hiya! I like that you're thinking in terms of UI, but currently I am much more interested in the technical details behind the semantic recommender -- that sounds like a lot of work, really. hence my question as to what tools you plan to use to make all this happen.
2025-03-17 07600, 2025
JuliaHusar[m]
Hey! So most likely I feel that Python or JS with a web scraper and interaction with music brainz RDMS would allow for these connections to be identified. I say a web scraper due to the fact that sentiments can be analyzed with information that is taken off of publicly available information about albums, and then put into bins/categories
2025-03-17 07605, 2025
JuliaHusar[m]
This would require a NLP approach combined with content-based recommendation
2025-03-17 07618, 2025
JuliaHusar[m]
So an example could be extracting information about an album's themes/moods, analyzing mlp, and then quantifying that in the context of both musical genres, but also moods
2025-03-17 07638, 2025
mayhem[m]
what sites would you scrape?
2025-03-17 07633, 2025
mayhem[m]
"content-based recommendation" what does that mean? acoustic qualities of the music?
2025-03-17 07643, 2025
mayhem[m]
ah yes.
2025-03-17 07607, 2025
mayhem[m]
so, we have the AcousticBrainz project that has this info. but the underlying algorithms suck, so the data is... well.. shit.
2025-03-17 07630, 2025
mayhem[m]
that is why the project is on life support now. and spotify shut down access to their acoustic API.
2025-03-17 07646, 2025
mayhem[m]
so, content based recommendation is outside of our reach for the time being.
2025-03-17 07612, 2025
JuliaHusar[m]
mayhem[m]: Well mainly I was looking at wikipedia due to how it contains aggregate info from different review sites. RYM is protected by cloudflare + it's against their TOS so that's not really viable lol
2025-03-17 07613, 2025
mayhem[m]
WP is perfect. anything else without clear usage rights would be a much harder sell.
2025-03-17 07635, 2025
mayhem[m]
and MB has a ton of WP links so we can download that stuff right quick.
2025-03-17 07643, 2025
JuliaHusar[m]
<mayhem[m]> ""content-based recommendation..." <- Acoustic Qualities could definitely be an aspect of it, but there are also things like subgenres that better explain a particular type of music (once again, wikipedia has in depth explanations of various subgenres), as well as general descriptors that describe themes within an album, or tags that describe emotions an album evokes.
2025-03-17 07634, 2025
JuliaHusar[m] uploaded an image: (87KiB) < https://matrix.chatbrainz.org/_matrix/media/v3/download/matrix.org/zHftkXbCVLobJfcOUkhLZtHW/Screenshot%202025-03-16%20at%208.15.26%E2%80%AFPM.png >
2025-03-17 07608, 2025
JuliaHusar[m]
AOTY for example has detailed descriptions of genres. I know that musicbrainz does too with tags, but I feel that there could be a finer level of granularity
2025-03-17 07613, 2025
mayhem[m]
defining a subset of tags that express moods and using those: cake.
2025-03-17 07634, 2025
mayhem[m]
not sure if that data is statistically significant, however.
2025-03-17 07600, 2025
JuliaHusar[m]
mayhem[m]: Which data?
2025-03-17 07611, 2025
mayhem[m]
tags that express moods in MB.
2025-03-17 07620, 2025
JuliaHusar[m]
Ah
2025-03-17 07621, 2025
JuliaHusar[m]
Yes
2025-03-17 07633, 2025
JuliaHusar[m]
I'd assume the sample size is under 30 😭
2025-03-17 07643, 2025
mayhem[m]
however, we could encourage users to tag tracks with moods in order to improve how this works.
2025-03-17 07616, 2025
mayhem[m]
yeah, scraping the bottom of the barrel.
2025-03-17 07625, 2025
mayhem[m]
what else could be used?
2025-03-17 07639, 2025
mayhem[m]
have similarity data, popularity data
2025-03-17 07658, 2025
mayhem[m]
we have a neighborhood where users are compared to other users based on their listening history.
2025-03-17 07622, 2025
mayhem[m]
so you get a user's musical neighbors
2025-03-17 07629, 2025
JuliaHusar[m]
mayhem[m]: I did see that which I thought was interesting
2025-03-17 07639, 2025
mayhem[m]
I'm just throwing out what we have, hoping something adds to the pile.
2025-03-17 07657, 2025
JuliaHusar[m]
I did not understand how anything was being calculated though 😭😭😭😭😭
2025-03-17 07630, 2025
JuliaHusar[m]
I have to go to bed rn because i have a long day tmrw but I'll definitely sleep on it and talk it out with a few friends!
but lets keep chatting, this sounds super interesting
2025-03-17 07608, 2025
JuliaHusar[m]
mayhem[m]: will be looking at this!!!!
2025-03-17 07650, 2025
vardhan_ joined the channel
2025-03-17 07650, 2025
vardhan joined the channel
2025-03-17 07622, 2025
vardhan has quit
2025-03-17 07602, 2025
JadedBlueEyes has quit
2025-03-17 07615, 2025
JadedBlueEyes joined the channel
2025-03-17 07609, 2025
MyNetAz has quit
2025-03-17 07604, 2025
MyNetAz joined the channel
2025-03-17 07653, 2025
MyNetAz has quit
2025-03-17 07648, 2025
MyNetAz joined the channel
2025-03-17 07608, 2025
minimal has quit
2025-03-17 07651, 2025
minimal joined the channel
2025-03-17 07627, 2025
minimal has quit
2025-03-17 07657, 2025
d4rk has quit
2025-03-17 07622, 2025
d4rk joined the channel
2025-03-17 07631, 2025
d4rk has quit
2025-03-17 07657, 2025
d4rk joined the channel
2025-03-17 07605, 2025
MyNetAz has quit
2025-03-17 07608, 2025
vardhan_ has quit
2025-03-17 07647, 2025
NishantKumar[m] joined the channel
2025-03-17 07648, 2025
NishantKumar[m]
Hi, can anyone tell me where I can find the files like mb_metadata.xml and index.json for the project Implement a daemon that corrects out-of-sync cover art and event art metadata on archive.org? Also, do I have to resolve only this ticket [IMG-129](https://tickets.metabrainz.org/browse/IM… for this project?
2025-03-17 07649, 2025
BrainzBot
IMG-129: CAA state on archive.org out-of-sync with site
2025-03-17 07614, 2025
MyNetAz joined the channel
2025-03-17 07616, 2025
pite_ has quit
2025-03-17 07639, 2025
d4rk has quit
2025-03-17 07603, 2025
d4rk joined the channel
2025-03-17 07640, 2025
Kladky joined the channel
2025-03-17 07609, 2025
MyNetAz has quit
2025-03-17 07604, 2025
MyNetAz joined the channel
2025-03-17 07615, 2025
reosarevok[m]
akshaaatt, jasje: Dunno if MOBILE-221 is the same person I forwarded to you (the content seems familiar) but in any case, it was added to AREQ instead of MOBILE at first so letting you know in case you depend on some sort of automatic assignation for these things :)
reosarevok: did you forward my mail to the user that i sent you?
2025-03-17 07610, 2025
reosarevok[m]
Eh. No :D I never realized you only sent it to me!
2025-03-17 07616, 2025
reosarevok[m]
Why wouldn't you just send it to the user :D
2025-03-17 07630, 2025
reosarevok[m]
Ok, will send it now I guess
2025-03-17 07617, 2025
reosarevok[m]
Unless you want to just reply to them directly, which honestly would be best since you'll want to get the answer as well - just answer them, and cc support@, jasje
2025-03-17 07615, 2025
reosarevok[m]
And if for some reason you want me to send stuff, please let me know next time! :D
[00:04] JuliaHusar[m] The general ideas in these wireframes are that recommendation should be a user-centred process where users work with algorithms to find new music to listen to based on existing material they are familiar with. Essentially we are taking a user's taste, and allowing them to explore it further and make further connections.
2025-03-17 07645, 2025
Jigen
hey that sounds good! it's what *i'd* expect from recommendation thing tbh, this thing where I'm the one in controll and things based on what *I* think are specifically made the centrepiece of it. if that's what I'm getting from this message.
2025-03-17 07605, 2025
reosarevok[m]
jasje: let me know if I really should forward that email myself, anyway - but I'd still recommend you send it unless there's a reason I'm missing why that's a bad idea :)
2025-03-17 07629, 2025
Jigen
reosarevok[m]: maybe it feels intimidating to directly email other users ?
2025-03-17 07629, 2025
Jigen
I know I feel intimidated with it, I prefer it done by like "official email people guy" (ie you) or discussed publicly so that both I and the other person's opinions/reactions are vetted through
2025-03-17 07629, 2025
Jigen
often a third person's input is vital to prevent misunderstandings 🤷♂️
2025-03-17 07654, 2025
Jigen
not knowing any of the context to this btw
2025-03-17 07604, 2025
Jigen
:D
2025-03-17 07627, 2025
jasje[m]
<reosarevok[m]> "jasje: let me know if I really..." <- Nvm ill send it
2025-03-17 07604, 2025
jasje[m]
jasje[m]: Can you send me users email again?
2025-03-17 07619, 2025
JuliaHusar[m]
<Jigen> "hey that sounds good! it's what..." <- Essentially yeah, it's moreso a response to all of the AI models that seem to be taking autonomy from users
2025-03-17 07603, 2025
vardhan_ joined the channel
2025-03-17 07619, 2025
vardhan_ has quit
2025-03-17 07657, 2025
Jigen
oh deff that.
2025-03-17 07611, 2025
Jigen
I hate when computers try to do things *for* me, I hate that.
2025-03-17 07624, 2025
Jigen
but if I *tell* it to do something. that's awesome
<reosarevok[m]> "bitmap: let's put a second..." <- sure, sounds like a good plan to me
2025-03-17 07614, 2025
bitmap[m]
<NishantKumar[m]> "Hi, can anyone tell me where I..." <- given that I've received a pre-proposal from you already, I assume you've found examples of those?
<bitmap[m]> "given that I've received a pre-..." <- Yeah , but I need to install it on my local environment
2025-03-17 07612, 2025
NishantKumar[m]
bitmap: could you please give feedback on my pre proposal like anything that I missed to mention or do I need another thing to solve this issue?
2025-03-17 07622, 2025
pite joined the channel
2025-03-17 07624, 2025
mayhem[m]
kellnerd: ping!
2025-03-17 07639, 2025
mayhem[m]
or, anyone who is good with regular expressions -- I could use your help!
2025-03-17 07647, 2025
rustynova[m] joined the channel
2025-03-17 07647, 2025
rustynova[m]
I can try
2025-03-17 07653, 2025
mayhem[m]
ok, in python regex \w selects all word characters, a-zA-Z and all of the international versions as well. In C++ 11 \w selects a-zA-Z,but not the international versions.
2025-03-17 07612, 2025
mayhem[m]
if I removed all \w in python, I get:
2025-03-17 07648, 2025
mayhem[m]
"This @is an _ artist!!! (モーニング娘。)" > "This is an artist モーニング娘。"
2025-03-17 07655, 2025
mayhem[m]
in C++ I get:
2025-03-17 07601, 2025
mayhem[m]
"This @is an _ artist!!! (モーニング娘。)" > "This is an artist"
2025-03-17 07611, 2025
mayhem[m]
how can I make C++ do a python style \w ?
2025-03-17 07651, 2025
kellnerd[m] joined the channel
2025-03-17 07651, 2025
kellnerd[m]
So you want to match all letters? I'm not familiar with regex flavors in C++, but you want to use Unicode character classes, \p{Letter} specifically.
2025-03-17 07606, 2025
mayhem[m]
yes
2025-03-17 07627, 2025
rustynova[m]
That's the exact conclusion I came too. Never used c++
2025-03-17 07645, 2025
mayhem[m]
rustynova[m]: yeah. its an emergency, lol.
2025-03-17 07632, 2025
kellnerd[m]
It probably depends on the C++ library if/how they can be used, but usually you need the "u" flag to enable them.
2025-03-17 07657, 2025
mayhem[m]
text = regex_replace(text, regex("[_]|[^\\p(Letter)]+"), "")
