School paper plagiarism

You can talk about anything related to Wikipedia criticism here.
Post Reply
User avatar
ericbarbour
Sucks Admin
Posts: 4601
Joined: Sat Feb 25, 2017 1:56 am
Location: The ass-tral plane
Has thanked: 1142 times
Been thanked: 1844 times

School paper plagiarism

Post by ericbarbour » Wed Nov 07, 2018 8:26 pm

One of Wikipedia's most popular applications with school children.

As I've noted for many years, there is a website that checks school essays for plagiarism called Turnitin. And exactly seven years ago, Turnitin ran a study of the uses their service. And discovered that Wikipedia was the #1 source for plagiarized text with both secondary and college students. The original report appears to have been removed from Turnitin's website. For what reason, dunno. There is a copy here.

And I also discovered, Wikipedia insiders absolutely hate Turnitin. So much, that:
Their Wikipedia article was first created in 2005. And tremendously expanded from January 2006 until 2009 by a number of accounts: Cutter20 (T-C-F-R-B) who did little on Wikipedia other than the Turnitin article, plus working journalist and administrator Daniel Case (T-C-F-R-B) (who was clearly working with Cutter20). Thence it was edited by a litany of IP addresses. It swiftly expanded to more than 15k bytes by late 2007, most of it highly critical of the company's Turnitin plagiarism service. Every possible media source of a negative nature was dug up and added.

Later editors included obvious SPA Wikibarber (T-C-F-R-B), an IP address in Pakistan with possible links to a "diploma mill", a likely Turnitin paid editor called "Plagiarismdetection", and a continuing mishmash of other IP addresses and editing addicts. In 2008 a higher-education professional named ElKevbo (T-C-F-R-B) tried to make the article more neutral; his changes were later reverted. From 2009 to the present day, patroller Tedickey (T-C-F-R-B) has been watching the article with great care and quickly reverting any changes. A brief editwar in May/June 2015 concerned the section "Copyright infringement in countries where fair use does not exist", which was almost entirely about Italian law in this area; it was repeatedly removed by Tedickey.

Since Turnitin has repeatedly pointed at Wikipedia as the most popular source of school-paper plagiarized material, this article appears to act as a "revenge" attempt. Daniel Case's deep involvement in the early history of the article does not bode well for its "neutrality".

Evidently insiders were aware they had a problem. As a result of iParadigms trying to "cooperate" with the Wikipedia "community", in March 2012 a special policy page was created on Wikipedia, WP:Turnitin. Ironically, it was created by administrator and Wikiproject Medicine principal Jake Orlowitz. He received some support from fellow medicine editor Dr. James Heilman, but as of 2015 there is little evidence any serious action was taken by the insider community or the WMF. The timeline ends in August 2012, with subsequent proposed actions left "pending".


Turnitin later "teamed up" with the Wikimedia Foundation....just one of those little things you guys missed that Ocaasi and Heilman did to "make Wikipedia a happy place". And to neutralize their public-relations problems. Teachers also use WP for checking facts; they shouldn't, but they do, because the damn thing is sitting in front of them and difficult to avoid now. Everyone Googles, and Google points them to Wikipedia content.
In their words, their students increasingly “equate research with Googling,” and use search engines in lieu of more traditional sources without sufficient ability to judge the quality of information they find online.
In a similar vein, AP and NWP teachers use the online encyclopedia tool Wikipedia at much higher rates than U.S. adult internet users as a whole (87% vs. 53%). Wikipedia relies on user-generated, crowd-sourced content, a process that sometimes calls into question the accuracy of its information. In focus groups with teachers and students prior to the survey, Wikipedia was often noted as a tool teachers discourage or bar students from using because of concerns about the reliability of its content.


Found on the front page of Reddit today:
https://www.reddit.com/r/UnethicalLifeP ... translate/
Top voted comment:
Turnitin is the real question

User avatar
Dysklyver
Sucks Critic
Posts: 391
Joined: Thu Jun 21, 2018 10:14 am
Has thanked: 8 times
Been thanked: 24 times

Re: School paper plagiarism

Post by Dysklyver » Fri Nov 09, 2018 4:39 pm

A little known fact but every single edit made to Wikipedia is automatically run through Turnitin by bots built into the mediawiki software. Offending edits are flagged on a toollabs portal (copypatrol) and as of a few weeks ago new articles which contain copyvios are flagged automatically to reviewers.

User avatar
ericbarbour
Sucks Admin
Posts: 4601
Joined: Sat Feb 25, 2017 1:56 am
Location: The ass-tral plane
Has thanked: 1142 times
Been thanked: 1844 times

Re: School paper plagiarism

Post by ericbarbour » Fri Nov 09, 2018 9:22 pm

Dysklyver wrote:A little known fact but every single edit made to Wikipedia is automatically run through Turnitin by bots built into the mediawiki software. Offending edits are flagged on a toollabs portal (copypatrol) and as of a few weeks ago new articles which contain copyvios are flagged automatically to reviewers.

Links please? All I see is this:
https://meta.wikimedia.org/wiki/CopyPatrol

User avatar
Dysklyver
Sucks Critic
Posts: 391
Joined: Thu Jun 21, 2018 10:14 am
Has thanked: 8 times
Been thanked: 24 times

Re: School paper plagiarism

Post by Dysklyver » Fri Nov 09, 2018 10:33 pm

ericbarbour wrote:
Dysklyver wrote:A little known fact but every single edit made to Wikipedia is automatically run through Turnitin by bots built into the mediawiki software. Offending edits are flagged on a toollabs portal (copypatrol) and as of a few weeks ago new articles which contain copyvios are flagged automatically to reviewers.

Links please? All I see is this:
https://meta.wikimedia.org/wiki/CopyPatrol


There is masses of links, I will try and give an overview here.

In https://phabricator.wikimedia.org/T179227 we see that both Copypatrol and Eranbot have direct database access to Wikipedia (after consultation with the WMF and a tech meeting, see https://phabricator.wikimedia.org/T127390), and are hosted on toollabs. While originally separate, copypatrol and Eranbot effectively merged in 2016 after https://phabricator.wikimedia.org/T137691. The interface everyone uses now was made then too, in the 50+ task https://phabricator.wikimedia.org/T131583.

https://en.wikipedia.org/wiki/User:EranBot is an important part of this system, but is not a bot as such, having been gradually integrated into Wikipedia itself. See also https://en.wikipedia.org/wiki/Wikipedia:Bots/Requests_for_approval/EranBot_3. It even has OAuth since https://phabricator.wikimedia.org/T132081.

The integration with Special:NewPagesFeed is part of https://phabricator.wikimedia.org/T193782 or specifically https://phabricator.wikimedia.org/T199359 and it's subtasks. A custom user-right and API module was created for Eranbot to integrate as detailed in https://phabricator.wikimedia.org/T202041.

Since https://phabricator.wikimedia.org/T110144 turnitin is also used in Earwig (https://tools.wmflabs.org/copyvios/). In https://phabricator.wikimedia.org/T198361 and https://phabricator.wikimedia.org/T141380 they considered also using Google's API but eventually decided against it. You will note iThenticate is used (see https://phabricator.wikimedia.org/T201071).

As it is a main tool they did eventually run out of credit, Turnitin gave them a million to start off with, and https://phabricator.wikimedia.org/T185163 when they ran out ended with a provisional supply.

They wanted to tell everyone about it, but their press release (https://phabricator.wikimedia.org/T185292) got stuck in the epic backlog (literally called that) and then forgotten about and shuffled around not being done.

User avatar
ericbarbour
Sucks Admin
Posts: 4601
Joined: Sat Feb 25, 2017 1:56 am
Location: The ass-tral plane
Has thanked: 1142 times
Been thanked: 1844 times

Re: School paper plagiarism

Post by ericbarbour » Fri Nov 09, 2018 10:49 pm

Thanks, saving it. No one ever talks about Eran and his bots. An Israeli Jew and prominent administrator on Hebrew Wikipedia:
With only 135,000 articles in 2013 and 200,000 as of 2017, the Hebrew Wikipedia is either very well-run and well-disciplined, and/or hopelessly bigoted against Arabs and Palestinians, depending on who is being asked.

Quote: "You might like the Hebrew Wikipedia. They have a "Parliament" with policy votes decided by a 60 or 65% majority, I don't remember which. AFDs are also essentially votes with a separate section for discussion if needed. Some of the policies on Hebrew Wikipedia are annoying, but at least they have a functioning system for making policy, and I dare say it's a happier place in general than English Wikipedia."
Quote: "Most minor-language Wikipedias are chock full of polemic screeds, making the English Wikipedia look like a paragon of neutrality and editorial restraint. The unrestrained nationalism of several of the eastern European language wikis is especially noticeable, and the Hebrew wiki's unabashed pro-Zionist, anti-Palestine politics are equally well-known. These political biases invariably affect decisions on whether or not a given source is "reliable" enough to "verify" something."
Quote: "The problem with democracy on contents is that it can be used to enforce cultural biases. Just think how the Hebrew, English, Russian and Arabic Wikipedias are likely to be positioned on the spectrum of opinions on the Israel-Palestine conflict."
Quote: "The Hebrew Wikipedia is hopelessly biased in its coverage of Israel and all matters related to Islam (surprise, that) and has been known to enforce opinion orthodoxy on its editors as well. "
Quote: "If you don't think it's "that bad", here's a little illustration. Go to Google Translate, put in "he.wikipedia.org", and translate from Hebrew to English. Then put the word "פלסטינה" (Palestine) in the search box and hit return, and look at some of the resulting articles. Items like "Arab conquest of the Land of Israel" or "War of Independence" should give you some idea. Plus articles about British people involved in the Arab revolt against the Ottomans, like Stewart Francis Newcombe or T.E. Lawrence, tend to be very negative. And the one about Yassir Arafat (יאסר ערפאת), hoo boy."
One of the few mentions of Hebrew Wikipedia in English-language media: "Israeli newspaper Haaretz has published an article about the Hebrew Wikipedia’s deletion discussions that it considers elite enough to put behind its premium content paywall, although it is still available though search engines. At issue was whether Israeli Prime Minister Netanyahu’s 25-year-old son Yair Netanyahu should have an article on the Hebrew Wikipedia. In the end, the younger Netanyahu got an article, but only at the cost of undeleting an article for another prime minister’s son, Shaul Olmert, the son of Ehud Olmert. In a seemingly unrelated discussion reminiscent of the Pokeman character argument for inclusionism, the article also points out that individual characters for the Game of Thrones have their own articles on Hebrew Wikipedia, as does the former caretaker of Netanyahu’s residence, who has lately been leading protests. The entry for the Likud party’s liberal caucus has likewise been undeleted, after a resurgence of the New Likudniks, however an article for an ultra-Orthodox Hasidic blogger failed to pass the notability discussion. There was also a nearly unanimous decision that the word kef (fun) should no longer redirect to hana’ah (enjoyment), but deserves a separate article."

Post Reply