Reproducing Hacker News writing style fingerprinting • Hacker News Board

Re: Reproducing Hacker News writing style fingerprinting

#101

Post by hammock » Thu, Apr 17, 2025, 12:49 AM UTC

milesrout wrote: post #92

hammock wrote: post #20

The "analyze" feature works pretty well. My comments underindex on "this" - because I have drilled into my communication style never to use pronouns without clear one-word antecedents, meaning I use "this" less frequently that I would otherwise. They also underindex on "should" - a word I have drilled OUT of my communication style, since it is judgy and triggers a defensive reaction in others when used. (If required,…

Should is a commonly used word and a fine one. You should feel free to use it. If someone gets hot under the collar because you said he should do something then he is an idiot. "Ought to" is essentially a synonym. Anyone that gets upset when you said they should do something but is fine when you say that they ought to do something is truly a moron.

Most people are more moronic than one might think

Re: Reproducing Hacker News writing style fingerprinting

#102

Post by SnorkelTan » Thu, Apr 17, 2025, 1:02 AM UTC

I remember the original post the author is referring to. I was captivated by it and thought it was cool. When I ran the original mentioned in the post, it detected my one of my alt's that I forgot about. OP's newer implementation using different methodologies did not detect the alt. For reference, the alt was created in 2010 and the last post was in 2012. Perhaps my writing style has changed?

Re: Reproducing Hacker News writing style fingerprinting

#103

Post by SchemaLoad » Thu, Apr 17, 2025, 1:08 AM UTC

SnorkelTan wrote: post #102

I remember the original post the author is referring to. I was captivated by it and thought it was cool. When I ran the original mentioned in the post, it detected my one of my alt's that I forgot about. OP's newer implementation using different methodologies did not detect the alt. For reference, the alt was created in 2010 and the last post was in 2012. Perhaps my writing style has changed?

I usually just create a new account every time I get a new computer or reinstall OS. I thought most of the results here were noise, but after closer inspection it just found 10 accounts I forgot having. Actually incredible and a little scary how well it works.

Re: Reproducing Hacker News writing style fingerprinting

#104

Post by SchemaLoad » Thu, Apr 17, 2025, 1:10 AM UTC

yorwba wrote: post #67

antirez wrote: post #57

Earlier quoted context omitted.

The system uses on purpose those simple words, since they are "tellers" of the style of the user in a context independent way. Burrows papers explain this very well, but in general we want to capture low-level structure, more than topics and exact non obvious words used. I tested the system with 10k words and removing the most common words, and you get totally different results (still useful, but not style matching),…

I think grouping users by interests would be a more interesting application. Most users don't have multiple accounts, but everyone probably shares some interests with other users, whom they might enjoy discovering.

Pretty sure the point here is to demonstrate how governments or other surveillance orgs can easily find your alt accounts even if you use Tor or any number of security tools.

Re: Reproducing Hacker News writing style fingerprinting

#105

Post by throwanem » Thu, Apr 17, 2025, 1:23 AM UTC

hammock wrote: post #20

The "analyze" feature works pretty well. My comments underindex on "this" - because I have drilled into my communication style never to use pronouns without clear one-word antecedents, meaning I use "this" less frequently that I would otherwise. They also underindex on "should" - a word I have drilled OUT of my communication style, since it is judgy and triggers a defensive reaction in others when used. (If required,…

I would prefer the "analyze" feature focus on content rather than structure words. I forget the specific linguistic terms but to a first approximation, nouns and verbs would be of interest, prepositions and articles not. Let's call the former "syntactic" and the latter "semantic."

I suppose it's possible the "analyze"-reported proportions are a lot more precise and reliably diagnostic than I imagine. I haven't yet looked in detail at the statistical method.

Also, of course, it would require integration with NLP tooling such as WordNet (or whatever's SOTA there something like a decade and a half on) and a bit of Porter stemming to do part-of-speech tagging. If one 0.7GB dataset is heavyweight where this is running, that could be a nonstarter; stemming is trivial and I recall WordNet being acceptably fast if maybe memory hungry on a decade ago's kinda crappy laptop, but I could see it requiring some expensive materialization just to get datasets to inspect. (How exactly do we define "more common" for eg "smooth?" Versus semantic words, all words, both, or some combination? Do we need another dataset filtered to semantic words? Etc.)

If we're dreaming and I can also have a pony, then it would be neat to see both the current flavor, one focused on semantics as above, and one focused specifically on syntax as this one coincidentally often seems to act like. I would be tempted to offer an implementation, but I'm allergic to Python this decade.

Re: Reproducing Hacker News writing style fingerprinting

#106

Post by declan_roberts » Thu, Apr 17, 2025, 1:26 AM UTC

This is exactly why HN needs to allow us to delete accounts.

Re: Reproducing Hacker News writing style fingerprinting

#107

Post by Loughla » Thu, Apr 17, 2025, 1:30 AM UTC

milesrout wrote: post #92

hammock wrote: post #20

The "analyze" feature works pretty well. My comments underindex on "this" - because I have drilled into my communication style never to use pronouns without clear one-word antecedents, meaning I use "this" less frequently that I would otherwise. They also underindex on "should" - a word I have drilled OUT of my communication style, since it is judgy and triggers a defensive reaction in others when used. (If required,…

Should is a commonly used word and a fine one. You should feel free to use it. If someone gets hot under the collar because you said he should do something then he is an idiot. "Ought to" is essentially a synonym. Anyone that gets upset when you said they should do something but is fine when you say that they ought to do something is truly a moron.

The only time to avoid command words like should is when the person could conceivably see them as a command. Because then you're being a dick.

Otherwise, if someone wants to take the time to dissect meaning from add-on meaningless words like should in a sentence, they should find something better to do with their time. Or just ask instead of being a moron.

Re: Reproducing Hacker News writing style fingerprinting

#108

Post by Loughla » Thu, Apr 17, 2025, 1:32 AM UTC

gus_massa wrote: post #82

weinzierl wrote: post #18

How does it find the high similarity between "dang" and "dangg" when the "dangg" account has no activity (like comments) at all? https://antirez.com/hnstyle?username=dang&threshold=20&actio...

It has a [dead] comment that says > Please don't post unsubstantive comments to HN. [link to guidelines] My guess is it was a parody/impersonator account. You can enable "showdead" in your profile to see [dead] comments ans posts. Most of them are crap, but there are some false positives an errors from time to time.

I like to leave dead comments on. It helps me feel better about living in a cultural wasteland to see that people who probably live near urban centers can be just as dumb! It really does help me worry less.

Re: Reproducing Hacker News writing style fingerprinting

#109

Post by scoresomefeed » Thu, Apr 17, 2025, 2:25 AM UTC

The original version nailed all of my accounts with terrifying accuracy. Since then I make a new account every few days or weeks. Against the rules I know. And I’ve learned a lot about HN IP tracking and funny shadowbanning-like tricks they play but dont cop to. Like I get different error messages based on the different banned ips I use. And j see different behavior and inconsistency with flagged messages (like one that got upvoted a day after it was flagged and not visible to other users).

Re: Reproducing Hacker News writing style fingerprinting

#110

Post by 6510 » Thu, Apr 17, 2025, 2:32 AM UTC

legohead wrote: post #62

Frieren wrote: post #33

It works for me. The accounts I used long time ago are there in high positions. I guess that my style is very distinctive. But I also have seen some accounts that seem to be from other non-native English speakers. They may even have a Latin language as their native one (I just read some of their comments, and, at minimum, some of them seem to also be from the EU). So, I guess, that it is also grouping people by their…

Didn't catch my original account when I tried it, not anywhere in top 100. But, if I do the reverse (search using my original account), this one shows up as #2. The main difference between the accounts is this one has a lot more posts, and my original account was actively posting ~11 years ago.

I never knew A can be like B without B being like A.