Wikipedia:Village pump (idea lab)

From Wikipedia, the free encyclopedia
Jump to: navigation, search
  Policy   Technical   Proposals   Idea lab   Miscellaneous  
Shortcuts:
The idea lab section of the village pump is a place where new ideas or suggestions on general Wikipedia issues can be incubated, for later submission for consensus discussion at Village pump (proposals). Try to be creative and positive when commenting on ideas.
Before creating a new section, please note:

Before commenting, note:

  • This page is not for consensus polling. Stalwart "Oppose" and "Support" comments generally have no place here. Instead, discuss ideas and suggest variations on them.
  • Wondering whether someone already had this idea? Search the archives below, and look through Wikipedia:Perennial proposals.
« Older discussions, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22


Centralized discussion
Proposals: policy other Discussions Ideas
For a listing of ongoing discussions, see the dashboard.


Accessible Citations (Experiment with .PDF#page links)[edit]

>>>> Updated Version =) <<<<<


The Goal:

  • Find the easiest/quickest route to follow a reference all the way back to the original paragraph/sentence on the original document in which the writer/editor was reading.

Current Experiment:

Where this can be Improved:

  • Ideally, use one of the simple {{command|name|#page}} when citing sources in the WikiCode, and automatically create http://example.pdf#pageNumber external links to the specific pages.
  • Use something other than {{efn}}

Example (still in experimental phase):

This is how it looks in WikiCode:

<nowiki>
Sentence.
{{efn|[https://ia601509.us.archive.org/18/items/LetterFromRome/Letter_from_Rome.pdf#56 ''Letter from Rome'', page 56.]}}

Another Sentence.
{{efn|[https://ia601509.us.archive.org/18/items/LetterFromRome/Letter_from_Rome.pdf#71 ''Letter from Rome'', page 71.]}}

== References ==

{{reflist}}

|- !Result |- | MIT Server downloaded articles.[1] Aaron Swartz is legendary.[2]

Normal Referenced content.[1]

open access publication – free to read Accessible Citations
1.^ JSTOR Evidence.[1] [PDF]. View Page 3127.
2.^ JSTOR Evidence.[1] [PDF]. View Page 3142.
References
  1. ^ a b c JSTOR (30 July 2013). "JSTOR Evidence in United States vs. Aaron Swartz" (PDF). Archived from the original (PDF-1.6) on 1 March 2017 – via Archive.org. 

|}

Popcrate (talk) 12:29, 1 March 2017 (UTC)

I'm dubious about using the {{ref}} template, which is unpopular. I believe that mw:Extension:Cite.php isn't meant to support nested refs like that anyway. Why don't you just make separate citations for the source when you change the page number? WhatamIdoing (talk) 02:01, 2 March 2017 (UTC)
Insert plug for the Comm Tech tasks related to phab:T138601. --Izno (talk) 12:52, 2 March 2017 (UTC)
@Izno: this looks very related to what I am thinking of =) It would definitely be better to program in some functionality, rather than misuse a bunch of other templates and references ;) How would you recommend I get involved over there? Popcrate (talk) 23:34, 2 March 2017 (UTC)
Whatever we do, it should allow the linked to page in the PDF to not be the page number that the person sees on Wikipedia. That is because page XYZ of the PDF is often not page XYZ of the document. For example, extra intro pages are included often or the PDF is just part of a larger document. AManWithNoPlan (talk) 21:05, 12 March 2017 (UTC)

Good point, AManWithNoPlan , and I agree... A further use I could see for this, is to add links to existing references with page numbers. For example... let's say a book becomes part of the public domain, or is already part of public domain, and can then add links to the new book. <- That would make it especially important to have a separation of actual page number versus PDF page number (in order to preserve the "real" page number from original text). Popcrate (talk) 04:58, 18 March 2017 (UTC)

Because User:Izno referred me to the phabricator page, it looks like I'm not the only one to consider this, and I would definitely be willing to program it in. I think there just needs to be some consensus on how it should appear on Wikipedia. I think there's a lot of potential having very accessible reference links in Wikipedia, and having a way for editors to easily create them will be essential. Popcrate (talk) 04:58, 18 March 2017 (UTC)

Clarification that some lists are not intended to be exhaustive, but only should include items which are standalone notable[edit]

We have quite a few articles in Wikipedia which are lists or contain lists. (More information can be found in the MOS - Lists)

There are quite a few types and they can be categorized in various ways but for the present discussion I'll break them into two categories:

  1. Lists intended to be exhaustive (The list itself must be notable while the items in the list may or may not be separately notable.) E.g. List of characters in The Lion King
  2. Lists not intended to be exhaustive, and inclusion means the entries should be individually notable. E.g. List of Medal of Honor recipients


At OTRS, it is quite common to receive an email from someone who notes the omission of an item in a list and asks us to add it. In some cases, the request is legitimate and we attempt to be helpful, but in many cases, they mistakenly think the list is intended to be exhaustive rather than a list of items which in addition to meeting the inclusion criteria, are also notable and have an existing article.

I think it would be useful to add a head note to lists of type 2, to let readers know that the list is not intended to be exhaustive with respect to items in the universe but only exhaustive with respect to items that are individually notable. Like investment my motivation is to cut down on the number of people who ride into OTRS to request that their friend or relative be added to some list, I think it would provide a useful service to readers who might otherwise be misled.

I emphasize the two types of lists, because if we could agree on wording for such a head note, it should not be indiscriminately added to all standalone or embedded lists, it should only be added to those that qualify as type 2.

I'm including this in the idea section rather than as a formal proposal because I think there are some details to be worked out. For example, it may be unobtrusive to add such a head note to standalone list articles, but I'm not sure how best to include this information in this case of school and location articles which often have a section for notable alumni or residents.--S Philbrick(Talk) 15:55, 11 March 2017 (UTC)

The difference between lists that are intended to be complete and those that are not can already be signaled: {{Dynamic list}} (not intended to be exhaustive) versus {{Incomplete list}} and {{Complete list}} (intended to be exhaustive).
Of course, there are various list inclusion criteria (notability being just one of them) and there has never been a convenient way to signal editors, let alone readers, what the criteria is. There are lists that are dynamic (not intended to be exhaustive) but individual entries don't need to be notable. – Finnusertop (talk ⋅ contribs) 16:09, 11 March 2017 (UTC)
I don't see that as adequate. For example, the specific motivation is this article: List of books written by children or teenagers. A reader wrote in to tell us about a book not on the list that was written by someone under the age of 20. I'm not sure which template you think belongs on that list but I don't see how any of them would send the message that a non-notable book should not be added.--S Philbrick(Talk) 16:12, 11 March 2017 (UTC)
List articles are expected to be clear about their inclusion criteria. However, many such lists identify the inclusion criteria without stating that this particular list also requires that the individual items be notable. My guess is that this is not mentioned because most experienced editors treat it as implicit. I don't disagree, but I can tell you from experience that many readers do not pick up on this. I am not attempting to propose ways to improve the general inclusion criteria discussion — I am only pointing out that the implicit need to meet the notability hurdle is misunderstood by many readers and I'd like to find a way to make it explicit rather than implicit.--S Philbrick(Talk) 16:18, 11 March 2017 (UTC)
Moving pages such as List of books written by children or teenagers to List of notable books written by children or teenagers could go a long way to solving the problem. Roger (Dodger67) (talk) 11:14, 13 March 2017 (UTC)
I suspect that most readers don't know what we mean by "notable". WhatamIdoing (talk) 20:17, 22 March 2017 (UTC)
If we're compiling a list of members, all of whom must be sufficiently notable to have their own article, maybe it shouldn't be a list at all but a category? Chuntuk (talk) 12:46, 30 March 2017 (UTC)

A "Wiki word processor"[edit]

I'm so comfortable and familiar with wiki-markup that I frequently find myself using it while working on MS Word. Then I get frustrated when the headings, italics, footnotes, etc don't appear as I want them. I find it so muck quicker and easier to simply type raw wiki-code than having to constantly click toolbar buttons. Is there a way to "teach" MS Word to parse some basic MediaWiki markup? Or how about some clever coder(s) create a "Wiki word processor" app that uses MediaWiki markup. Roger (Dodger67) (talk) 11:10, 13 March 2017 (UTC)

You can look at this. Ruslik_Zero 20:46, 14 March 2017 (UTC)
Ruslik0, thanks but that is actually the exact opposite of what I'm looking for. That MS addon converts the MS Word document format to MediaWiki markup. What I want is a word processor that natively uses wiki-markup. The rules of Wikipedia does not allow me to use my sandbox for my own personal, business and academic writing. What I'd like is the MediaWiki "source editor" as a standalone app. Roger (Dodger67) (talk) 15:31, 15 March 2017 (UTC)
Almost certainly not. There's no single standard for wikitext, and even within WMF projects there's variation between how formatting is handled—those footnote templates are mostly en-wiki specific templates, not Mediawiki modules. I suppose theoretically you could embed Parsoid into a word processor, but I can't imagine anyone would bother, especially since the devs are trying to discourage the use of Wikitext markup. (As a very clumsy fudge, you could use Word's autoreplace function to turn Wikitext markup into Wordstar markup (e.g., '''text''' becomes *text* and ''text'' becomes _text_), which Word can be coaxed into understanding, but it seems like more trouble than it's worth. (Is typing three apostrophes really any easier than ctrl-b, anyway?) ‑ Iridescent 16:25, 15 March 2017 (UTC)
(Adding) This may be a stupid question, but if you want something with the look and feel of Mediawiki but without using Wikipedia, why not just install Mediawiki on your own computer? ‑ Iridescent16:29, 15 March 2017 (UTC)
(edit conflict) It's not exactly what you're asking, but you don't actually have to "constantly click toolbar buttons" in MS Word. You just need to learn the keyboard shortcuts. For example, you can tell it to put things in italics by holding down the "Ctrl" button and pushing "i" ("Ctrl+i"), bold is Ctrl+b, underline is Ctrl+u, etc. Headings get a LOT easier if you learn to use styles. The biggest problem with MS Word is that people don't know how to use it, and the designers want it to be as intuitive as possible, so they hide some of the more powerful features like styles. ~ ONUnicorn(Talk|Contribs)problem solving 16:31, 15 March 2017 (UTC)
If you know (or are willing to learn) Markdown, pandoc can be used to convert it to html, pdf, epub, or even .docx formats. Many text editors support it – vim comes with markdown syntax highlighting out-of-the-box, I believe; emacs, if it doesn't have it by default, certainly has a package for it. For a fuller-featured, but more complex and harder to learn, alternative, LaTeX might be worth investigating. Caeciliusinhorto (talk) 10:55, 17 March 2017 (UTC)
mw:Extension:VisualEditor can be run stand-alone (with or without MediaWiki). With its new built-in wikitext mode (which might or might not be packaged up for third-party installation right now), you could type in wikitext markup, switch to visual mode, and then copy and paste the contents to a word processor if you needed a particular format. I would be worried about it not handling complex formatting though: character formatting and links would probably convert nicely, but ref tags would be hopelessly lost or garbled.
The last time I used Microsoft Word (which was a l-o-n-g time ago), it was also possible to define all manner of keyboard shortcuts, and if that's still possible, then you could probably define keyboard shortcuts that way. Whatamidoing (WMF) (talk) 20:25, 22 March 2017 (UTC)
Yes, it's still possible to define keyboard shortcuts; I do it all the time. How you go about doing it depends which version of Word you're using as they're always moving features around and hiding them in different places. ~ ONUnicorn(Talk|Contribs)problem solving 20:35, 22 March 2017 (UTC)

Idea: proof of self teaching with Wikipedia[edit]

Just an idea that could be discussed.

Wikipedia has a lot of knowledge and many people are eager to learn from it. What if an automated quizz could be offered to the reader at the end of an article so that eventually, if passed, the latter would receive a "proof of knowledge", like a MOOC but self-taught.

At the end of the day this person eager to learn could receive this kind of certificate and put it on his/her CV and LinkedIn profile for example.

If this idea is taken further, companies could ask their new recruits (or even candidates for a position) to pass these tests so that people could prove they have some kind of knowledge on a specific topic.

One issue we could see: the answers could be put online and everyone could cheat very easily and obtain 100% on any topic. Thus one solution would be to develop a smart algorithm to draw more or less randomly generated quizz. — Preceding unsigned comment added by 2A02:120B:2C4C:7BB0:D8E:38E8:7AF2:752 (talk) 11:27, 17 March 2017 (UTC)

That's not a bad idea. I wouldn't do it but I could see how some people would eat it up. I'd suggest it be limited to FA articles, though. There's a lot of terrible articles here which study of shouldn't entitle one to even a PDF download of a certificate. BlueSalix (talk) 03:44, 18 March 2017 (UTC)

Wiki 4 Coop[edit]

Hello everyone,

I come to you to invite to re-read the submission of a new partnership project between the Wikimedia movement and the Belgian NGOs. The project is titled Wiki 4 Coop and I invite you to discover its submission page on Meta-Wiki. Do not hesitate to endorse the project if you like it and even correct my English if you have a little time. A beautiful end of day for all of you, Lionel Scheepmans ✉ Contact (French native speaker) 11:44, 17 March 2017 (UTC)

The effects of milestones on achieving A-class articles[edit]

I've been improving articles in an effort to get GA status, and eventually FA status, but I came to the realization that it completely undermines the value of A-class status. A-class is supposed to be when an article is "considered 'complete'", according to Wikipedia:Article classes. In fact, look at the statistics: there are more Good Articles than A-class articles, not because A-class status is hard to achieve, but because GA and FA status are both milestones that are preferable to A-class (which, incidentally is why there are far more FA as well; note that I don't have the full statistics, but in WikiProjects such as CGR and Military History that is the case, whereas WP such as Video Games have eschewed A-class altogether). Why do we settle with GA status, if A-class is supposed to be the goal? People just edit until they've reached GA status and stop because that's considered the milestone, and only settle for A-class if they can't get if Featured. A-class IS the goal, because a complete article is the goal.

I'm not suggesting we get rid of GA and FA status, but that we should give reason for people to bother writing an A-class article - which is supposed to be the highest point of completion achievable in articles, spare Featured. So:

  1. What are everyone's thoughts on some WikiProjects eschewing A-class altogether?
  2. Thoughts on ways to encourage people to achieve A-class after GA status?
  3. Is there a point in claiming A-class is the point of completion when, in practice, FA status is that point? Psychotic Spartan 123 07:07, 19 March 2017 (UTC)
One factor is that virtually no wikiprojects have functioning A-class review processes - milhist does; I have not seen CGR's a-class review used since I have been around, and iirc the last discussion on CGR's a class review essentially concluded that it was dead... Caeciliusinhorto (talk) 09:05, 19 March 2017 (UTC)
The other issue is that these X-class schemes are effectively obsolete; a remnant of the now dormant Wikipedia:Version 1.0 Editorial Team process and should probably be scrapped except in these projects where they are still in use. Jo-Jo Eumerus (talk, contributions) 09:52, 19 March 2017 (UTC)
I looked at Wikipedia:Version 1.0 Editorial Team and it has the full statistics there: only 1,678 A-class articles, but 5,945 FA and 27,948 GA as of this time. Unrelated, but it's sad that the majority of all articles are stubs and vast majority are below C-class. Should it be proposed that we give A-class a review process independent of WikiProjects, i.e., have an A-class project page similar to GAN and FAC? That way at least you won't need to rely on possibly dead projects to get an article promoted. Psychotic Spartan 123 13:31, 23 March 2017 (UTC)
I don't see that there's a great deal of point. We already have two review processes which would ideally be more active; adding a third would just further split this effort. And I don't see any great advantage of having A-Class reviews back in action. It's another stepping-stone on the way to FA status, but many FAs don't go through the existing stepping-stones already. Caeciliusinhorto (talk) 13:55, 23 March 2017 (UTC)

Fixing Special:WantedCategories[edit]

Users have created user categories for the purpose of collaboration and humor. However, some categories that people place on their userspace haven't actually been created. This causes the category to appear on Special:WantedCategories, which creates clutter on the page and hinders users who try to create wanted categories. For a more detailed explanation, you can read this RfC[3]

Technical Proposal[edit]

  • Add a new namespace called the User Category: namespace. It functions exactly like the Category: namespace, however categories in this namespace do not appear in Special:WantedCategories.
  • Naturally, a User Category Talk: namespace will have to be created for their talk pages.
  • For the purpose of this discussion, let's assume that the above proposal is possible. We can ask the devs to implement it if there is consensus

Policy Proposal in conjunction with technical proposal[edit]

  1. Categories (red and non-red) intended solely for use in the userspace should be moved/changed to the User Category: namespace
  2. Pages moved via the above criteria shall have their talk pages moved to the User Category Talk: namespace.

Discussion[edit]

If anyone can think of any ideas to improve these proposals (or suggest a better alternative proposal) please share. Gamebuster19901 (Talk║Contributions) 18:27, 24 March 2017 (UTC)

Maybe we need a MediaWiki:Wanted-categories-exceptions message to list all these unwanted categories rather than an ad hoc namespace. Jo-Jo Eumerus (talk, contributions) 19:31, 24 March 2017 (UTC)
Yeah. That'll still allow us to find redlinked categories for userspace that we genuinely want to be blue, and would likely to be much easier to implement. —Cryptic 21:32, 24 March 2017 (UTC)

Personally, I'd expect a request for this to be declined as "no, just don't do that". There are many things developers' time could be better spent on than making it so people could fool around having redlink categories on their user pages. Anomie⚔ 23:10, 24 March 2017 (UTC)

That, and the implementation details are not trivial. There's a *lot* of weird logic behind how "special" namespaces like NS_FILE and NS_CATEGORY work. Replicating those for other namespaces isn't easy (I've tried before and gave up). ^demon[omg plz] 21:43, 31 March 2017 (UTC)

Developing the deleted articles[edit]

I think it would be a very good idea to help the users to move the deleted articles into their own user space or at Wikia, so they can keep working on them. I noticed there are a lot of articles about artists and writers who are just a bit below the notability line. For example this article Ioan Groșescu which was deleted from the Romanian Wikipedia. We are talking about people who, even though they are not important enough for Wikipedia to keep an article about them, they are quite significant for the Romanian culture, and it's a big waste to completely forget about them. And many times those people are notable, but the users who created the articles are simply don't have enough Wikipedia experience to prove it when building the articles. IMO, when the users are notified that the article they created was proposed for deletion, they should be aware of the fact that:

  • They can ask for the article to be moved in their userspace so they can keep improving it until it's good enough for Wikipedia
  • Or they can ask for moving the article at Wikia, so the article will not be completely lost, even when the person the article is about is not notable enough for Wikipedia. Or to keep working at the article at Wikia until they satisfy the higher standards at Wikipedia.

The users should have access to very easy to use tools that can place such requests.

I think such a facilitation would have a significant beneficial effect on developing Wikipedia. —  Ark25  (talk) 22:18, 25 March 2017 (UTC)

Anyone who wants the text of a deleted article can already just go to Wikipedia:Requests for undeletion and ask for the deleted text back. We're not going to host userspace drafts that aren't viable Wikipedia articles indefinitely (we're not a free webhost, and we delete drafts after a while if nobody's actively trying to improve them). What Wikia—which is a private company and has nothing to do with Wikipedia—does, is entirely a matter for them. ‑ Iridescent 22:27, 25 March 2017 (UTC)
Sure but I think that Wikia + Wikipedia can have some cooperation that is beneficial for both of them and for developing the access to knowledge in general. There is a lot of valuable knowledge that was deleted from Wikipedia simply because the users didn't know how to prove that those articles meet the notability criteria. —  Ark25  (talk) 10:05, 26 March 2017 (UTC)
Take a look at http://speedydeletion.wikia.com/wiki/Main_Page which is a "Speedy deletion Wiki ". It does not have everything, and afaik prople don't develop pages there. Graeme Bartlett (talk) 02:14, 29 March 2017 (UTC)

Intelligent Wikipedia that talks to you interactively[edit]

@Rich Farmbrough, Anomie, EpochFail, Headbomb, Beeblebrox, Redrose64, Magioladitis, and Od Mishehu:

Either in voice or text. Or both.

It could start out by saying:

"How can I help you?"

or...

"What would you like to learn about today?"

or...

"What would you like to know?"

You answer its inquery, or ask it a question. It then uses algorithms to access its vast collection of data (Wikipedia, as a corpus, and indices/tables constructed for this project), and provides an answer.

This concept was proposed as a WikiProject to develop something like this way back in 2010, and with the technology advancements that have occurred since then, this project idea may deserve another look:

Dear Wikipedia,

I am Giuseppe (Joseph) Zaccaria (Zachariah), an Italian citizen resident in Iran (my mom is Iranian) born in 19 October 1993, I study mathematical physics in the third year of high school in an Italian school in Tehran. I speak 3 languages fluently (English, Italian, Persian) and have also studied for 3 years Latin and 1 year Spanish. My dream has always been to become a scientist in NASA (although the odds for that are overwhelmingly low), since I love space and anything related to that. I really hope to be able to come to the US, possibly to study.

Now, I use Wikipedia constantly, especially for school projects and studies, and I must say BRAVO!! i think all the students of the world are more than glad to have a site in which you can find almost anything. Nevertheless, each time i have had a question (the answer of which might have been just a few words), Wikipedia gave me a huge array of articles that it thought could be related to what I was looking for, and i must say that that is always a bit disappointing. So, i thought why not add another part in Wikipedia (you know, like wikibooks...) called something like Wikibrain, or WikiAnswers..., where you write a question, and the program automatically searches the articles it thinks might be related to the question, and then finds the answer, without giving you the option to go and look for the answer in the articles yourself. I mean a new program capable of relating words given from the user (i mean the words that are in the question) to other words already existing in the domain (pretty much what it already is doing), and then finding the words in the article (found in the domain) which are most likely related to the primary words that formed the question (given from the user). So, when you ask "distance between Paris and Rome?" the program would automatically find the answer in its articles and say "500 kilometers" (of course this is just an example, I'm sure that if you work on it, it will resolve more complex questions). I am sure that something like this would be a huge step in improving Wikipedia, so I really hope that my idea has interested you and that i might help you in some way. I am really looking forward to receive an answer from you,

Yours sincerely,

Joseph Zachariah

The technology exists for this, right?

What would it take to develop something like this? Would it be any harder than developing WP:AWB, for example? How was AWB developed?

What other features could this thing feasibly have?

What should it be called? WP Genie? Alfred? Jarvis? WikiJarvis? The Transhumanist 20:23, 27 March 2017 (UTC)

You may want to consider just why Wikipedia:Knowledge Engine is a red link—User:GorillaWarfare's summary here is a very good place to start, if you're not familiar with the back-story—for why this is not a rabbit-hole you want to go down. (And no, the technology doesn't exist for this; multinational corporations have spent—literally—billions developing Siri, Google Now and Cortana, and they struggle to answer questions more complicated than "What is the capital of France?".)  ‑ Iridescent 20:29, 27 March 2017 (UTC)
I understand why the idea might be cool, but even if the technology existed, it would be too expensive to implement for a website that, every year or two, has to beg for donations just to remain ad-free (I don't mean "beg" to be an insult). Psychotic Spartan 123 20:36, 27 March 2017 (UTC)
Technology is being developed to do these kinds of things, but as Iridescent said above, it requires immense resources and isn't something we could just sort of "plug in" to Wikipedia. As for your question about whether it would be any harder than developing WP:AWB: yes, considerably. The two are really not comparable.
This is not a project that could be developed by a handful of bot developers; it would require buy-in from the Wikimedia Foundation and the devotion of significant resources. Again, as Iridescent pointed out, this would be a similar move to when the Knowledge Engine project was created, and that's not something I expect the WMF is keen on repeating. GorillaWarfare (talk) 20:42, 27 March 2017 (UTC)
I know the idea lab exists to entertain ideas and not to vote on anything, but it hurts my head thinking about the resources in time and money it would take to do this. Wikipedia's dedicated editors can barely keep up with vandalism, struggle to create quality articles, and I think even hypothesizing on something of this natural is asking a lot. If you think about it, Wikipedia is not a browser, it is a destination. Services like Ceres and Cortana already do this, and under the (basic, usually crappy) answer, a list of sites are given. Among those sites, Wikipedia is usually on the top of the list with an article to answer all the questions on that subject you could possibly imagine (because it, of course, would be a quality article and not a stub). The last thing we need is another project, and another backlog, to add to the list of projects with backlogs that we cannot even keep up with as it is. The idea is cool, but the work it would take far exceeds the return. Just look at all the other backlogs we have and imagine adding another one. Psychotic Spartan 123 21:26, 27 March 2017 (UTC)
Please save the bucket of cold water for the proposal, if this brainstorming session ever inspires one. :) Right now, we're just kicking around some ideas. The most interesting aspect of this exploration (IMHO) is on how much and how could this be automated? It would give interested persons something to apply WikiBrain on. How could that tool be applied to something like this? The Transhumanist 21:36, 27 March 2017 (UTC)

What would it entail? A question generation algorithm, and a database of stock answers? And algorithms to generate the answers in the database? Optionally, editors could provide answers in the database? This is the idea lab -- a place to explore ideas. And so, I'm interested in exploring this idea further. What would such an idea consist of? What would it be built out of? What would its structure be? Any and all thoughts are welcome. The Transhumanist 21:01, 27 March 2017 (UTC) P.S.: Let's brainstorm some more. How about question redirects? (See section below).

One place you can start to see what it would entail is Watson (computer) § Description. isaacl (talk) 05:55, 28 March 2017 (UTC)
What would be needed? Teetering on the edge of The Singularity. In other words, computers that design, build, and program computers that design, build, and program computers that design, build, and program computers... — Neonorange (talk) 06:21, 28 March 2017 (UTC)
More generally, it's a leading-edge, high-technology proposal, and Wikipedia is a trailing-edge, low-tech operation. Yes, we'll get this kind of capability, many years after it has been heavily used by Microsoft, Apple, Google etc. Jim.henderson (talk) 19:28, 28 March 2017 (UTC)
I have a feeling, that this is a very farfetched idea and might be implemented to the Wiki years after this proposal. Cheers, FriyMan talk 18:03, 30 March 2017 (UTC)
Strong Oppose. Is this a joke? KMF (talk) 02:11, 8 April 2017 (UTC)

Question redirects[edit]

What about a supplemental set of redirects, which are questions, and which point to the answer in the encyclopedia? The question redirects could be prefixed with "Q:" or something. Such a system could be worked on by editors and bots. Thoughts, more ideas?The Transhumanist 21:06, 27 March 2017 (UTC)

A bot could be created to capture questions asked for which we don't yet have a Q:redirect for, and place it on a backlog list of such question redirects to consider. The Transhumanist 21:10, 27 March 2017 (UTC)
  • I agree, if we end up with all the "agents" belonging to multinationals, humankind's status will be at risk.
  • I further agree that a lot of the technology is available. These are problems I have been thinking about since about 1972/3 and while they are by no means trivial, the apparatus is there.
  • I don't have significant time to devote to this now - the years I was intending to spend on similar Wiki problems were unfortunately disrupted.
All the best: Rich Farmbrough, 21:19, 27 March 2017 (UTC).
(Though the "stock questions and stock answers" machine has already been made by a company in Cambridge.) All the best: Rich Farmbrough, 21:21, 27 March 2017 (UTC).


I tried out some questions in the search box...
"Who is John Wayne?" returned John Wayne at the top of the search results. Why shouldn't this be reprogrammed to go straight to the article instead, and save the reader a point and click?
What is the meaning of life? actually points somewhere. :) Unfortunately, it is just a reasking of the question, rather than an answer. But at least it gets you to the right page.
Who killed Abraham Lincoln? returns Abraham Lincoln: Vampire Hunter at the top of the list.
How far is it to Washington D.C.? gets Save Outdoor Sculpture! as the top result.
The search box currently isn't very good at answering questions, but it answered two of the questions above correctly! Maybe this warrants more testing. The other two may indicate another way to generate semi-random links. :) The Transhumanist 22:05, 27 March 2017 (UTC)
Has anyone working on this talked to the researchers at OCLC about their QuestionPoint[1] product? It does a lot of this already and connects to a real librarian via chat or email (library decides format) when there isn't a stock answer. TeriEmbrey (talk) 14:50, 28 March 2017 (UTC)
Of the four questions you give, one is not answerable by a simple redirect, one already redirects somewhere useful, and two give search results that provide the answer: while Abraham Lincoln: Vampire Hunter is the top search result for "who killed abraham lincoln", Abraham Lincoln is the second result and assassination of Abraham Lincoln the fourth.
More generally, I would like to know more about how this proposal could work. Obviously it would be impractical to create redirects for every conceivable question that is answered by any of our articles. Does wikipedia collect statistics on what the most common search terms are? If it did, could we parse those to find questions? Probably not accurately without a lot of work. How else could we do this? Leave it up to editors' judgements? Is there anything stopping editors creating these kinds of redirects already if they think they are useful? Clearly some of these kind of redirects already exist: see What is the meaning of life?
Alternatively, we could create a set of automatic redirects for a limited subset of possible questions: say, any search in the form "Who is foo?" which doesn't itself exist as an article could automatically redirect for the article for foo. But I can think of only two ways of doing this: 1) use a bot to automatically create these redirects, which would leave us with millions of redirects, many likely to be totally unused, all of which would require editorial maintenance, or 2) patch the search software, which would require persuading a developer that this would be sufficiently useful to be worth development time – and I am not sure that it is... Caeciliusinhorto (talk) 18:55, 28 March 2017 (UTC)
"Is there anything stopping editors creating these kinds of redirects already if they think they are useful?" Yes. The type is not included at WP:POFR, and they get targeted for deletion via WP:RFD, or speedy deleted. It would take a proposal, which is why we're banging the idea around here first. The Transhumanist 22:57, 28 March 2017 (UTC)
The sole work on this that I know of is this very discussion. Thanks for the QuestionPoint link. I had never heard of that. The Transhumanist 23:00, 28 March 2017 (UTC)
"Does wikipedia collect statistics on what the most common search terms are?" I don't know. I've posted your question over at WP:VPT, so we will soon find out. The Transhumanist 23:08, 28 March 2017 (UTC)
See:
https://phabricator.wikimedia.org/T115085
https://lists.wikimedia.org/pipermail/wikimedia-l/2016-July/084745.html
The Transhumanist 00:57, 29 March 2017 (UTC)
"How else could we do this?" Perhaps with WikiBrain? See its homepage on Github The Transhumanist 23:12, 28 March 2017 (UTC)
What is the meaning of life? only exists because it is the topic of an article section. Likewise, Who am I? is the title of many works, and has a disambiguation page. I was thinking that questions would be prefaced with "Q:", in order to avoid clashing with article titles. The Transhumanist 23:17, 28 March 2017 (UTC)
@Caeciliusinhorto: Concerning automatic "redirects", that would most likely be a search engine function. The Transhumanist 23:20, 28 March 2017 (UTC)

Creating a set of redirects is a kludge. It would be better to improve the search engine, which is clearly possible: if I type the first three questions into a very popular search engine, an appropriate Wikipedia page is either the first answer or among the top three results. The last question doesn't seem to be suited to an encyclopedia, but the search engine answers it quite effectively. – Jonesey95 (talk) 23:34, 28 March 2017 (UTC)

But how is the search engine doing it? With a table of redirects? Regex parsing? Something else? The Transhumanist 00:20, 29 March 2017 (UTC)
The search engine uses the redirects and keyword searches of title and content. This is not always fantastic, especially when the search includes a mis-spelling: the engine will prioritise articles with redirects from the mis-spelling, rather than synonymise the word. All the best: Rich Farmbrough, 19:29, 29 March 2017 (UTC).

What's the search engine and extensions used for Wikipedia? The Transhumanist 00:59, 29 March 2017 (UTC)

Wikipedia uses mw:Extension:CirrusSearch. Special:Version has a list of installed extensions at the English Wikipedia. PrimeHunter (talk) 02:00, 29 March 2017 (UTC)
While search engine intelligence is great, redirects are a Chinese room type of AI. Clearly there are context issues which could not be covered well by such a mechanism "Who did John Smith marry?" would do better to redirect the disambiguation page than to any section on any article about a John Smith. All the best: Rich Farmbrough, 19:27, 29 March 2017 (UTC).

Theoretically, Wikidata is just great for this. There is one tool ask (which probably isn't finished), which can answer to some questions. And with SPARQL queries you can query for many things, like distance between Kanzas City and Washington. --Edgars2007 (talk/contribs) 12:57, 30 March 2017 (UTC)

If my Firefox can pass a search argument to Google Search, I don't see why Wikipedia couldn't do the same with site:en.wikipedia.org appended. ―Mandruss ☎ 19:26, 30 March 2017 (UTC)

References

  1. ^ https://www.oclc.org/en/questionpoint.html

Would like to see wikiproject hotels more active[edit]

I feel that articles in hotels are lacking in quality and quantity. Many of them are written like advertisements and wikiproject hotels seems abandoned. I would like to see more quality articles about hotels. I think the way to do that would be to make wikiproject hotels more active, I do not know how to propose this/where to propose this. Admins feel free to move this to where it should be. — Preceding unsigned comment added by Yonikasz (talk • contribs) 02:53, 3 April 2017 (UTC)

I think the most effective way to make a dormant wikiproject more active is to do it yourself. If the former coordinator (or whatever they called themselves) is still active, you might want to drop a note on their talk page. Otherwise, don't wait for permission - start improving articles. Anyone can edit still has some meaning here, after all. Eggishorn (talk) (contrib) 16:24, 6 April 2017 (UTC)

Improving backlogging of WP:templates for discussion[edit]

The insufficient participation at WP:templates for discussion was discussed in September 2016, the same month when the stagnation of the process was also discussed. Currently, the process still has a huge amount of nominations that have very little participants. Now that the consensus agreed to apply WP:PROD to "File:" namespaces, what about applying PROD to templates? Alternatively, what about creating a separate PROD for templates? If neither solution is helpful, how else do we resolve the backlogging and participation issues? --George Ho (talk) 04:17, 4 April 2017 (UTC)

Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_(idea_lab)&oldid=774382752"