Pieter Colpaert

thoughts

Subscribe to the atom feed

• 2026-01-08

  Eventual Interoperability

  A pragmatic approach to interoperability: model your data richly for your own needs, then allow the eventual creation of one‑to‑many alignments.

  interoperability
• 2025-11-28

  SEMIC2025 Trip report

  In November 2025, I went to the SEMIC conference in Copenhagen. This is the trip report, in which I talk about data spaces, LDES and the EIF.

  conferences
• 2025-09-30

  Named graphs and RDF messages

  I’ve seen named graphs being called the most horrendous mistake in RDF standardization, to a necessary feature for which, without it, RDF wouldn’t be where it is today. In this blog post I try to explain my interpretation of all this, and how I believe named graphs are here to stay. I also introduce a future outlook, in which I believe we need to talk about RDF messages.

  linkeddata
• 2025-09-08

  SEMANTiCS2025 Trip report

  I went to SEMANTiCS2025 with the team in Vienna. I will especially remember: the keynote of Hannah Bast, the fish bowl discussion, and the work on the Jelly serialization. Our team took on quite some responsibilities as well, organizing a bunch of workshops and a tutorial.

  conferences
• 2025-09-03

  Four types of specification artefacts

  Interoperability isn’t about creating the one standard to rule them all. It’s about reusable artefacts. I argue there are four types: vocabularies, application profiles, interaction patterns, and implementation guides. Together, they form a toolbox for connecting systems across domains.

  interoperability

talks

• 2021-03-16

  Linked Data Event Streams as the base API for European datasets

  ENDORSE2021 — keynote

  Keynote at the ENDORSE2021 conference introducing Linked Data Event Streams

• 2019-10-09

  GraphQL vs. REST for Open Data

  GhentJS — dev-talk

  Talking about GraphQL vs. REST:

• 2018-07-11

  Open data to create power for the many, not the few

  TEDxUCLouvain 2018 — TEDx

  The story so far with Open Data: we’re not there yet. Let’s make sure data can be adopted by the many, by making sure a machine can access it for you.

• 2017-09-22

  Keynote on Open Data: a Belgian perspective

  Open Sea Lab — Keynote

  Answers to questions what Open Data, Linked Data and Querying Web APIs means, with examples from Belgium.

• 2017-02-04

  Predicting the crowdedness of your next train

  Data Beers Brussels — Lightning talk

  In 2017 we ran a successful crowdfunding campaign for getting to know how busy your next train will be. We ended up coding this into iRail.be!