Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Using stylometry to find HN users with alternate accounts

stylometry.net

Post reply on HN
Print view View original

551–560 of 562 posts

Re: Show HN: Using stylometry to find HN users with alternate accounts

#551

Post by InGoodFaith »

b800h wrote: post #513
InGoodFaith wrote: post #215

Earlier quoted context omitted.

What you are describing is also known as an eggcorn. https://en.wikipedia.org/wiki/Eggcorn

An eggcorn is a soundalike though, isn't it? Deliberately altering idioms to catch people's attention isn't an eggcorn IMO.

> An eggcorn is a soundalike though, isn't it?

Not necessarily, you might be thinking of malapropisms but yes probably a closer word would be the general term: protologism.

Another commenter added some useful info on the evocative alteration of metaphors [2]

1: https://en.wikipedia.org/wiki/Malapropism

2: https://news.ycombinator.com/item?id=33757097

Re: Show HN: Using stylometry to find HN users with alternate accounts

#552

Post by dysoco »

forgotpwd16 wrote: post #153
zxcvbn4038 wrote: post #99

How long until this becomes the algorithm for a dating site? “Find hot single women who write just like you”

Wouldn't be surprised if dating sites already used similar algorithms.

Do dating sites really use clever algorithms to match up people together? I was under the impression that, the less likely you are to meet your perfect match, the more you're going to use the app.

In my experience I don't see a relevant list of potential matches aside from gender and age preference, it's all completely random, even frequently I see people outside the settings I've specified (i.e. men or older women).

Re: Show HN: Using stylometry to find HN users with alternate accounts

#553

Post by fancybouncy »

bee_rider wrote: post #545
reducesuffering wrote: post #540

Earlier quoted context omitted.

> like to be clear about our unknowns or opinions. So, lots of “sounds likes,” “probably,” “could be” and so on. Commonly called just “hedging” like hedging your bets.

That’s a kinder description than I gave it in my next paragraph, so thanks I suppose. I do think it is an under-emphasized aspect of honesty, though, that we should be clear about our level of experience/understanding. Especially online — people like to discuss things, even (especially?) when we are just getting started. So if we’ve picked up opinions through osmosis and we start repeating them without testing them,…

> we should be clear about our level of experience/understanding

there are many languages that encode this info as mandatory grammatical affixes, it's called evidentiality.

Re: Show HN: Using stylometry to find HN users with alternate accounts

#554

Post by bb88 »

sillysaurusx wrote: post #52

Wow. This gives a lot of false positives, but it found all ~10 of my old accounts over the years. The most interesting thing is that my writing style changed pretty drastically since a decade ago. Searching for my oldest account matches my earliest usernames, whereas searching this account matched the rest. The details of the algorithm are fascinating: https://stylometry.net/about Mostly because of how simple it is.…

sillysaurus3 was in mine. :) Clearly we're not the same.

Re: Show HN: Using stylometry to find HN users with alternate accounts

#556

Post by matsemann »

costco wrote: post #238
dsr_ wrote: post #224

This is interesting. I'm 0.566 correlated with logfromblammo -- and while we are definitely not the same person, I could easily imagine writing a sentence such as: "For some bizarre reason, management has not yet assigned a task to their programmer underlings to automated themselves out of existence. I can't imagine why." which is theirs, not mine, from about a year ago. I like that. On the other hand, I'm nearly as…

> On the other hand, I'm nearly as correlated with peterwwillis: 0.5485 -- who has no comments and no submissions. This is due to the Firebase API not updating when users ask the admins to move their comments to another account.

Yeah, I got a good match with my previous nick here. Which to me proves the tool works well.

Re: Show HN: Using stylometry to find HN users with alternate accounts

#557

Post by nickstinemates »

The bias is interesting here.

https://stylometry.net/user?username=nickstinemates

Number 2 for me is someone I worked closely with for a few years, and then putting his name into this results in all of the people we worked with for a few years. So it seems content>style, or, we are all more alike than we thought.

Re: Show HN: Using stylometry to find HN users with alternate accounts

#559

Post by bee_rider »

fancybouncy wrote: post #553
bee_rider wrote: post #545

Earlier quoted context omitted.

That’s a kinder description than I gave it in my next paragraph, so thanks I suppose. I do think it is an under-emphasized aspect of honesty, though, that we should be clear about our level of experience/understanding. Especially online — people like to discuss things, even (especially?) when we are just getting started. So if we’ve picked up opinions through osmosis and we start repeating them without testing them,…

> we should be clear about our level of experience/understanding there are many languages that encode this info as mandatory grammatical affixes, it's called evidentiality.

I hadn’t heard of that. Neat!

I find it interesting that the first example they use in the Wikipedia article is Turkish. I’ve only met a couple Turks, but they were all quite good engineers. I wonder to what extent embedding this kind of information in the language helps organize your thoughts.

Post reply on HN