Yeah I am really not sure what it really wants to measure; I came back with a 1.69% Mr Rogers level of hate. My worst comment was referencing Apple product pricing as a dick move.
I would be more interested in measuring how often a person up/down votes, how often they are up/down voted compared to others within a similar posting. It is one thing to get down voted into oblivion but if you have a lot of company does that mitigate it?
I like the idea of scoring people on their "troll" number. This might fix other conversation sites where there's lots of noise. Someone could sell the solution to the Wall Street Journal.
Actually disappointed I scored so low on this. I think it's healthy for community discussion that HN users to have sizable hater scores - otherwise we're a bunch of yes people (like me, it seems).
"Ah ok. I looked at it more as something you force on your co-workers :)"
Looks like 'ah' implies negativity to the algorithms driving this.
Edit: It's also only pulling your 50 most recent comments by default. That covers about the last two weeks for me and is probably why it isn't very accurate. 50 comments is a very small dataset to work from. The other option it gives is 50 oldest comments - again not very representative for someone who has been commenting regularly for 1301 days. I try to keep my comments useful but I know for a fact I've gotten into emotional debates in the last 4 years in which I was negative/stupid/hating. Maybe it's not possible but I would suggest pulling a random sample of 100-200 comments for more accurate data.
Refreshing, isn't it? It give us a semi-sensible insight on how others may perceive our comments. One of the biggest things I learnt in Australia is never to use irony when you're a foreigner. People either wonder whether you said what you intended to say or it delays their understanding of the joke, if there was any. It's sad, because English coworkers love this kind of humour and they're excellent at not smiling when they do it.
Great idea! I'll make another option that does exactly this. I think it's potentially a better way to think about it. Since I have to make an individual call for each comment it gets pretty hairy past 50 but I'm sure I can do it differently in the future. I tried to not put a limit on it at first and realized that @pg has something like 13,000 comments... lol. I could also just pull back every user's comments and run them through the model off line and update it every night. I just need a list of every user :)
My most negative comment also had a simley face, and can totally see that it's quite negative without it. It's pretty terse and sharp sounding.
*"Fitting all that into 7 minutes will likely lead to injury for someone who is in the audience for any kind of short N minute workout, IMO. I concur with your general sentiment and good intentions though. :)"
I wrote this a couple days ago though, and am a bit miffed because I would have expected to have written a more acerbic comment sometime in the past.
"Yeah, that book I linked is the same guy as that site you linked. Good stuff." is the worst thing I've said according to this. I'm positive I was much closer to trolling territory on that Angular 2 thread a couple days ago.
Comment listed as my worst (without enabling "Back In The Day" option, id=7642888) is presented with wrong link to my other comment (id=7701968). Chronologically 7701968 is my next comment after 7642888. Off-by-one error somewhere?
Hello! I wrote this in on of the other threads about so I figured I would leave it here too.
A few things:
1. Thanks for posting my blog post (https://news.ycombinator.com/item?id=8517727) @chippy. :) The actual app ( haternews.co ) kept getting booted off HN... And now thanks @melling for posting it.
2. There have been a lot of interesting comments on the three (now four) threads on here. People pointed out some bugs and overall issues which I will be fixing (also, the site should not crash half as much now). This is just a fun side project I have been messing around with so I can get better at using data science in various applications. If you would like to help build it out for fun further let me know! Also, feel free to submit a bug or suggestion for an improvement if you really want to.(https://github.com/kevinmcalear/hater_news/issues)
3. I wanted to build the "hater score" for two reasons. First, to see how accurately I could build a model to measure insulting comments in the wild and second (if it's accurate), to see how people would react to seeing how positive or negative they usually are on Hacker news (or other social networks).
4. I wanted to make sure everyone knows that just because something is your "Worst Comment" doesn't mean it is negative. Most people have very low scores and most of your comments are not identified as insulting. (It would be over 50% if it is actually an insulting comment.) So most people on HN are not actually haters. I just had a more "hater" focused design just for fun. There are in fact actual haters though, if you look hard enough.
5. Something I found interesting is clicking the "Back In The Day" checkbox. It takes your 50 oldest comments and analyses them, instead of your 50 most recent.
6. Finally, if you're not sure why some comments are getting ranked higher than others, feel free to look at the training data I used (it's from a kaggle competition from a while back.) and read my blog post. If you don't want to here are additional features I used on top of standard bag-of-words (CountVectorizer):
* badwords_count – A count of bad words used in each comment.
* n_words – A count of words used in each comment.
* allcaps – A count of capital letters in each comment.
* allcaps_ratio – A count of capital letters in each comment / the total words used in each comment.
* bad_ratio – A count of bad words used in each comment / the total words used in each comment.
* exclamation – A count of "!" used in each comment.
* addressing – A count of "@" symbols used in each comment.
* spaces – A count of spaces used in each comment.
If you have suggestions on other features I could collect let me know! I'll also be building a way to get actual training data from HN itself and letting HN users determine if a comment is actually insulting or not so that the predictions constantly improve.
This kind of mindset ("Let's check I have the least negative impact on the community") has to come from the HN crowd. Would it be relevant to adapt it to Reddit?
I'd say that the fact that it doesn't even work - arbitrarily doling out 'hater' marks due to a poor sentiment engine - makes it a good candidate for removal.
This insidious flaw results in two things:
1. accounts being deemed as 'haters' without warrant
2. the belief that the tool is 'correct' - perpetuating #1.
42 comments
[ 3.5 ms ] story [ 87.9 ms ] threadI would be more interested in measuring how often a person up/down votes, how often they are up/down voted compared to others within a similar posting. It is one thing to get down voted into oblivion but if you have a lot of company does that mitigate it?
"These things make me genuinely happy." -sandmanxc
I'm a horrible person.
Apparently I'm a saint, or I don't comment that much.
I want to make the world a better place.
I guess the algos need more training, which I'm helpfully providing with this comment.
https://github.com/kevinmcalear/hater_news
I like the idea of scoring people on their "troll" number. This might fix other conversation sites where there's lots of noise. Someone could sell the solution to the Wall Street Journal.
"https://news.ycombinator.com/item?id=8497356*"
How does this thing work?
I'm sure I've said stuff on HN which is more negative than that.
"Ah ok. I looked at it more as something you force on your co-workers :)"
Looks like 'ah' implies negativity to the algorithms driving this.
Edit: It's also only pulling your 50 most recent comments by default. That covers about the last two weeks for me and is probably why it isn't very accurate. 50 comments is a very small dataset to work from. The other option it gives is 50 oldest comments - again not very representative for someone who has been commenting regularly for 1301 days. I try to keep my comments useful but I know for a fact I've gotten into emotional debates in the last 4 years in which I was negative/stupid/hating. Maybe it's not possible but I would suggest pulling a random sample of 100-200 comments for more accurate data.
*"Fitting all that into 7 minutes will likely lead to injury for someone who is in the audience for any kind of short N minute workout, IMO. I concur with your general sentiment and good intentions though. :)"
I wrote this a couple days ago though, and am a bit miffed because I would have expected to have written a more acerbic comment sometime in the past.
(maybe now?)
Works for me. https://transfer.sh/6y7eK/riverbashing.png
https://news.ycombinator.com/item?id=7960514
which I can understand. Including that one I have a hater score of 3.5% which makes me a pretty positive person :)
"No way you're a failure, that's all I'll add." -BasDirks
It's pretty obvious how it was misinterpreted.
"With Wine everything is fine" - izietto
:(
"Have you ever tried the 11" Air?"
It certainly wasn't intended as an insult at the time, but I do like the way it now appears to be one. "I hate these Mac casuals..."
Application Error An error occurred in the application and your page could not be served. Please try again in a few moments.
If you are the application owner, check your logs for details.
A few things:
1. Thanks for posting my blog post (https://news.ycombinator.com/item?id=8517727) @chippy. :) The actual app ( haternews.co ) kept getting booted off HN... And now thanks @melling for posting it.
2. There have been a lot of interesting comments on the three (now four) threads on here. People pointed out some bugs and overall issues which I will be fixing (also, the site should not crash half as much now). This is just a fun side project I have been messing around with so I can get better at using data science in various applications. If you would like to help build it out for fun further let me know! Also, feel free to submit a bug or suggestion for an improvement if you really want to.(https://github.com/kevinmcalear/hater_news/issues)
3. I wanted to build the "hater score" for two reasons. First, to see how accurately I could build a model to measure insulting comments in the wild and second (if it's accurate), to see how people would react to seeing how positive or negative they usually are on Hacker news (or other social networks).
4. I wanted to make sure everyone knows that just because something is your "Worst Comment" doesn't mean it is negative. Most people have very low scores and most of your comments are not identified as insulting. (It would be over 50% if it is actually an insulting comment.) So most people on HN are not actually haters. I just had a more "hater" focused design just for fun. There are in fact actual haters though, if you look hard enough.
5. Something I found interesting is clicking the "Back In The Day" checkbox. It takes your 50 oldest comments and analyses them, instead of your 50 most recent.
6. Finally, if you're not sure why some comments are getting ranked higher than others, feel free to look at the training data I used (it's from a kaggle competition from a while back.) and read my blog post. If you don't want to here are additional features I used on top of standard bag-of-words (CountVectorizer):
* badwords_count – A count of bad words used in each comment.
* n_words – A count of words used in each comment.
* allcaps – A count of capital letters in each comment.
* allcaps_ratio – A count of capital letters in each comment / the total words used in each comment.
* bad_ratio – A count of bad words used in each comment / the total words used in each comment.
* exclamation – A count of "!" used in each comment.
* addressing – A count of "@" symbols used in each comment.
* spaces – A count of spaces used in each comment.
If you have suggestions on other features I could collect let me know! I'll also be building a way to get actual training data from HN itself and letting HN users determine if a comment is actually insulting or not so that the predictions constantly improve.
This insidious flaw results in two things:
1. accounts being deemed as 'haters' without warrant
2. the belief that the tool is 'correct' - perpetuating #1.
> "Excellent! thankyou"
Talk about making me feel good about myself.