Around 1 year ago we got hit badly on our [blogging platform][0] by people/groups submitting fake customer support description of other big companies, either being Microsoft, Facebook, Comcast etc. Rolled out a machine learning model and trained it on the database. 99% of them vanished. Next day, the machine didn't work and success rate was around 5%. Found out, they have learned the trick and now using symbols from…
You either die an MVP or live long enough to build content moderation
171–180 of 429 posts
Re: You either die an MVP or live long enough to build content moderation
#172Earlier quoted context omitted.
The internet is not moderated in any useful sense for content. Drug markets like white house market, and before that silk road have perpetuated for years. Tor and other darknet websites host content that is nearly universally disdained by governments and even most individuals, which I hesitate to even name here what that heinous content is (you and I both know some examples). > We assume NASA's site has NASA-posted c…
>The internet is not moderated in any useful sense for content. Drug markets like white house market, and before that silk road... You mean the Silk Road that the US government "moderated" out of existence, along with other Tor marketplaces over the years? The same ones that suggest White House Market's existence is also likely to be limited?
Re: You either die an MVP or live long enough to build content moderation
#173Earlier quoted context omitted.
I built a Python library for finding strings obfuscated this way. Was critical when moderating our telegram channel before an ICO. https://github.com/wanderingstan/Confusables E.g. "𝓗℮𝐥1೦" would match "Hello"
You should post this to Show HN. Also you have a typo on your README ("characgters")
>>> import unicodedata
>>> unicodedata.normalize('NFKD', '𝓗℮𝐥1೦𝗵𝗲𝗹𝗹𝗼')
'H℮l1೦hello'
Obviously it isn't going to remap leetspeak characters like 1 -> l but it covers a lot of cases.Re: You either die an MVP or live long enough to build content moderation
#174Here's a thought. Have a platform where identity is verified. Users can post publicly or within their circle. Any illegal or fraudulent content can be handled by the legal system due to the lack of anonymity. Beyond that, let users form groups for topics like reddit. Where does this fall down?
What I'd like is something identical to Facebook (preferably hosted in a box on my desk) that I invite all my friends to, and they have the ability to invite people to. 2nd gen is probably far enough. Anyone cuts up rough, I go by their house and make them miserable.
Re: You either die an MVP or live long enough to build content moderation
#175Earlier quoted context omitted.
For adversarial problems like this, a shadowban approach can sometimes be necessary. Perhaps people can still see their blogs but GoogleBot gets blocked from indexing them, or they only appear to someone with the spammer's cookies. That way it takes them longer to catch on and evade the model. Of course, that means you'll need to at least spot check your bans because you can't rely on legit users escalating to you.
Yep, shadow ban was in place as well. The thing is, the people that weren't the ones posting the content. It appeared their computer was affected by some type of malicious file to be part of a bigger network. (botnet?) I could see from the thousands of different IPs in different countries around the globe, that it could be affected personal computers. Very few of them were computers from hosting companies, the rest w…
Re: You either die an MVP or live long enough to build content moderation
#176Earlier quoted context omitted.
What I'd like is something identical to Facebook (preferably hosted in a box on my desk) that I invite all my friends to, and they have the ability to invite people to. 2nd gen is probably far enough. Anyone cuts up rough, I go by their house and make them miserable.
You never plan to have any friends beyond your immediate neighborhood?
Re: You either die an MVP or live long enough to build content moderation
#177Earlier quoted context omitted.
I always read "dead" comments. I believe that many of them should not have been "killed".
Reddit rule of thumb: the downvoted comments are far more likely to introduce a new idea than the upvoted ones. However, it's like comparing the likelihood of dying in a car crash instead of a plane wreck: both events are rare enough as not to be worth considering in planning your daily life.
Re: You either die an MVP or live long enough to build content moderation
#178Earlier quoted context omitted.
Yep, shadow ban was in place as well. The thing is, the people that weren't the ones posting the content. It appeared their computer was affected by some type of malicious file to be part of a bigger network. (botnet?) I could see from the thousands of different IPs in different countries around the globe, that it could be affected personal computers. Very few of them were computers from hosting companies, the rest w…
Did they map to ISP ASNs? Country geolocation doesn't say much anymore since there's so many VPN providers whose business is to buy a CIDR in every country and resell access.
Very few of them were from AWS, OVH and other hosting providers, very very few.
We ran each IP towards black lists ips, paid IP reputation checkers. Majority of the IPs were clean.
Back then we had IP reputation check, but it was a headache to maintain, so we disabled it later, however that time very very few them got stopped at IP reputation checks.
Re: You either die an MVP or live long enough to build content moderation
#179This article opens up with an anecdote about policing morality and imposing one's own local norms on the legal speech of others. NSFW is a euphemism. Please don't push your American puritanical worldview on every user of your UGC platform.
If they own the platform, why shouldn't they get to make the rules?
Practically, it makes them an asshole.
Just because you are legally entitled to do something doesn't mean it's good for your users or society: e.g. the widespread practice of IAPs, or Apple's censorship of the App Store.
Re: You either die an MVP or live long enough to build content moderation
#180Earlier quoted context omitted.
If you can identify text written with mixed glyphs just ban it outright. Normal users don't use text like this, the pure binary presence of such "homomorphic" text at all is probably a better signal for spam than whatever your neural net when running it after normalization.
> Normal users don't use text like this Sounds like you live in a filter bubble. (╯°□°)╯︵ ┻━┻ (ノ◕ヮ◕)ノ*:・゚