User talk:Headbomb/unreliable

gen-ph

Headbomb, do you think it would be possible for your script to detect if an arXiv pre-print is in gen-ph? They do this wonderful service of putting the crackpot pre-prints in a separate section, it would be nice to take advantage of their work and mark the links accordingly. The difficulty is that you can't know that from the URL: the "new style" (e.g. https://arxiv.org/abs/2104.05395 ) doesn't show the section at all, and the "old style" (e.g. https://arxiv.org/abs/physics/0609007 ) only shows the larger section "Physics", which has a lot of legitimate research. Your script would need to query the page and check whether the subsection is "physics.gen-ph", I don't know if this is technically possible. Tercer (talk) 20:45, 15 April 2021 (UTC)

With a bit of tweaking, it could probably detect {{cite arxiv}}s with |class= set to physics.gen-ph or similar, but I'm not super keen on distinguishing within the archive (even if physics.gen-ph is more likely to be bunk nutcase ramblings than hep-ex). The arxiv isn't a reliable source in general, so whenever you see it you should double check if it's used for routine information, or for novel claims. But let it simmer in the brainjuice for a while, see if I can't come up with something. Headbomb {t · c · p · b} 20:53, 15 April 2021 (UTC)

• Skilling, John; Knuth, Kevin H. (2020). "The arithmetic of uncertainty unifies quantum formalism and relativistic spacetime". arXiv:2104.05395 [physics.gen-ph].
• Chalmers, Gordon (2006). "Modification to Special Relativity". arXiv:physics/0609007.

@Tercer: I tried something, see above. Headbomb {t · c · p · b} 20:57, 15 April 2021 (UTC)

You can hunt for them here. Headbomb {t · c · p · b} 20:59, 15 April 2021 (UTC)

Yep, it does work, thanks! The |class= parameter is seldom set, though, so I guess the vast majority of the cases will be hard to find. Tercer (talk) 21:41, 15 April 2021 (UTC)

Well, for {{cite arxiv}}, citation bot does set it. Headbomb {t · c · p · b} 21:58, 15 April 2021 (UTC)

I don't know how active citation bot is, but for example in Gleason's theorem it is not set anywhere, and this is typically the case. Tercer (talk) 22:16, 15 April 2021 (UTC)

@Tercer: that's because those are {{cite journal}}s, which are things that have gone through peer-review (well, forgetting predatory/quack journals, but those should be flagged because of their predatoriness/quackness, not because of physics.gen-ph). I've done a rough check and all ~690 or so {{cite arxiv}} with newer links do have their |class= set. Headbomb {t · c · p · b} 22:18, 15 April 2021 (UTC)

Ah, I see. Well, there are also borderline journals which occasionally publish stuff from physics.gen-ph. If I recall correctly there was recently a case with Int. J. Theor. Phys., which is not flagged as quack. Tercer (talk) 22:23, 15 April 2021 (UTC)

@Tercer: BTW, there's about 12 citations to gen-ph remaining or so. So may be OK, but since you're interested in doing that cleanup, I figured I'd ping you on the progress. Headbomb {t · c · p · b} 16:49, 27 June 2021 (UTC)

In my first pass I removed the cases where it the citation was unambiguously unwarranted, the cases left are those where it should remain or I find it unclear. I'll take a close look at them tomorrow. Tercer (talk) 20:41, 28 June 2021 (UTC)

Ok, I finished it now. Removed some, fixed some, in the end I left 4 citations:

• Brian Josephson: only one item in a long list of unreliable references. No point in removing only this one, and I'm not in the mood for fixing the whole article.
• Steven Weinstein (philosopher): it's being cited as one of his works, I don't see a problem.
• Safety of high-energy particle collision experiments: a large part of the article is about the controversy caused by this paper, and it is specifically mentioned as being unreliable. I think it definitely should stay.
• Energy Catalyzer: again, the paper is important for the article, and it is specifically cited as being unreliable. There's also a debunking, also published in gen-ph. Not acceptable in ordinary circumstances, but given the subject matter it's as reliable a source as it gets. Tercer (talk) 14:12, 1 July 2021 (UTC)

Google books highlighted yellow

I see that google books urls are now highlighted as yellow. This is making it much harder for me to find the script useful since there are just so many gbooks URLs in articles I work on. Is there a way I could customize the script to stop highlighting books.google.com for me? No worries if not... Eddie891 ^Talk _Work 12:28, 28 May 2021 (UTC)

@SD0001: I've setup custom rules (for myself) at User:Headbomb/unreliable-rules.js, but it doesn't seem to overrule User:Headbomb/unreliable.js. Is there a way for User:Headbomb/unreliable-rules.js to not just supplement, but rather entirely overrule User:Headbomb/unreliable.js? Headbomb {t · c · p · b} 13:34, 28 May 2021 (UTC)

Remove return false; from lines 138 and 153. – SD0001 (talk) 14:11, 28 May 2021 (UTC)

@Eddie891: I'm exploring a solution (see above), but in the meantime remember that yellow simply is the least problematic. Google Books was added because it's a general repository of books, and will have all sorts of unreliable books, like self-published books (e.g. [1]). Headbomb {t · c · p · b} 13:39, 28 May 2021 (UTC)

Thanks for checking it out. Your point about gbooks is undoubtedly valid, but for me I'm well aware of the limitations of google books. The frequency with which articles include links there means the yellow highlighting won't help me identify potentially unreliable sources, it will just cause me to gloss over all yellow highlighted links and miss more useful highlights, such as to forbes Eddie891 ^Talk _Work 13:55, 28 May 2021 (UTC)

@Eddie891:, create User:Eddie891/unreliable-rules.js and paste

unreliableCustomRules = [
	{
		comment: 'Plain google books',
		regex: /\b(books\.google\.com)/i,
		css: { "background-color": "" }
	},
];

Should fix it. Headbomb {t · c · p · b} 18:18, 28 May 2021 (UTC)

Works like a charm, tyvm!! Eddie891 ^Talk _Work 01:35, 29 May 2021 (UTC)

Sportskeeda

Questionable-at-best source currently used in hundreds of articles, focussed on Indian sports but also on other topics. The editorial process sounds extremely lax -- they publish hundreds of articles a day, most by non-employees who are expected to submit already-factchecked articles. There've been a couple of short discussions at RSN, but nothing much has been done to start to remove this source or check the assertions it supports. I think it would be helpful if it appeared highlighted. —valereee (talk) 14:05, 29 May 2021 (UTC)

@Valereee: I'll add to "depends on contributor/topic" for now. But an RSN discussion would be best to determine how to classify it in general. Headbomb {t · c · p · b} 14:09, 29 May 2021 (UTC)

Sounds good! I was just thinking about opening a discussion at RSN. —valereee (talk) 14:12, 29 May 2021 (UTC)

Physics Essays

I don't know much about specific journals and figuring out if they are reliable seems to be pretty tricky. I see "Physics Essays" is included in the list but can't figure out why it has been included. Could you clarify Headbomb? I tried a couple of google searches and couldn't find much.

I think you should establish clear criteria for inclusion in the list if you want to have more users adopt this script. At the moment some of the inclusion criteria seem too opaque.

-- {{u|Gtoffoletto}}  ^talk 16:32, 30 May 2021 (UTC)

@Gtoffoletto: See User talk:Headbomb/unreliable/Archive 1#Physics Essays and WP:JCW/Questionable3#Physics Essays. You can also search the archives of WT:PHYS for several discussions related to this. Headbomb {t · c · p · b} 16:38, 30 May 2021 (UTC)

Thanks! Wow. Identifying reliable journals is a real mess... -- {{u|Gtoffoletto}}  ^talk 19:46, 30 May 2021 (UTC)

Weird result

The following citation "Conversion Therapy is Torture". International Rehabilitation Council for Torture Victims. Retrieved 31 May 2021. to the website of reputable organization International Rehabilitation Council for Torture Victims is marked as a predatory journal. I'm guessing it has a suspicious combination of letters in it? (t · c) buidhe 17:17, 31 May 2021 (UTC)

@Buidhe: Can't see why it would have been flagged. It's possible IJICRT got cut off to ICRT by accident. Either way, I've fixed it. Thanks for reporting the error. Headbomb {t · c · p · b} 22:29, 31 May 2021 (UTC)

Lulu.com

This is a self-publisher/on-demand, should probably be added to the list —valereee (talk) 14:33, 21 June 2021 (UTC)

@Valereee: It's already listed. Headbomb {t · c · p · b} 19:32, 21 June 2021 (UTC)

Huh...when I checked in my sandbox this morning, it didn't highlight, but now it is. Weird. Sorry! —valereee (talk) 19:46, 21 June 2021 (UTC)

@Valereee: If you mean User:Valereee/sandbox, it's not highlighted because there is no link to Lulu.com, instead it highlights the link to GoogleBooks and warns you that GoogleBooks hosts self-published material. Headbomb {t · c · p · b} 21:38, 21 June 2021 (UTC)

OOOOH. I'm an idiot. Hahahahaha... —valereee (talk) 21:41, 21 June 2021 (UTC)

Press TV Deprecated

A discussion surrounding the reliability of Press TV on the reliable sources noticeboard was recently closed with clear consensus to deprecate. Would it be possible for someone to add the source's four urls that are listed on WP:RSP to the list, in light of this? Currently, the Iranian TLD is not being highlighted at all, though the now-defunct .com TLD is showing only as generally unreliable. — Mikehawk10 (talk) 02:28, 25 June 2021 (UTC)

Added here. I don't know which other URLs you are talking about however. Headbomb {t · c · p · b} 22:11, 26 June 2021 (UTC)

@Headbomb: The station's alternative sites include the [presstv.com now-seized] presstv.com. The group also includes includes presstv.co.uk, and presstv.tv, according to their website. — Mikehawk10 (talk) 22:52, 26 June 2021 (UTC)

Added the last two. If you find more, let me know. Headbomb {t · c · p · b} 22:53, 26 June 2021 (UTC)

ABS-CBN websites highlighted yellow

ABS-CBN is considered a generally reliable source in the Philippines, and its reliability has not been contested in the perennial sources list and in the reliable source noticeboard archives, though Christian Broadcasting Network is also lumped into these results, so I assume this must be to do with the similarity of their urls? Akbermamps 10:35, 8 July 2021 (UTC)

Appears to be the case - the regex in question simplifies to \bcbn\.com, and the \b matches the dash in abs-cbn.com. GeneralNotability (talk) 02:44, 9 July 2021 (UTC)

I'll take care of the false positive by the end of the day. Going to be a bit busy for the next few hours however. Headbomb {t · c · p · b} 14:24, 9 July 2021 (UTC)

@Akbermamps: should be fixed now. Headbomb {t · c · p · b} 18:29, 9 July 2021 (UTC)

Surgical Neurology International is marked as marginal

Some DOIs of Surgical Neurology International are marked as marginal (e.g. 10.4103%2F2152-7806.63896 10.4103%2F2152-7806.103019) but links to the PMC of articles are not (e.g. PMC3514915 PMC2908352). I do not believe this source should be marked as marginal.

There does not exist an RSP or RSN post about this source. The marginal entry in this add-on says: "Borderline source, which often (but not always) fails higher sourcing requirements. For Bentham journals, only those published by Bentham Open are of concern. For Copernicus journals, those from the EGU are fine." SNI is not a part of Bentham or Copernicus.

It is owned by Wolters Kluwer Medknow [2]. Scimago puts SNI in Q3 for Neurology, but Q2 for Surgery [3]. It is ranked 11th of most cited neurosurgery journals by Google Scholar [4]. It is indexed in Pubmed Central [5] and Scopus [6]. The editor-in-chief is a well-respected neurosurgeon at UCLA, James I. Ausman.

There are, truthfully, not that many general neurosurgical journals. And it is a small field (of roughly 3700 practicing neurosurgeons in the US [7], compared to roughly 110,000 internal medicine physicians [8]). So what counts as a "good" journal in this field is skewed in comparison to the rest of medicine. All index factors in neurosurgery are pretty small, comparatively.

For all these reasons, I don't think this source should be listed as "marginal."

Full disclosure, I have also published in this journal: [9] [10].

Anyway, let me know what you think.--Shibbolethink ^{(♔ ♕)} 19:11, 25 July 2021 (UTC)

Medknow Publications (who did published that article, you can tell by the DOI prefix 10.4103) is borderline , yes. See the note at WP:CITEWATCH#Medknow Publications. Medknow was listed on Beall's list, it then got delisted, but several of its journals got individually re-listed. Doesn't mean SNI itself is problematic (especially after it broke ties with Medknow; it is now published by Scientific Scholar, which you can tell by the DOI prefix 10.25259), but Medknow in general is... to be taken with a grain of salt. Headbomb {t · c · p · b} 00:26, 26 July 2021 (UTC)

Headbomb, ohhhh, I see. So the coloration is based upon DOI prefix? That makes a ton of sense, thank you. So the publications that were put out while under Medknow are tainted, but the new DOIs will not be. Okay that makes sense, and is probably fair. I don't really have any intention of ever citing anything from there anyway, as there are better topic reviews in better journals. Thank you for the explanation.--Shibbolethink ^{(♔ ♕)} 00:37, 26 July 2021 (UTC)

It's based upon DOI prefixes and URL domains, yes. It could technically match any pattern inside of a URL, but those are usually the ones that make the most sense. Headbomb {t · c · p · b} 05:58, 26 July 2021 (UTC)

"Predatory" publishers

Journals from European Center for Science Education and Research (EUSER) are coming up red as "predatory journal or publisher" but it's not clear why. They appear on Beale's archived list and are carried over to the Kscien list. But they claim to have double-blind peer review and editorial boards, and I can't find anyone disputing that other than by a bare list entry. I don't see any RFC on Wikipedia, and EUSER doesn't appear in any of the linked lists on this template.

I think the problem is that if we are going to flag a publisher as predatory, we better have developed our own case. Otherwise this a potentially slanderous statement with careless disregard for the truth. If we are just copying someone else's list, that's a problem. More specifically, if we are using Beale's list, he only called it "Potential, possible, or probable predatory scholarly open-access journals" so by shortening this to "predatory journal or publisher" we are making an extra leap without looking. This should be changed to "potential predatory journal or publisher". If we are using kscien, they only call it "publishers with suspicious scientific peer review" which should only be flagged as "suspicious peer review".

I believe this is an error that needs to be fixed. Dhaluza (talk) 01:26, 16 August 2021 (UTC)

"claim to have double-blind peer review and editorial boards" See WP:MANDY, basically. Their website is clearly set up as a puffery piece, claiming prestige from being indexed in Google Scholar, WorldCat, DOI and other things which are basically automatic or trivial, or non-existent indices [like 'sicilit']. They're not even in the DOAJ. Headbomb {t · c · p · b} 02:58, 16 August 2021 (UTC)

There's also affiliated with Open Journal Systems, another predatory publisher. Headbomb {t · c · p · b} 03:06, 16 August 2021 (UTC)

So guilt by association? Who says they are affiliated, and who says Open Journal Systems is predatory? Mandy only applies after a credible allegation. Dhaluza (talk) 12:30, 16 August 2021 (UTC)

Yes, I realize the EUSER website may be an elaborate fraud. Or maybe not. The question is who, exactly, is saying they are a fraud? And how is an editor using your tool to decide if they don't know that?

I hope the irony is not lost that this tool is questioning sources without citing its sources.

We would not consider Beall's list a reliable source for calling a publisher predatory in an article because it's a SPS. It was also criticized for having nearly 20% false positive rate, especially with open source journals in non G7 countries, which leads to first world bias. And it's not clear to me that Kscien is any better. The latter claim to have a committee, but don't name them. You can't see why or when a publisher was put on the list. But I'm left guessing what sources you are using because they are not clearly cited in the tool, or in the docs. An editor should not have to spend hours hunting around to see why a source is flagged. Dhaluza (talk) 12:06, 16 August 2021 (UTC)

The preponderance of evidence, from Beall, to the fairly obvious scam/promotional puffery of their website and obvious pay-to-publish nature of their books, to hitting multiple red flags, a website riddled with typos as soon as you leave the main page, affiliations with other known predatory publishers, combined with the lack of indexing of their journals in reputable/selective services, means EUSER has several strikes against it, and little to nothing going for it. Headbomb {t · c · p · b} 12:41, 16 August 2021 (UTC)

I wasn't aware Beall published a preponderance of evidence, so that would be interesting. Please link to your source.

But that would also be beyond the point. This is not about whether EUSER is predatory or not, it's about why it is flagged as predatory by this tool. I keep asking you to cite your sources, and you keep citing yourself. Are you telling me that you curate your own list and load that into the tool? Why can't I find this list in the documentation? Dhaluza (talk) 22:39, 16 August 2021 (UTC)

Why it is flagged by the tool? Because EUSER is a predatory publisher. See above, and yes I curate the list. Headbomb {t · c · p · b} 23:50, 16 August 2021 (UTC)

And is the list secret, or can any editor see what is on the list? Dhaluza (talk) 00:51, 17 August 2021 (UTC)

The source code is public at User:Headbomb/unreliable.js. Headbomb {t · c · p · b} 00:56, 17 August 2021 (UTC)

OK, so the predatory black-list is hardcoded? And is there any metadata on when items are added/dropped and why? Dhaluza (talk) 01:07, 17 August 2021 (UTC)

You can browse the script's edit history if you really want. As for the 'why', see the above. I add sources based on discussions, and whenever I encounter sources that are less-than-stellar that should be flagged. Headbomb {t · c · p · b} 01:10, 17 August 2021 (UTC)

OK, but you don't keep notes on why items are on the list, so an editor who sees a site flagged just has to take your word that it's definitely predatory (not potentially predatory, or something less definitive), and there is no way for them to evaluate that conclusion on their own, because there is no backup metadata?

As for EUSER, I just emailed a couple of their academic references to see if their comments were legit and both confirmed they were. One also offered that when his students publish there, they don't pay a fee. Dhaluza (talk) 01:40, 17 August 2021 (UTC)