There's no way to keep up with all the "Canonical-isms" that show up weekly thanks to marketing trying to give things a marketing name.
Actually that is quite feasible and uses the same learning methods as the current wave of AI tools.
If a Canonicalism is introduced it is picked up. Every online document needs to be thrown into a giant mill and words (strings, tokens) pop out like a coffee bean grinder. I worked on that more than 20 years ago but today we have the modern Chat-GPT.
If I have time (not right now) I might knock together a proof of concept experiment. Just needs scraping of Canonicalisms buried deep in URL's.
Bookmarks