Slashdot Mirror
34 'Highly Toxic Users' Wrote 9% of the Personal Attacks On Wikipedia (bleepingcomputer.com)
Researchers used machine learning to analyze every single comment left on Wikipedia in 2015. An anonymous reader shares their results:
34 "highly toxic users" were responsible for 9% of all the personal attacks in the comments on Wikipedia, according to a research team from Alphabet's Jigsaw and the Wikimedia Foundation. They concluded that "significant progress could be made by moderating a relatively small number of frequent attackers." But at the same time, in Wikipedia's comments "less than half of attacks come from users with little prior participation; and perhaps surprisingly, approximately 30% of attacks come from registered users with over a 100 contributions. These results suggest the problems associated with personal attacks do not have an easy solution... the majority of personal attacks on Wikipedia are not the result of a few malicious users, nor primarily the consequence of allowing anonymous contributions."
The researchers "developed a machine learning algorithm that was able to identify and distinguish different forms of online abuse and personal attacks," reports Bleeping Computer, adding that the team "hopes that Wikipedia uses their study to build a comments monitoring dashboard that could track down hotspots of abusive personal attacks and help moderators ban or block toxic users." The paper describes it as a method "that combines crowdsourcing and machine learning to analyze personal attacks at scale."
The researchers "developed a machine learning algorithm that was able to identify and distinguish different forms of online abuse and personal attacks," reports Bleeping Computer, adding that the team "hopes that Wikipedia uses their study to build a comments monitoring dashboard that could track down hotspots of abusive personal attacks and help moderators ban or block toxic users." The paper describes it as a method "that combines crowdsourcing and machine learning to analyze personal attacks at scale."
In further news, it was discovered that all 34 of the "toxic users" were Administrators or Wikipedia employees.
The entrenched fiefdoms are 100000%+ more harmful than the random drive-by. The drive-by will be deleted while the entrenched (college professors with beards, etc.) will be considered ***absolute truth***.
Do you mean that tab on the articles where all the aspies spend hours bickering about whether sentence 4 of paragraph 32 should have an oxford comma? Because if the people there in charge of distributing the personal abuse are overworked I could probably volunteer one or two hours a week.
were in regard to overly territorial Wikipedia moderators?
Build a Man a Fire, and He'll Be Warm for a Day. Set a Man on Fire, and He'll Be Warm for the Rest of His Life.
With 9 comments here already, I see the problem being blamed on:
- Freedom of speech
- Admins
- Muslims
- Liberals
No surprise that they're all ACs.
I'm sure it would be very upsetting for the ACs if /. started tracking IPs, but I suspect that a disproportionate number of "Trolls" come from the same IPs.
Mimetics Inc. Twitter
On behalf of myself and my 33 sock-accounts.
Sorry.
"That's the way to do it" - Punch
I wonder how these guys would behave in real life. What kind of mind do you have, when you're one of the most productive in the area of spewing hate, anger and vitriol.
8 of 13 people found this answer helpful. Did you?
I'm not really sure this is a problem for Wikipedia, but the ABC guys seem to think so. But take a look at their methodology. "Crowdsourced" "Machine Learning" via proprietary website, after we removed "common comments" which they assume to be bots. I'm sure anyone using the same data set would be hard pressed to recreate their results. They are very fuzzy despite all the algorithmic pruning.
We use this data to train a machine learning classifier, experimenting with features and labeling methods
Isn't this what they're really testing? An unspecified machine learning with "features" and "labeling"? Absolutely bewildering.
The entrenched fiefdoms are 100000%+ more harmful than the random drive-by. The drive-by will be deleted while the entrenched (college professors with beards, etc.) will be considered ***absolute truth***.
The entrenched fiefdoms, pages where one user (or a small cabal of users) believe that they own the article and will dispute and revert every change to their perfect prose are indeed a problem in Wikipedia-- their motto should be "the encyclopedia everybody can edit, except don't bother trying with these articles." But in my experience it's rarely college professors-- it's dedicated amateurs who have simply decided that they are the world's expert in this field.
Many of them actually are quite knowledgable-- there are some pretty good articles there. But sometimes these are by people who just don't have a good grasp on writing for clarity and sticking to the topic.
Most of the college professors I know are at best amused by wikipedia, and in general disdain it.
http://www.geoffreylandis.com
The best way to troll Wikipedia is to insert [citation needed] next to all the most obvious parts of the article.
I was going to do this to the Wikipedia article about "the sky", next to the part where it says "the sky is blue", but there are already 4 separate citations for that particular fact. There is a weakness in the following sentence, however - "At night, the sky appears to be a mostly dark surface". There is no citation for this so-called "fact" - until I have read a newpaper article or academic paper confirming that the sky is dark, I will go on believing the opposite. Because that's Wikipedia law.
Here you go. Add these to the article:
At night, the sky appears to be a mostly dark surface.[1][2][3][4]
[1] Harrison, E. R. "The dark night-sky riddle: a" paradox" that resisted solution." Science 226, (1984): 941-946.
[2] Jaki, Stanley L., and H. L. Armstrong. "The paradox of Olbers' paradox." American Journal of Physics, 40.9 (1972): 1354-1355.
[3] Harrison, E. R. "Olbers' paradox." Nature 204, (1964): 271-272.
[4] Wesson, Paul S., K. Valle, and R. Stabell. "The extragalactic background light and a definitive resolution of Olbers's paradox." The Astrophysical Journal 317, (1987): 601-606.
http://www.geoffreylandis.com
you want to verbuse him? chillax, there's no need for that.
sag
Why do people think that having a recognizable user name makes them right? You do realize that is just an ad homonym attack, right?
1. It's hominem, as in person.
Did you realize that you just made an ad homonym attack?
You attacked the AC post for using a homonym.
http://www.geoffreylandis.com
Having a recognizable user name doesn't automatically make someone right. But having the ability to go back and view their comments in prior conversations sure makes it easier to gauge if their opinion is worth a shit or not.
Yep. And after a while you get to notice that some usernames are usually very insightful.
There are only 2 uses I've seen for AC: Trolls, and people who claim they can't comment under their name because their employer would recognize them (or some flavor of that).
Yes on 1, no on two: these people could simply chose a username like "haX0r42" or "Pringleeater" that their boss won't recognize. /., there's one more reason a person might comment as AC: they have already moderated the thread and don't want to remove their moderations..
You're right, though, the worst of the drive-by flaming and pugnacious idiocy is almost always anonymous.
Due to the particular nature of
http://www.geoffreylandis.com
Leave Toxicity to the chemistry and use real words to describe what you want to accuse people of. Terms like toxic behaviour and hate speech are cool, because there is no clear definition, which means you can redefine them each time you use them. If somebody refutes a claim, you tell them, that the word was used in another context than what he refuted.
This news supports the frequent observation that the Internet and associated communications platforms give a very disproportionate voice to a very small minority of jackasses who seem to have nothing better to do.
This posting is provided 'AS IS' without warranty of any kind, implied or otherwise.
Poe's law in action folks.
Unicode killed the ASCII-art *
Poe's law in action folks.
A useless meme based on anecdotes.
If you want news from today, you have to come back tomorrow.
I can understand where you are coming from, but to give you some insight on how odd this situation gets: First of all, there are people who genuinely believe this sort of thing. Second, it's not _entirely_ without merit, since once or twice even the co-founder of WP, the great Jimbo Wales has gotten out of line and behaved inappropriately; although it's always related to boring administrative stuff like arbitration committee cases, and when it has happened he was either taken to task or noticed his mistaken and taken himself out of the matter. Third, there at least used to be a group of submitters from an anti-wikipedia activism site (yeah, that's a thing that exists) that would spam submissions to Slashdot. Many of them were ridiculously biased or entirely without merit, and more than one of those bad ones managed to get posted.
Did you reference the correction to a verifiable source?
If so, correct twice, then at next revert dispute the reversal. Is their source disproving yours? If it isn't, then the article at worst should contain both versions, with an entry that shows it is disputed which one is correct. If your source disproves their, they are asking for a ban.
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
although... I must supply that there's a problem with Wikipedia's acceptable sources. In particular, press articles are regularly accepted - even as sources in articles about issues of journalist integrity. So if you have 'people vs journalists', all the 'defensive' voice of journalists is being heard, sourced to their own articles, while people's proofs these articles are BS - usually posted in social media - is considered 'invalid' - 'original research' or otherwise unverifiable.
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
You mean the admins who will revert an article when someone adds to it, citing vandalism or not enough citations... but they will copy the same text -verbatim- to the article and have that stand?
Nobody bothers editing Wikipedia anymore... if you are not in the "A"-list crowd, you will just get your changes reverted on you, no matter how good you are.
Trump is a bad example for you point.
Since Trump has a reputation of being a liar it is very easy for people to see that the source is Trump and disregard it as a fabrication.
If what he said happens to be true it is probably just a coincidence.
The real problem occurs when you have someone with a reputation of being right but happens to be wrong in a single case.
If people have the habit of looking at the source they might assume that this false statement is true just because it came form a source with a long history of providing reliable facts.
The reverse can also be the case. There is the story about the boy who cried wolf.
Because of his history of telling lies other people wouldn't believe him when he told the truth since they considered the source and disregarded him because of past statements.
Looking at the source isn't a very accurate test to determine if something is true or not.
The upside of looking at sources is that it is very low effort verification, other methods to test a statement could be very time consuming.
In the case of Trump you could probably just disregard his statements because they are logically inconsistent or otherwise nonsensical.
Are you trolling, or are you another person who doesn't understand the difference between Freedom of speech and the freedom of owners to moderate a private platform of their own, such as Wikipedia however they please?
Yeah, it can be a pain to fight those situations, but there are at least mechanisms to put a stop to that for anyone who cares to put forth that sort of effort.
Have gnu, will travel.
If the source is provably wrong, and you provide the proof (e.g. the source that has that author's admission) then that's an entirely bullshit argumentation. Provide proof of wrongness, get two first steps of edit war (revert their revert, each time citing the proof), then dispute their revert if they do this again.
Note, just deleting a section will likely get you nowhere. It's hard to source absence of text. Instead edit it, with the right followup. "It was believed that 2+2=5[source], although later studies[source] disproved that claim."
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
Interesting. I was thinking more of that small minority of rude and obsessed people who seem to take over every conversation, start flame wars everywhere, send death threats to everyone, etc. much like spam, it doesn't take that long to generate a lot of noise if you want to take the time. So it can appear there are a lot of jerks because they make so much noise, even though only a very small percentage of the audience.
This posting is provided 'AS IS' without warranty of any kind, implied or otherwise.
It shows that a small group of people can make a difference.
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
It was clearly not intended to be a truthful statement. No Slashdot reader could genuinely believe it to be true. How, then, can it constitute an attack?
A statement has to be true for it to be an attack?
All I know is if I was a Wikipedia editor, and was honestly trying to do the best job I could, which I believe most of them actually are, I'd probably feel kindof shitty. It's snark criticizing volunteer work without knowing the person or what they actually do. It doesn't help the discussion at all, it's part of the general trend of casually insulting people so we can look superior, be funny, or just generally feel better about ourselves. It's mild as far as these things go, so I don't know that I would have bothered using a mod point on it, but it falls into that category.
And your father smelt of elderberries.
The Russians have won. They have made the world a cesspool of distrust, greed, fear and hate.
Keeps diluting the elitism of the internet.
I wonder if this isn't true on more general forums.. especially political forums like TheHill. Some posters have over 100000 posts and those generally are the obnoxious ones.
Post in social media containing detailed, referenced research? Or video taken at given events? Or recording of Q&A with the 'subject' of the article? Social media is just location, not content. The content itself can be created in such a way that it's a fully reliable and verifiable research - but it's [original research], not acceptable as Wikipedia source. Meanwhile, entirely unverified editorial pieces that were published by a newspaper are considered valid sources, no matter how many direct witnesses provide videos proving the contrary on their Youtube, Twitter and Tumblr accounts.
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
"the general trend of casually insulting people" - yep, I hate this too.
If I paraded through the town singing "Rakarra is the secret ruler of Mars and he is trying to kill us all with his mind control waves", would you really consider that an attack on yourself?
First, I might be quite intrigued by it, and I would certainly hope I could live up to the reputation. What an honor, what a responsibility!
But really, it would feel like satire, or just playful banter. I wouldn't take it seriously, no. But it was discovered that all 34 of the "toxic users" were Administrators or Wikipedia employees does not sound like playful banter to me, it feels a bit more like a stinging rebuke. Reading that, it would be hard for me to come to another conclusion than you really do think that Wikipedia moderators and administrators are bad, toxic people. It's a sentiment I've seen here more than once, echoed sometimes in summaries of actual Slashdot articles.
I've edited Wikipedia a number of times, and most of my contributions were insta-reverted. Until the culture of "n00b-bashing" is eradicated from wikipedia, they could use a little satirical scrutiny.
I think the culture could certainly come under scrutiny.