seeing match what others are experiencing too.)

5 comments

· Sorted by Oldest

Andrew P.
·
yes been thinking alot about this for longer than I am willing to admit!
Alastair
·
Ha — same here. I’ve been working on applying proven practices from software engineering — like versioning, dependency management, and governance — to semantic modelling. Some of it’s already being used in projects like the UN/CEFACT Transparency Protocol, which has been a great pressure test. Always keen to swap ideas. I honestly think this is one of the big hurdles stopping semantic modelling from going mainstream.
Bob DuCharme
·
You may find https://www.bobdc.com/blog/selling-rdf-technology-to-big/ interesting, in which I described thirteen years ago why I stopped using the phrase "semantic web". The hot thing at the time was "Big Data," so I was addressing that context, but the ideas still hold, and RDF-related technology continues to do more great things. (For most people, forgetting about OWL helps, so you may also find https://www.bobdc.com/blog/dontneedowl/ interesting.) All of the ideas in your piece make sense, but arguing against aughts-era notions of the value of RDF technology is kind of shooting fish in a barrel.
Ellie Y.
·
Alastair, Andrew P., did anything come from your conversations? Michael DeBellis is working on ideas now to describe an agile framework for the development of knowledge graphs - I think this is one part of a larger "data supply chain" , fuelled by knowledge graphs -- Jessica T. writes a lot in this vein as well (and collabs with Andrew)
Andrew P.
·
Hi Ellie Y. I haven’t participated in any followup with the people on this thread