Treemap of the top 30 note topics we’ve classified so far from nostr.wine over ~12 hours of kind 1s. I suspect we are over labeling technology a bit but then again everything is technology. 
Login to reply
Replies (86)
I’d be interested to see if technology and bitcoin overlap some, but this seems to disprove the “nobody talks about anything but bitcoin” trope so many complain about.
I think it’s more likely that the biggest accounts talk about bitcoin, but not most accounts, if that makes sense.
Given today's date I'd say this is under counting
I expect that number to continue pumping throughout the day 😂
love what you’re doing, keep up the good work!
There is lots of overlap for sure. Check it out yourself:
Elastic
Thank you!!!
Less than I thought - 89/2290 have technology + bitcoin.
Photography? 0?
I think it goes to show a lot of content is already here. The effort must be put into curating your own feed.
This chat made me feel bullish on #Nostr . Thanks for posting this experiment. ❤️
Other is ten times the smallest section?!
I’m going to add back programming as a topic as well because I think technology is a bit generic for posts like @e6ddbf8c64cc8ef864df0d8ec0c371e8410909d67ac5c3f071f8ad0207e80262
By technology you mean bitcoin.
And greeting is the old GM GN 😅
I’ve been trying to hashtag all my code related posts #coding to help with discovery. I’ll bet this tool will be useful in identifying the best labels (by common use) to use.
Such as programming vs coding.
Other is a place holder because I’m only showing the top 30. So the 7% in other are a mix of topics shared by <0.77% of all notes. At 50 topics (in this photo), you’ll see other drops down to 2%. At 100 topics, it’s barely visible and unlabeled.
10/2300 for photography. 70 for art.
"Good morning coffechain" is what?
Food? Greeting? Diet?.... 😁
Tragedy is Iran ans Ukraine 😢
Check it out for yourself! You can query the results easily here:
Elastic
This cannot be true. Memes are at least 69%.
Only 12% greetings. We need more GMs now!
this is interesting. though i keen to know how do you decide the categories?
69.69% to be exact
GM & GN 🫂💜☕
View quoted note →
It’s very cool to see… I found it funny to see “Health & Fitness” at the bottom of the list… even a lil sad… #HealthEqualsWealth #YouAreWhatYouEatEATS #WhenYouFeelGoodYouDoGood
It’s a mix of a classification model + LLM for when none of our topics are a high probability fit.
It only shows "loading elastic" for me.
But don't worry. We just like to randomly give devs a hard time.
The stats and chart is actually very cool.
Ahhh perhaps a browser/javascript issue? Wasn’t trying to be sassy just wanted you to see/play with the data.
Short answer: it depends on how it’s worded! 
interesting!
and this data, will it be open sourced or paid kn demand?
Yeah I had a typo in coffee.
And I'm from Iran, we have many unnamed issues when it comes to internet. And everything else tbh.
For now anyone can access the results here as a guest:
Long term our plan is to allow filter.nostr.wine subscribers to create custom feeds using our topics/language/sentiment labels through our relay. We can also offer the classifier as an API (that will not be free) and classification results through an API (perhaps a freemium model). All of that is TBD. This is my weekend proof of concept.
Elastic
good job!
Here's one for the health category. I'm about to take a big ole shit 💩
View quoted note →
Well done. 
👏👏
We sure do love to greet one another here 🫂💜🤙🏻
View quoted note →
good morning
Good afternoon.
Art 2.35%
These are rookie numbers
We’ve gotta pump these numbers up
good evening
I gotchu.
Wait, where are the feet? 🤔
#footstr
No way
Uhm gotta pump the woodworking talk, I can't see it in any category in here 😁
Coming back on the monotheme of Nostr.
Lookie look, nostr is evolving from being a niche protocol to people expanding their interests 🤙
View quoted note →
Damn nice son
surface greetings and bovine tech worship has replaced meaningful discussion god worship...
are you really all this shallow or is it the fear of losing your #socialcapital that turns you into voiceless conformists ?
signs of things to come... 
Can I subscribe to fitness?
Deploying more GM, steady lads
Wouldn't it be cool if we could subscribe to these?!
#asknostr
View quoted note →
Eventually, we’ll let filter.nostr.wine users create custom feeds using our topics/language/sentiment labels.
Synchronicity is beautiful.
gm ☕️
Buen día para ti
2.07% on #memes is quite disappointing
Unfortunately, they will always be underrepresented because we aren’t currently looking at posts with ONLY images at all (mostly memes) We simply pull the content type from the url and set it aside for later classification. Eventually…
Shitposters and #meme frens, we've got work to do! 2.07% is simply unacceptable.
View quoted note →
How did you generate this?
#ga & a #ge from the uk 17:17 BST
First pass of our topic classifier + elasticsearch + Kibana. View quoted note →
Are hashtags enough to get classified into the correct category?
Great job, btw 🤙💜
As with any model, there are no guarantees but explicit hashtags of a common topic/similar to a common topic definitely carry weight.
And thank you!
As an Information Scientist, I love this
You can explore the data yourself here:
Elastic
No way its that low
🤯
Thanks, gracias 🙏🏾
View quoted note →
Treemap de los 30 temas más importantes que hemos clasificado hasta ahora en nostr.wine
Unfortunately probably victim of the same problem as memes, though they could certainly be pumped nonetheless. View quoted note →
Thanks 🤙
1.79% of posts were #DateChain (relationships) ? 😂
A categoria XXX deveria ser criada também.
This is great to see!
View quoted note →
This is a really cool experiment! Do the labels you're creating get published to the wine relay or a set of relays as well?
music is really low! let's make this place a little bit more vibey =)
Nothing yet, still a very early experiment. Lots of refining to go before its production ready but we will try to offer them in as many ways as possible!
Ahh, ok so it's not that new labeling nip32 then yet? I was reading up on that. It's a little bit confusing now to use labels the 'right' way but I'm pretty excited that some smart people are giving this a lot of thought..
Nope, nothing nostr native for now (besides the source of the notes of course). You can explore the labels here:
Elastic
I am happy with my contribution to #bitcoin , #nostr and #meme 😊
View quoted note →
nostr is 12.29% "GM" 😂
View quoted note →
Interesting. My timeline goes the other way😐😆
It'd be fascinating to watch as that chart changes over time
Great 👍
Good morning, from ring of fire
View quoted note →
GM!
Actually surprised Bitcoin is only the third biggest topic. But positivity surprised at the same time... a wider range of topics being discussed = Nostr appealing to a wider range of people.