> imagine what a company with millions of dollars and a couple dozen PhD linguists could do.
Could they do much better?
Show HN: Using stylometry to find HN users with alternate accounts
311–320 of 562 posts
Re: Show HN: Using stylometry to find HN users with alternate accounts
#312Forget the alternate accounts — if two users are close in style, there’s a decent chance they should be friends. This is an HN friendship machine.
Re: Show HN: Using stylometry to find HN users with alternate accounts
#313This is an evil website. We won’t have any anonymity soon. The highest match is my years old banned account that I forgot about. Where did you get the data from?
Please don't shoot at the messenger. costco shared this voluntarily and I can see no bad intention. We should see it as an opportunity to learn how easy it is to associate different pseudonymous accounts. Nothing drives this point home better than a practical demo. We can be pretty sure stylometry is used widely by bad actors already and we should not punish people who help to spread the word about these technical po…
And this is actually quite a simple approach--which is interesting in and of itself. While there would be diminishing returns, there are a ton of other techniques you could use to make stronger inferences about similarity.
Re: Show HN: Using stylometry to find HN users with alternate accounts
#314This is one reason why I like legal doctrines such as "beyond a reasonable doubt." Even a 0.9 match in a tool like this could be a coincidence, if there are millions of users. But that won't stop people from casually believing "aha it must be an alt account", based on some anecdata.
It's so easy for something like this to be turned into a tool for a witch hunt, targeting innocents.
Re: Show HN: Using stylometry to find HN users with alternate accounts
#315Ha, gruseom shows up for pg, which is dang’s old account. A worthy successor. This is a fascinating way to find similar HN users who aren’t the same person. It’s a surprisingly great recommendation engine. “If you like pg, you might also like…” Sure, the privacy concerns are valid, but the cat’s out of the boot. Might as well enjoy the benefits. montrose is almost definitely pg. Someone who talks about ancient histor…
Why would montrose be pg ? The correlation is not that high. Looks like a few people have picked up pg's mannerisms.
Re: Show HN: Using stylometry to find HN users with alternate accounts
#316Earlier quoted context omitted.
Say you see user2 listed in bold on user1's page. That means that user1 is also in user2's top 20 users. In my experience it is often an indicator of a good match (but not always).
Huh, that's a somewhat non intuitive property.
It is a bit, but if stylometric equality was a thing you'd expect it to be symmetric, so if stylometric simmilarity is a thing....
Re: Show HN: Using stylometry to find HN users with alternate accounts
#317ColinWright is Dang?
Woah
Re: Show HN: Using stylometry to find HN users with alternate accounts
#318Earlier quoted context omitted.
> but this functionality turned out way creepier than I thought the moment I tried it Hopefully this raised awareness means that people who actually need anonymity will be more likely to know to take precautions.
Genuinely asking, what way is there to combat this? Is there a tool that takes out stylistic elements of your comment?
The site mentions a service called Quillbot which apparently does just that. https://stylometry.net/avoid
Re: Show HN: Using stylometry to find HN users with alternate accounts
#319After a few tries on boring accounts, I thought to try the account of somebody who was notorious for an incident outside of HN, and had a (deservedly) bad time at HN for a couple of years before the account went dark. And yeah, there's a bunch of high confidence (.6-.8) hits for that account, and from a quick browse of the comments of the recently active ones, they look really likely to be alts. Like, all three that…
.6 is high confidence? I did my own username, wondering what it would return, since I know I don’t have any alt accounts. The top results are in the .6-.7 range. If they aren’t alt accounts, is it just coincidence that we have similar writing styles?
Re: Show HN: Using stylometry to find HN users with alternate accounts
#320Oh god, that thing starts with direct focus on the search field, opening it showed a bunch of old nicknames, I thought it was the result of some study.