User talk:Headbomb

User	Talk	Archives	My work	Sandbox	Resources	News	Stats

---

subst belongs inside the curly brackets

Regarding [1], subst belongs inside the curly brackets {{subst:PAGENAME}} and not before. I have fixed all the Template:Globalize subpage nominations. PrimeHunter (talk) 10:10, 10 April 2019 (UTC)

Redirecting my question from Village Pump (Technical)

Moved to Wikipedia:Village pump (technical) § Changing the case of a template

Please keep the discussion there, since others may be more able to help than I can. Headbomb {t · c · p · b} 19:13, 10 April 2019 (UTC)

Mouse Genome listed at Redirects for discussion

An editor has asked for a discussion to address the redirect Mouse Genome. Since you had some involvement with the Mouse Genome redirect, you might want to participate in the redirect discussion if you wish to do so. Citing (talk) 04:07, 17 April 2019 (UTC)

Mobile versions

Correct me if I'm wrong but isn't the mobile version of WP a hot topic in that many editors dislike it? Do you know if anything is being done to make it easier to edit using the mobile version or is there an easy way to switch to the desktop version? ^Atsme _Talk 📧 15:14, 20 April 2019 (UTC)

@Atsme: Well, I don't personally have much love for the mobile version, and don't know many who do, but I'm no expert on the topic. I'm sure there are efforts to make things better, but like anything it takes time to get there.

As for how to switch to the desktop version, usually you just have to scroll to the bottom of a page and there's a link there. Headbomb {t · c · p · b} 15:35, 20 April 2019 (UTC)

Rationale for removal of previous edit on ʻOumuamua page?

Moved to Talk:ʻOumuamua

Headbomb {t · c · p · b} 17:00, 20 April 2019 (UTC)

RetractionBot

You might be interested to know that I've just filed the BRFA for version 1 of RetractionBot - Wikipedia:Bots/Requests for approval/RetractionBot! It's only looking at a set of Crossref DOIs right now - with it only being my second bot I wanted to get some more eyes on it and have that first run approved, and then I'll dive back in and take a look at incorporating PubMed data too. Sam Walton (talk) 22:48, 21 April 2019 (UTC)

Update

Hello again. The thread on Doc James's user talk is now archived. So here I am to explain what happened next.

I spent time last week on coding up Beall's list, as much as is possible, on Wikidata. The main result is at d:User:Charles_Matthews/Focus list and journal queries#Clear matches. The SPARQL query directly below can now be used to search the ScienceSource working list (focus list), and see what comes up.

The publishers involved are Baishideng Publishing Group, Bentham Science Publishers, Frontiers Media, Impact Journals LLC, and Pulsus Group. Well, what I actually did first was to go over around 80 journals without statement of publisher, most of which were published by or for medical societies of various kinds. For Impact Journals, and Canadian Center of Science and Education, I then set up as items on Wikidata so that they could be used to filter. When I add more to the focus list, shortly, I expect to see more suspect publishers.

So you mentioned Bentham before. Baishideng, per the Wikipedia article, is trying to clean up its act. Frontiers Media likewise. Pulsus is in the shadow of its takeover by OMICS. Impact Journals — is there any reason to dispute Beall's verdict?

There is clearly a time dimension in some cases. It is quite possible to treat articles in the Pulsus journals as OK up to the takeover, for example. This is getting fancy, but the same principle could be applied to DOAJ approval dates for whitelisting too. While it wouldn't hurt too much to exclude all articles from these publishers, code that has a more precise rationale for exclusion and inclusion is doing a better job, and raising the right kinds of issue in so doing.

Do you have comments? What you said last time did prompt me to do work, adding to the project. Charles Matthews (talk) 16:44, 22 April 2019 (UTC)

a bot that automatically identifies if scientific references are supported, contradicted, or mentioned.

Hi Headbomb!

I came across this tweet (https://twitter.com/samwalton9/status/1120267906762334209) and after doing some digging came across your name and work in the bot approval group.

I am the co-founder of scite.ai, a new platform that aims to make science more reliable by classifying citations as supporting, contradicting, or just mentioning using a deep learning model. In short, allowing anyone to see how an article has been cited.

I would like to create a bot that shows if scientific articles on Wikipedia are supported or contradicted, similar to the approach with retracted articles but wanted to know if you thought such a bot would be approved.

Would love to know your thoughts on this!

Thanks! Josh — Preceding unsigned comment added by Jdiogenes86 (talk • contribs) 19:30, 23 April 2019 (UTC)

My thoughts is its going to be *insanely* hard to train a neural net to determine if any piece of information in Wikipedia is supported or not by a certain reference. You're welcomed to try, but I don't foresee such a bot editing articles anytime soon. It might have a better chance producing reports of things that have a good likelyhood of being false, so encouraged human review, but even there training the network will be very difficult. Headbomb {t · c · p · b} 19:43, 23 April 2019 (UTC)

And if the bot is simply reading Wikipedia for an external project, there's not much issue. You might want a WP:BOTFLAG to have higher API limits, but you are likely better off downloading database dumps. Headbomb {t · c · p · b} 19:46, 23 April 2019 (UTC)

Sorry, I don't think my original message was clear enough. Effectively, I would like say if scientific citations in wikipedia are supported or not but not the page itself, which would indeed be exceedingly difficult! To give you an example, for citation #50 in this page (https://en.wikipedia.org/wiki/Amygdala), I would like to show the high-level information from our scite with a link to our open report (https://scite.ai/reports/association-between-amygdala-hyperactivity-to-gVamGz). That make sense?