Last week I virtually attended the Knowledge Graph Conference. Among the company pitches and academic minutiae there were a few thoughtful and creative talks:

• Dave Raggett on “Representing and Reasoning with Imperfect Knowledge” using plausible reasoning. This recalled the DARPA Machine Common Sense project I was involved with.
• Peter Lawrence on “Creating a digital twin graph model”. Lawrence helped start TopQuadrant’s Professional Services team, which I joined some years later.
• Ruben Verborgh on “Web APIs For Decentralized Knowledge Graphs”, which covered some of the same ground as his blog posts on SOLID.
• Mike Dillinger on “Validating Knowledge Graph Quality”. The speaker described himself as a “free-range knowledge scientist” with experience as an epistemologist, analytical linguist, cognitive psychologist, AI technologist, and former technical lead for knowledge graphs at LinkedIn.
• Benjamin Steer presenting “An Introduction to temporal graph analytics with Pometry”. I’ve gotten into this area recently after reading a great survey on temporal networks.
• Stuart Jennings on “Graph Projects at the BBC”. The speaker has one of those charming CVs you only seem to find in Europe. He studied Classical Archaeology and Ancient History at Oxford and UCL, interned at the Royal Museums Greenwich, switched to software development, and is now a Lead Data Scientist at the BBC. The company has been involved with the Semantic Web since the early 2000s. Their editorial tone tagging in program metadata reminded me of NovelList’s appeal terms.