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

Show HN: Hater News – Find trolls and anyone's hater level on Hacker News

haternews.herokuapp.com

Post reply on HN
Print view View original

11–20 of 57 posts

Re: Show HN: Hater News – Find trolls and anyone's hater level on Hacker News

#12

Post by dang »

valarauca1 wrote: post #4

This already made the front page once today before it was nearly at the same moment 1) Taken down via HN hug of death 2) Nuked from orbit by mods. Also I'm only at 2.5% I honestly figured I was far more toxic.

> Nuked from orbit by mods.

No moderator touched that post. It was killed by user flags. (Also, why on earth would we "nuke" a post sharing original work done with the HN API? That's why we made the API in the first place.)

https://news.ycombinator.com/item?id=8515160

Reposts of user-flagged items, though, we do tend to penalize. Otherwise flags wouldn't mean very much.

Re: Show HN: Hater News – Find trolls and anyone's hater level on Hacker News

#13

Post by KevinMcAlear »

sbenario wrote: post #8
krapp wrote: post #6

Apparently, this is my worst comment... which can't possibly be true: "Even then, you're aware of the contacts or the implant, unless you're living in the Matrix and were genetically engineered with a AV jack in your spinal column." Only 3.15%, i'm kind of disappointed with myself.

@KevinMcAlear - can you share any details on the algo? :-)

Sure! I wrote a dry blog post about it actually. http://kevinmcalear.com/thoughts/building-hater-news/

@krapp there were some challenges building out a great model but you can download the whole repo and pull out just the machine learning part and see what I did, I have it commented out in an iPython Notebook. :)

The Repo: https://github.com/kevinmcalear/hater_news

It basically uses word tokenization using scikit-learn's count vectorizer and some extra features I added like "bad words", ratio of bad words to total words used, speaking in all CAPS, and a few other features. I then took the features and use logistic regression to predict the likely hood that a specific comment is insulting then average all a user's comments into one score.

I used training data from a kaggle competition and was able to score near the same level as the winners but it will definitely be improved as I keep working on it.

Re: Show HN: Hater News – Find trolls and anyone's hater level on Hacker News

#16

Post by adventured »

"The internet is covered in trolls who are filled with hate. It's time we started to expose them and put them in their place."

Can't say I agree with that sentiment. It comes across almost as bad as the hate it proclaims to wish to expose.

I don't think the Internet is covered in trolls, the trolls are an extremely small percentage of users.

Re: Show HN: Hater News – Find trolls and anyone's hater level on Hacker News

#17

Post by valarauca1 »

dang wrote: post #12
valarauca1 wrote: post #4

This already made the front page once today before it was nearly at the same moment 1) Taken down via HN hug of death 2) Nuked from orbit by mods. Also I'm only at 2.5% I honestly figured I was far more toxic.

> Nuked from orbit by mods. No moderator touched that post. It was killed by user flags. (Also, why on earth would we "nuke" a post sharing original work done with the HN API? That's why we made the API in the first place.) https://news.ycombinator.com/item?id=8515160 Reposts of user-flagged items, though, we do tend to penalize. Otherwise flags wouldn't mean very much.

I assume flag killed was something mods could do :x I know they can super flag posts, I assumed that is what [Flag Killed] meant.

Re: Show HN: Hater News – Find trolls and anyone's hater level on Hacker News

#18

Post by dang »

valarauca1 wrote: post #17
dang wrote: post #12

Earlier quoted context omitted.

> Nuked from orbit by mods. No moderator touched that post. It was killed by user flags. (Also, why on earth would we "nuke" a post sharing original work done with the HN API? That's why we made the API in the first place.) https://news.ycombinator.com/item?id=8515160 Reposts of user-flagged items, though, we do tend to penalize. Otherwise flags wouldn't mean very much.

I assume flag killed was something mods could do :x I know they can super flag posts, I assumed that is what [Flag Killed] meant.

[flagkilled] means that a post was killed by user flags. I got tired of typing "This post was killed by user flags."

Re: Show HN: Hater News – Find trolls and anyone's hater level on Hacker News

#20

Post by cmdrfred »

KevinMcAlear wrote: post #15
cmdrfred wrote: post #14

An error occurred in the application and your page could not be served. Please try again in a few moments. aww I always get here too late for the cool stuff.

Should be fixed now! :)

It is! My most hateful comment is about the Apple Air: "Please provide an example why a screw or removable panel would provide less value to the consumer than gluing the thing and making it impossible to replace the battery on your own. I eagerly await your response."
Post reply on HN