Feedback request: Project to show bias in wikipedia articles

[justinbrainwave]

One more question I have is, if you had to calculate a "bias score" for a user in an automated way, how would you do it?

This would be a difficult question but I can share some general pointers:

a) if they focus in a very narrow subject area
b) and get into editwars and are reverted frequently
c) or, are obviously are "protected" by certain friendly administrators (the Guerrilla Skeptics and the Doc James "Medical Mafia" being two major examples, who are still active today).

Subject areas that habitually attract abuse:
--political figures, especially conservatives (watch for aggressive editwarring)
--political concepts that attract a LOT of sectarian rage, like the British Isles/Ireland or Israel-Palestine issues
--people with "outside" medical or social ideas
--anything to do with Scientology/Dianetics, Transcendental Meditation, or extremist religious cults
--anything that the notorious Chip Berlet is interested in controlling, he's basically a leftist conspiracy crank
--celebrities in the midst of a personal scandal
--Follow that epic asshole David Gerard around to see the worst of Wiki-Cranks....
--same for Guy "JzG" Chapman

There are other items but they tend to be more obscure or much lesser in severity or general interest.

This is useful, thanks!

[WWHP]

Let me k now if you think this would work and let me know if you have any more questions.

My case study is here https://rome-viharo.medium.com/aiki-wik ... dc707bdd22

Hi! I read the article but I found the amount of links a bit overwhelming, I am interested on the aiki wiki algorithm, have you written about it or do you have have it as a proprietary product? Any algorithm to be able to find "bad faith actors" would be helpful.

I can pinpoint where in the case study would apply if you had a specific question, it is an overwhelming amount of links and work, even makes me want to throw up at times

Hit me up for a Aiki Wiki walkthrough of the algorithm. Here is an article on the methodology and overview, but it is one of those things that you have to see to fully grok.

https://rome-viharo.medium.com/an-intro ... 06eb04f722


The whole system already can identify all types of actors, from good to bad, but I am curious about your work too. can you reach out on twitter?

[justinbrainwave]

Let me k now if you think this would work and let me know if you have any more questions.

My case study is here https://rome-viharo.medium.com/aiki-wik ... dc707bdd22

Hi! I read the article but I found the amount of links a bit overwhelming, I am interested on the aiki wiki algorithm, have you written about it or do you have have it as a proprietary product? Any algorithm to be able to find "bad faith actors" would be helpful.

I can pinpoint where in the case study would apply if you had a specific question, it is an overwhelming amount of links and work, even makes me want to throw up at times

Hit me up for a Aiki Wiki walkthrough of the algorithm. Here is an article on the methodology and overview, but it is one of those things that you have to see to fully grok.

https://rome-viharo.medium.com/an-intro ... 06eb04f722


The whole system already can identify all types of actors, from good to bad, but I am curious about your work too. can you reach out on twitter?

will do thanks!

[Jake Is A Sellout]

I found it a bit weird that their study didn't focus on an obvious source of bias, which is removal of content. You can have the most neutral stance on someone while only adding negative details about them while removing any positive details.

The effect of such removals is still detected in their methodology though.

That paper already proves what you're trying to reflect. Wikipedia is biased, certainly when compared to a real encyclopedia, and that's down to two things they're never going to be able to fix - too few editors, and too many articles.

You're wasting your time engaging with Wikipediocracy seeking help or advice, they're nothing but Wikipedia enablers and cheerleaders. They are even less interested in hearing about studies like this than the actual Wikipedia editors, largely because the only people they think are to blame for anything that is wrong with Wikipedia, are the owners and staff. Who of course, have nothing to do with editing.

You're wasting your time with this place too. Eric Barbour has no time and even less intellect for this work. He's only going to send you down blind alleys.

For example, a classic example (perhaps the canonical one) of a biased Wikipedia editor these days, is the one called Snoogansnoogams. Every edit he makes, is to add positive material about the things and people he likes, and add negative material about those he does not. Similarly, he will remove both positive material and negative material where its presence negates his efforts. A programme that presented a pre and post Snoog Wikipedia article side by side would certainly be a helpful tool, but it's not like it's hard to do it manually.

Contrary to Eric's out of date and probably half remembered knowledge, the Snoog roams the whole encyclopedia, or at least the entire US politics area, he rarely gets into trouble or edit wars, and he isn't protected by a small clique, the entire community protects him, because in their eyes, he is doing nothing wrong.

This is Wikipedia, 2021 style. In the beginning, you weren't supposed to be able to tell what an editor's personal views were from their edits. That would be a sign they were biased, and Administrators were supposed to advise, warn and ultimately ban such people if they couldn't change their ways and be more like a professional editor (who is obviously still biased, but is trained to be aware of it and mitigate it).

As Larry Sanger famously said, things went very pear shaped with Wikipedia quite early on, certainly once direct control from the top was lost, and the lunatics are now fully in control of the asylum (i.e., some of the most biased people on there, are the actual Administrators who are supposed to get rid of biased editors). That accounts for the active efforts to bias Wikipedia content, while the manifestation of the passive element, the background noise of a clear Democratic lean, is well explained by the Barack Obama situation.

Don't listen to Eric. He'll have you wasting your time. He asks you to follow Guy Chapman for example. A serious Wikipedia critic would know Guy Chapman hasn't been seen on Wikipedia since May, and it seems highly likely it was because he was targeted precisely because he got a little too big for his boots, and when it came to the already broad liberty on Wikipedia for Administrators to be biased and abuse their position, he was given an inch and he took a mile. He's dead in a ditch somewhere, either murdered or by suicide.