Dan Romero pfp
Dan Romero
@dwr.eth
A few other thoughts 1. We use a machine learning algorithm to classify spammy users. 2. That requires a manual input when new spam "metas" emerge. 3. Best way to model spam (and fraud or any other "gaming" of a system) is like an organism that evolves based on the environment (the social network) and the various inputs into the environment, i.e. antibiotics / spam filters 4. So what works effectively today will not be effective tomorrow (antibiotic resistance!). 5. So you have to flag the new metas to the model using signals—user reports, manual classification—and then the model can relabel users.
7 replies
28 recasts
28 reactions

Theleroi.eth pfp
Theleroi.eth
@theleroi
Another question: How does the report system work, and how do we value the different user reports? Can someone behind a big bot-net of bot users spam report a user to trick the algorithm in order to put that reported user under spam label?
1 reply
0 recast
2 reactions

basement5k.⌐◨-◨ 🟡 pfp
basement5k.⌐◨-◨ 🟡
@basement5k.eth
Just checked my following since the update. about half my followers are gone. I honestly dgaf about how many followers I have, im actually stoked that there are still 2k left after the update. Keep trimming the fat. sincerely, Ratio Guy
0 reply
0 recast
3 reactions

Icetoad 🎩🍕 pfp
Icetoad 🎩🍕
@icetoad.eth
Thanks for looking into this. I know we don't see eye when it comes to politics but if you can't tell I am bullish on Farcaster and want to see you, the team and everyone on here succeed :)
0 reply
0 recast
3 reactions

0xChris pfp
0xChris
@0xchris
Thank you for investing time, money, and effort into this. This assignment reminds me of the concept of insurance: it may not seem like a way to maximize wealth, but it’s essential for maintaining the system’s value.
0 reply
0 recast
0 reaction

lucas pfp
lucas
@elesel.eth
Kind of sounds like the machine learning algo which eventually has stored every single available spam meta could be a valuable product by itself.
0 reply
0 recast
0 reaction

Juliettemeon.degen.eth pfp
Juliettemeon.degen.eth
@juliettemeon.eth
Hi Dan, why hasn’t the number of followings decreased?
0 reply
0 recast
1 reaction

Jessi ⌐◨-◨ 🤫🌭  pfp
Jessi ⌐◨-◨ 🤫🌭
@zwillow
There are lots of accounts marked not spam, but their bios had “airdrop hunter, what are we farming” etc in them. Clearly spam/farmers. Then my friends IRL were labeled spam because they haven’t casted much. Definitely agree with the purge. Went through one by one and checked who I was following. Helps to be more mindful. Thanks for everything, was painful but good. ;)
0 reply
0 recast
0 reaction