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

431–440 of 562 posts

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

#431

Post by highwaylights »

drc500free wrote: post #155

This is a super interesting tool for self reflection. Looking at the top 10 similar accounts to mine, it gives me an arms-length view of how other people probably interpret my tone. I appear to be a well-educated, over-confident know-it-all.

Same. Looking through some of the handles on my list tells me that I come across like a not-particularly-well-educated McSmug that needs to take a good long look at myself. Wouldn’t be so bad if I wasn’t reading the posts thinking I definitely could see myself writing this.

This was certainly eye-opening.

Update: It’s actually a little strange that reading through some of the matches it’s not just style that overlaps but perspectives in quite a few cases too. I’m definitely not the unique little snowflake that some others are finding themselves to be.

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

#433

Post by costco »

jonas-w wrote: post #426
costco wrote: post #81

Earlier quoted context omitted.

HN Firebase API. I just wrote a program in C++ with libcurl to get https://hacker-news.firebaseio.com/v0/item/1.json , https://hacker-news.firebaseio.com/v0/item/2.json , https://hacker-news.firebaseio.com/v0/item/3.json , ...

Why didn't you use the google bigquery? https://news.ycombinator.com/item?id=10440502

I was aware there was a HN dataset on BigQuery but I had never used a library to work with it before and when I played around on the website the posts I got were all from 2015 at the latest. It probably would've made my work easier but there's not really anything I can do about it now.

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

#434

Post by 00F_ »

stupendous_luck wrote: post #271
00F_ wrote: post #57

ive had maybe a hundred throwaway accounts on HN over the past ten years. generally, i make an account, say something that is apparently wildly offensive to someone else, get flagged and down-voted and then muted or hell-banned. then i make another account because i never did anything wrong and start the process over again. ive emailed the admins, tried to reason with the admins, it never does any good. the power is…

You really don't need advanced AI to do it. Just a bunch of scrapers and some run of the mill statistics. And guess what, it's been done by many companies already. They just don't care to create such a site.

you have no idea what im talking about. you dont realize how much data is out there. you dont comprehend how much smarter than you something can be.

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

#435

Post by stavros »

rogual wrote: post #391
stavros wrote: post #13

Oh wow, it's really sure that I'm stavrosk, which I am: https://stylometry.net/user?username=stavros The next person is 30% less certain, that's huge! This would basically identify any alt I might have with near certainty.

Funny thing is, it thinks I'm you, but it doesn't think you're me! https://stylometry.net/user?username=rogual I'd have thought this stylometry thing would be commutative.

I guess it's a multidimensional space, so you can have someone closer to you than me, but they aren't also closer to me than you. Basically, they're close to you, but on the "other side" of me, I guess?

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

#436

Post by MikePlacid »

jll29 wrote: post #216

The method used, i.e. to calculate the cosine of the two authors' word vectors, is poorly suited for stylometric analysis because it is based on a poster's lexicon and the word frequencies of each word, but ignoring stylistically relevant factors like word order. Also, the cosine of the vectors of word frequencies conflates author-specific vocabulary and topics; in other words, my account is grouped (with >51% simila…

> based on a poster's lexicon and the word frequencies of each word, but ignoring stylistically relevant factors like word order.

Interesting. I was expecting to be grouped with other Russian speakers and I am (based on some nicknames). But I thought the most telling feature will be exactly word order - it’s absolutely relaxed in Russian. Word frequencies? Well, probably the absence of articles, lol (but I swear to God that I often spend some extra time trying to insert as many articles in my texts as I could).

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

#437

Post by chriskanan »

I have no alternative accounts besides making a single throwaway account to post one "Ask HN" five years ago, but I have a decent number of matches above 0.5. I think this is due to the relatively uniform style of "who is hiring posts," since my matches did that in a similar way for other companies. I made many of those for about two years when I was at a start-up.

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

#439

Post by drc500free »

seydor wrote: post #236
drc500free wrote: post #155

This is a super interesting tool for self reflection. Looking at the top 10 similar accounts to mine, it gives me an arms-length view of how other people probably interpret my tone. I appear to be a well-educated, over-confident know-it-all.

we must be a good match

I'd love a version of this where you enter two usernames and get a match score.

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

#440

Post by lzooz »

meta2023 wrote: post #403
lzooz wrote: post #374

Earlier quoted context omitted.

I don't think there's a public list of all HN usernames is there? Found this, it includes 250k usernames, but it's not there. https://www.kaggle.com/datasets/hacker-news/hacker-news-corp...

The username in question isn't in this dataset but maybe it was created in the past 10 days, as the max(timestamp) is Nov 16th, 2022. https://console.cloud.google.com/marketplace/details/y-combi...

It isn't there, and given the "story" it happened years ago so it should be there, so I guess we've been played.
Post reply on HN