The 20 Lessons about Building Ontologies and Knowledge Graphs

This post links to all the 20 posts I created on LinkedIn.

I just finished posting my 20 lessons from 20 years of building ontologies and knowledge graphs, every day on LinkedIn.

I’m working on a longer post that puts it all together which I’ll publish here. In the meantime, you can find the links to all the lessons here:

1. Knowledge Graphs Make LLMs Accurate

2. You Can’t Automate Understanding

3. Knowledge Graphs from Structured Data ≠ Text

4. Ontology Is a Means, Not a Goal

5. Don’t Be Pedantic

6. Don’t Boil the Ocean

7. You Need a Methodology

8. This Is Socio-Technical Work

9. KG Use Cases Are About Reusability and Managing Uncertainty

10. Application-Centric vs Data-Centric Thinking

11. You Need to Figure Out the Incentive Mechanism

12. We Need Knowledge Engineers

13. Governance Is What Gets You Out of POC

14. Identity Is the Root of Integration Pains

15. Mappings Are Where Semantics Live

16. Your First Knowledge Graph Is of Your Metadata

17. Separate Knowledge from Data

18. Knowledge Graph ≠ Graph Database

19. RDF vs Property Graph Is a Distraction. STOP!

20. We Still Lack Tooling