Wikipedia:Village pump (idea lab)

From Wikipedia, the free encyclopedia
  Policy   Technical   Proposals   Idea lab   Miscellaneous  
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.


Question redirects

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

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 (talkcontribs) 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)
Yonikasz, you can try to WP:REVIVE the WikiProject. You might also be interested in wikivoyage:, which is a travel guide. WhatamIdoing (talk) 19:51, 11 April 2017 (UTC)

Unify Templates in wikidata

is it possible to Unify Templates and add it to wikidata for all languages. only Variables can be translated. so it will make it easy for newbies, save time and storage space for Wikipedia hard drives.Waso99 (talk) 06:42, 11 April 2017 (UTC)

@Waso99: See meta:Community Tech/Central repository for gadgets, templates and Lua modules. ~barakokula31 (talk) 12:37, 11 April 2017 (UTC)
@Barakokula31: that's great. Thanks Waso99 (talk) 13:58, 11 April 2017 (UTC)

Factual Accuracy indicator in the Revisions

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Wikipedia articles undergo a lot of revisions everyday, and a casual reader going through an article has to re-check to be doubly sure about the veracity of everything he is reading. I propose the introduction of an indicator (ideally a symbol) beside the last revision which has been checked for factual accuracy, in the View History section.

  • Initially we can begin with Good/Featured Articles (which are subject to factual accuracy checks while nomination), and place a symbol besides the revision that granted the GA/FA status indicating that the article has been checked for factual accuracy till that point.
  • Subsequent edits can also be checked for factual accuracy by other trusted users (we may give the privilege to extended confirmed users to check the diffs since the GA/FA and ensure that the factual accuracy is still maintained). These users can then move the symbol to the newer revision that they are authorising.
  • In case the proposal turns out to be successful, we can extend the accuracy indicator to all the articles.

This can help improve the readability and trustworthiness of the encyclopedia. Here is an example - List of Delhi Daredevils cricketers is an outdated Featured List. The list is erroneous and incomplete and has not been actively edited post its successful nomination in 2012. It would therefore be ideal that a symbol be added besides [1] (the FL nomination revision) indicating that the article is factually correct as of this revision. Subsequent editors can continue editing the article, and trusted editors can move the symbol to a subsequent revision, post a fact check of the diff between the revision on which the symbol is currently placed to the current unchecked revision. Jupitus Smart 13:09, 11 April 2017 (UTC)

@Jupitus Smart: Wikipedia:Flagged revisions? --NeilN talk to me 21:07, 20 April 2017 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Verifier

Is it possible to design a verifier, a bot maybe, for assisting GA, FA and peer reviewers. The function will be simple. To verify the information in article from the source cited. This can be achieved by using simple AI, which finds the text of article in the cited source and returns errors and mismatches for editors to look upon. -- Pankaj Jain Capankajsmilyo (talk · contribs · count) 09:09, 23 April 2017 (UTC)

I'm aware of the be creative and positive when commenting on ideas instructions for this page, but the simple answer is "no". At the level you're talking about the overwhelming majority of sources are offline—indeed, "all the sources are online" is an automatic red flag when reviewing at FA/GA as except for a few hyper-recentist fields like videogaming it's a clear sign that the author hasn't conducted "a thorough survey of the representative literature"; Google Books has an (openly admitted) huge systemic bias towards the United States and out-of-copyright English-language materials. Thus, what you're proposing is that we digitise every book ever published (which would itself be seriously illegal), and write a bot to compare the digitised texts with the Wikipedia text which has been specifically written with the intention of avoiding close paraphrasing and thus won't retain the sentence structure or narrative flow, and will generally synthesise material from multiple sources in each paragraph. This is a task even humans with the source texts in front of them find difficult; to automate it (even if one presupposes that copyright law were repealed to allow in-copyright works to be digitised en masse and all the source materials were digitised for free by an unspecified benefactor) would take software orders of magnitude above anything currently available even to intelligence agencies and multinational corporations, let alone a non-profit with an annual budget of $66 million. ‑ Iridescent 17:53, 23 April 2017 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_(idea_lab)&oldid=776856481"
This content was retrieved from Wikipedia : http://en.wikipedia.org/wiki/Wikipedia:Village_pump_(idea_lab)
This page is based on the copyrighted Wikipedia article "Wikipedia:Village pump (idea lab)"; it is used under the Creative Commons Attribution-ShareAlike 3.0 Unported License (CC-BY-SA). You may redistribute it, verbatim or modified, providing that you comply with the terms of the CC-BY-SA