Cleared a bunch of notes from #OpenStreetMap. Did some ground truthing on my bike yesterday and then did the updates today. My helmet cam was being uncooperative and I lost some footage but I was still able to get what I needed.
Still lots of open notes but I put something of a dent in them.
The Groundtruth project has documented the costs of Trump's cuts in detail.
From Arkansas to the Amazon, Pennsylvania to Pakistan, they are staggering. This is hurting a wide swath of people across the nation and the world.
https://storymaps.arcgis.com/stories/69072891d2694f278d9cf168dfc1a006
New dataset from BL colleague Alex Hailey: Ground truth transcriptions of 18th &19th century English language documents relating to botany from the India Office Records; related datacard and blog posts https://doi.org/10.23636/gfva-yn20 #Transkribus #dataset #GroundTruth #transcription
In #machinelearning, #GroundTruth is a term of #art.
@Cdespinosa
In #MachineLearning, #GroundTruth is a term of #art.
I think Israel should exist.
I think Palestine should exist.
I think Ukraine should exist.
See?
Not hard.
Ooh, addresses and what's displayed on the buildings... this goes against the #GroundTruth principle of #OpenStreetMap, right? :)
#SOTMBaltics #BalticGIT2023
@pesasa Gah! I'm misreading my own previous thread 
And yes, I'd introduced belief there, so that's on me...
Context: that toot was based on an earlier HN thread. I'd looked to that to try to suss out What I Was Even Thinking in writing that, and the context in which it emerged. That's well worth taking a look at (it's a short thread).
To clarify for discussion:
The grid is based on "known" and "unknown", and "truth" and "falsity". K, U, T, and F.
Apparently I'd equated or at least connected "known" with "belief" writing the toot in Sept., 2021. That is, "Believed truth".
"Ground Truth" status is the T/F dimension.
By way of clarification, I'm reading myself as saying that T/F are independent of knowledge or belief, that is, T/F is ground truth, not some awareness or belief, which ... falls more into the K/U axis. A belief is a subset of that which is known.
That noted, the grid is what I call a naive extrapolation based on a state matrix. If you have knowledge and truth each with two states each, you can expand them to the sixteen combinations shown. And as I'd originally said, I'm not sure how useful that is, though it seems interesting.
as I'd said on HN:
In the known-knowns model, you have knowledge and metaknowledge (what you know, what you know you know).
That is, "unknown unknowns" is about your awareness of your own scope of knowledge and ignorance.
In that context, then your comment seems accurate: an unknown unknown is not a believed truth. It is in fact a domain in which there is neither belief or awareness that a belief need be necessary.
That leaves us with the question of a fact falling into the status of unknown unknown, but having possible truth values. And again, I'm not sure how useful that is, outside simply filling out the combinatorial matrix.
