Skip to content

Peaking at all the Knowledge Graph Tools

Date: 2026-09-01 The dream of cool knowledge graph's has been accomplished many times but none of it sticks, like look at all these awesome examples.....

Yet I don't know anyone that uses any of them....

Above we saw some cool knowledge graph software but what about cool knowledge graph examples....

Those are some interesting exampels, but I can't imagine using any of them in my life or sharing them with friends to use.

Well back to the use case then, "How can we retroactively link videos, websites, and the world wide web in general to wikidata?" There ought to be a chrome extensions that uses AI then link everything to Wikidata? Let's check out what's out there

Well turns out they have a bunch, * Wikilinker - Chrome Web Store * Wwwyzzerdd for Wikidata - Chrome Web Store * Entity Explosion - Chrome Web Store * Knowledge Tooltip - Chrome Web Store

Okay cool so this stuff has been invented a bunch of times, but none of it actually plugs into some grand knowledge graph trying to map the World Wide Web, all these Browser Extensions generate ephemeral links using simple heuristics that don't even touch LLMs and NLP.

But what's the point of having all the web linked into wikidata, it's not like people read that many articles anymore, and when they do they do so in the little echo chamber of their own mind. I only say this because I am hanging out in the echo chamber of my own mind. If I had a list of my friends favorite top 10 youtube accounts, podcasts, writers, shows, books etc. etc. I could have a fun time mapping them out with the goal of learning about them.

Wait isin't this the same problem as the Discord Binding, a bunch of stats, or in this case links that nobody cares about? The stats the discord binding produced were like, which users posted a single message on the most number of unique days, which message got the most unique emoji reactions, what user got the most replies to their messages etc. etc. For more examples check out Questions for the Data

I stopped working on Discord Binding because I knew I needed to do NLP and Knowledge Graph stuff next but got overwhelmed with the complexity, that's when I got obsessed with Nostr and social annotation software like Margin.at, Hypothes.is, and Lantern (Nostr) but then got stuck in the weeds of the RBAC problem that none of those social annotation software solved.

It would be awesome if I had a ETL Pipeline to RDF Podcasts, Youtube Videos, Web Pages, and store the data within a single medium.