Re: [Wikidata-l] Suggestions for improvements of Wikidata
Any chance you could put this list up on the wiki? Perhaps in your user space. It'd be interesting to see these issues end up being tracked in Phabricator and hopefully fixed. :) Yours, -- Tom Morris http://tommorris.org/ ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
2015-04-13 14:54 GMT+02:00 Cristian Consonni kikkocrist...@gmail.com: With this premise, I think that Romaine's proposal for a game is absolutely doable and a good idea. I want to clarify, I mean that I agree with Gerard that the best indication for country for Battle of Stalingrad is URSS, I simply say that a game should keep it simple (so in this case, either the system is able to infer URSS as a possibility to present to the user or otherwise the user should (be instructed to) say not sure). I am much less convinced about the citizenship violations. Even if I believe that citizenship is a concept introduced with the modern nation-state, for a variety of reasons this is anyway applied to people that have lived before that state(at least in is modern form) was established. For example, Galileo Galilei is reported as an error but all the biggest Wikipedias (and some others that I am able to read) state that Galileo Galilei was Italian (catalan Wikipedia says that he was Tuscan in the artcle, but caegorizes him in the category Físics italians (Italian pysicists) and Astrònoms italians (Italian astronomers). On the other hand, the use of the name Italia to indicate at least a portion of present-day Italy goes back in history and there are mentions in documents from at least 42 b.C. (and possibly this will be the same for most Europe). C ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
Hoi, The point is very much that the battle WAS in the USSR. It is not not applicable it is one of the most important battles in the second world war. My point is that we should not forget this. The battle of Uhud was not in Saudi Arabia either... Thanks, GerardM On 13 April 2015 at 12:10, Cristian Consonni kikkocrist...@gmail.com wrote: 2015-04-09 8:29 GMT+02:00 Gerard Meijssen gerard.meijs...@gmail.com: Because the battle of Stalingrad as a battle was not fought by modern day Russia, it was fought by the USSR and Nazi Germany. Associating the battle of Stalingrad with modern day Russia is wrong on so many levels. At the time it was Stalingrad, hence the name. It will never be the battle of Wolgograd. I believe that you should have a Not applicable button to click for these cases. C ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
Well, getting a list of violations per country would not be hard, given the dates. There are, for example, 2,300 UK citizens who died 1706 or earlier: https://tools.wmflabs.org/autolist/?language=enproject=wikipediacategory=depth=12wdq=claim%5B27%3A145%5D%20and%20between%5B570%2C0%2C1706%5Dstatementlist=run=Runmode_manual=ormode_cat=ormode_wdq=notmode_find=orchunk_size=1 It would be possible to generate a daily constraint violation report for more such conditions, given a list of valid data ranges (e,g, Q145 / 1701 / now for UK). I'd volunteer, if someone makes a machine-readable list (table?) on a wiki page :-) A more fine-tuned bot could actually auto-replace some, if the new country is the same or larger as the old one. But given the numbers, it is probably not necessary to toy with such forces (we can fix a few thousand by hand once; new entries should be low in numbers). On Mon, Apr 13, 2015 at 2:22 PM Andrew Gray andrew.g...@dunelm.org.uk wrote: This is an example of a more general problem, I think - country is treated as an indefinite concept, which breaks down for historic people as well. To take Magnus's example, Wikidata records that Henry VIII was a citizen of the UK, which would no doubt have surprised him (https://www.wikidata.org/wiki/Q38370). Perhaps what we want is to figure out some way that country (P17) and citizenship (P27) can have robust constraints based on date of birth/death or on date of an event, so that - for example - anyone who is reported as having citizenship of the UK has to have been born before or died after 1707. For something like the battle, the constraint would be that the event has to have happened while the P17 country was in existence. I don't know if we can do anything this sophisticated with the current constraints system - perhaps it would have to be organised on a country-by-country basis, one report for the UK, then the USSR, and so on as we define the cases. Perhaps something to look at doing a year down the line, when we've imported a lot of data we can fix ;-) Andrew. On 13 April 2015 at 13:00, Gerard Meijssen gerard.meijs...@gmail.com wrote: Hoi, The point is very much that the battle WAS in the USSR. It is not not applicable it is one of the most important battles in the second world war. My point is that we should not forget this. The battle of Uhud was not in Saudi Arabia either... Thanks, GerardM On 13 April 2015 at 12:10, Cristian Consonni kikkocrist...@gmail.com wrote: 2015-04-09 8:29 GMT+02:00 Gerard Meijssen gerard.meijs...@gmail.com: Because the battle of Stalingrad as a battle was not fought by modern day Russia, it was fought by the USSR and Nazi Germany. Associating the battle of Stalingrad with modern day Russia is wrong on so many levels. At the time it was Stalingrad, hence the name. It will never be the battle of Wolgograd. I believe that you should have a Not applicable button to click for these cases. C ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l -- - Andrew Gray andrew.g...@dunelm.org.uk ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
2015-04-13 14:00 GMT+02:00 Gerard Meijssen gerard.meijs...@gmail.com: The point is very much that the battle WAS in the USSR. It is not not applicable it is one of the most important battles in the second world war. My point is that we should not forget this. The battle of Uhud was not in Saudi Arabia either... Ok, but I think that having a system that, for examples, cross checks dates and presents URSS as a possibility would be much more complicated to build. I think that the Wikidata game (or a similar game-like system) can not address all possible complicated scenarios, and thus there will always be some cases that should be handled directly editing Wikidata. I was following Magnus here, in the post where he introduces the Wikidata Game[1]: «So what’s the approach here? I feel the crucial issue for gamification is breaking complicated processes down into simple actions, which themselves are just manifest decisions – “A”, “B”, or “I don’t want to decide this now!”. [...] Of course, this simplification misses a lot of “fine-tuning” – what if you are asked to decide the gender of an item that has been accidentally tagged as “person”? What if the gender of this person is something other than “male” or “female”? Handling all these special cases would, of course, be possible – but it would destroy the simplicity of the three-button interface. The games always leave you a “way out” – when in doubt, skip the decision. Someone else will take care of it, eventually, probably on Wikidata proper.» With this premise, I think that Romaine's proposal for a game is absolutely doable and a good idea. C [1] http://magnusmanske.de/wordpress/?p=203 ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
Of course there will always be some things too complicated to be reasonably expressed in Wikidata, or hard to process by software. But in the case of historical datas, we better have to think of a common and practical representation and ways for tools to process datas, because this is totally a WIkipedia common usecase :) There is already tools to draw wars in a map and chronological datas. 2015-04-13 14:54 GMT+02:00 Cristian Consonni kikkocrist...@gmail.com: 2015-04-13 14:00 GMT+02:00 Gerard Meijssen gerard.meijs...@gmail.com: The point is very much that the battle WAS in the USSR. It is not not applicable it is one of the most important battles in the second world war. My point is that we should not forget this. The battle of Uhud was not in Saudi Arabia either... Ok, but I think that having a system that, for examples, cross checks dates and presents URSS as a possibility would be much more complicated to build. I think that the Wikidata game (or a similar game-like system) can not address all possible complicated scenarios, and thus there will always be some cases that should be handled directly editing Wikidata. I was following Magnus here, in the post where he introduces the Wikidata Game[1]: «So what’s the approach here? I feel the crucial issue for gamification is breaking complicated processes down into simple actions, which themselves are just manifest decisions – “A”, “B”, or “I don’t want to decide this now!”. [...] Of course, this simplification misses a lot of “fine-tuning” – what if you are asked to decide the gender of an item that has been accidentally tagged as “person”? What if the gender of this person is something other than “male” or “female”? Handling all these special cases would, of course, be possible – but it would destroy the simplicity of the three-button interface. The games always leave you a “way out” – when in doubt, skip the decision. Someone else will take care of it, eventually, probably on Wikidata proper.» With this premise, I think that Romaine's proposal for a game is absolutely doable and a good idea. C [1] http://magnusmanske.de/wordpress/?p=203 ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
This is an example of a more general problem, I think - country is treated as an indefinite concept, which breaks down for historic people as well. To take Magnus's example, Wikidata records that Henry VIII was a citizen of the UK, which would no doubt have surprised him (https://www.wikidata.org/wiki/Q38370). Perhaps what we want is to figure out some way that country (P17) and citizenship (P27) can have robust constraints based on date of birth/death or on date of an event, so that - for example - anyone who is reported as having citizenship of the UK has to have been born before or died after 1707. For something like the battle, the constraint would be that the event has to have happened while the P17 country was in existence. I don't know if we can do anything this sophisticated with the current constraints system - perhaps it would have to be organised on a country-by-country basis, one report for the UK, then the USSR, and so on as we define the cases. Perhaps something to look at doing a year down the line, when we've imported a lot of data we can fix ;-) Andrew. On 13 April 2015 at 13:00, Gerard Meijssen gerard.meijs...@gmail.com wrote: Hoi, The point is very much that the battle WAS in the USSR. It is not not applicable it is one of the most important battles in the second world war. My point is that we should not forget this. The battle of Uhud was not in Saudi Arabia either... Thanks, GerardM On 13 April 2015 at 12:10, Cristian Consonni kikkocrist...@gmail.com wrote: 2015-04-09 8:29 GMT+02:00 Gerard Meijssen gerard.meijs...@gmail.com: Because the battle of Stalingrad as a battle was not fought by modern day Russia, it was fought by the USSR and Nazi Germany. Associating the battle of Stalingrad with modern day Russia is wrong on so many levels. At the time it was Stalingrad, hence the name. It will never be the battle of Wolgograd. I believe that you should have a Not applicable button to click for these cases. C ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l -- - Andrew Gray andrew.g...@dunelm.org.uk ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
Huh, just when I sent this mail, I realized that there is a database with nation dates, it's called Wikidata... So I present: https://tools.wmflabs.org/wikidata-todo/wrong_nationality.html Have fun! On Mon, Apr 13, 2015 at 3:43 PM Magnus Manske magnusman...@googlemail.com wrote: Well, getting a list of violations per country would not be hard, given the dates. There are, for example, 2,300 UK citizens who died 1706 or earlier: https://tools.wmflabs.org/autolist/?language=enproject=wikipediacategory=depth=12wdq=claim%5B27%3A145%5D%20and%20between%5B570%2C0%2C1706%5Dstatementlist=run=Runmode_manual=ormode_cat=ormode_wdq=notmode_find=orchunk_size=1 It would be possible to generate a daily constraint violation report for more such conditions, given a list of valid data ranges (e,g, Q145 / 1701 / now for UK). I'd volunteer, if someone makes a machine-readable list (table?) on a wiki page :-) A more fine-tuned bot could actually auto-replace some, if the new country is the same or larger as the old one. But given the numbers, it is probably not necessary to toy with such forces (we can fix a few thousand by hand once; new entries should be low in numbers). On Mon, Apr 13, 2015 at 2:22 PM Andrew Gray andrew.g...@dunelm.org.uk wrote: This is an example of a more general problem, I think - country is treated as an indefinite concept, which breaks down for historic people as well. To take Magnus's example, Wikidata records that Henry VIII was a citizen of the UK, which would no doubt have surprised him (https://www.wikidata.org/wiki/Q38370). Perhaps what we want is to figure out some way that country (P17) and citizenship (P27) can have robust constraints based on date of birth/death or on date of an event, so that - for example - anyone who is reported as having citizenship of the UK has to have been born before or died after 1707. For something like the battle, the constraint would be that the event has to have happened while the P17 country was in existence. I don't know if we can do anything this sophisticated with the current constraints system - perhaps it would have to be organised on a country-by-country basis, one report for the UK, then the USSR, and so on as we define the cases. Perhaps something to look at doing a year down the line, when we've imported a lot of data we can fix ;-) Andrew. On 13 April 2015 at 13:00, Gerard Meijssen gerard.meijs...@gmail.com wrote: Hoi, The point is very much that the battle WAS in the USSR. It is not not applicable it is one of the most important battles in the second world war. My point is that we should not forget this. The battle of Uhud was not in Saudi Arabia either... Thanks, GerardM On 13 April 2015 at 12:10, Cristian Consonni kikkocrist...@gmail.com wrote: 2015-04-09 8:29 GMT+02:00 Gerard Meijssen gerard.meijs...@gmail.com: Because the battle of Stalingrad as a battle was not fought by modern day Russia, it was fought by the USSR and Nazi Germany. Associating the battle of Stalingrad with modern day Russia is wrong on so many levels. At the time it was Stalingrad, hence the name. It will never be the battle of Wolgograd. I believe that you should have a Not applicable button to click for these cases. C ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l -- - Andrew Gray andrew.g...@dunelm.org.uk ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
2015-04-13 18:46 GMT+02:00 Magnus Manske magnusman...@googlemail.com: So I present: https://tools.wmflabs.org/wikidata-todo/wrong_nationality.html All links to Wikidata are missing the /wiki/ part. C ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
Hoi, Because the battle of Stalingrad as a battle was not fought by modern day Russia, it was fought by the USSR and Nazi Germany. Associating the battle of Stalingrad with modern day Russia is wrong on so many levels. At the time it was Stalingrad, hence the name. It will never be the battle of Wolgograd. Thanks, GerardM On 7 April 2015 at 23:27, Roland Cornelissen metamatter...@gmail.com wrote: Hi, On 05-04-15 16:10, Gerard Meijssen wrote: It does not make sense to couple coordinates with countries... The battle of Stalingrad for instance was firmly in the USSR and not in modern day Russia. Why not? Every Place has a history that can't be evaded. The beautiful part of Linked Data is these little graphs one gets when joining data on defined properties. For instance joining data based on coordinates would provide an insight into the history of a certain Place. The battle of Stalingrad is an event that has taken Place at Stalingrad, which is now Wolgograd, both have the same coordinates. There are even more places at these coordinates :-) Cheers, Roland ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
Hi, On 05-04-15 16:10, Gerard Meijssen wrote: It does not make sense to couple coordinates with countries... The battle of Stalingrad for instance was firmly in the USSR and not in modern day Russia. Why not? Every Place has a history that can't be evaded. The beautiful part of Linked Data is these little graphs one gets when joining data on defined properties. For instance joining data based on coordinates would provide an insight into the history of a certain Place. The battle of Stalingrad is an event that has taken Place at Stalingrad, which is now Wolgograd, both have the same coordinates. There are even more places at these coordinates :-) Cheers, Roland ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
Hi, On 05-04-15 16:10, Gerard Meijssen wrote: It does not make sense to couple coordinates with countries... The battle of Stalingrad for instance was firmly in the USSR and not in modern day Russia. Why not? Every Place has a history that can't be evaded. The beautiful part of Linked Data is these little graphs one gets when joining data on defined properties. For instance joining data based on coordinates would provide an insight into the history of a certain Place. The battle of Stalingrad is an event that has taken Place at Stalingrad, which is now Wolgograd, both have the same coordinates. There are even more places at these coordinates :-) Cheers, Roland ___ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Re: [Wikidata-l] Suggestions for improvements of Wikidata
As an example of a step towards E1 (outside Wikidata), the list might
remember the code-snippet that the DJ wrote for Commons,
https://commons.wikimedia.org/wiki/User:TheDJ/wdcat.js
If you add
importScript('User:TheDJ/wdcat.js');
to your common.js file on Commons, then whenever you go to a Commons
category that is the target of a P373 on Wikidata, it adds a link to the
page that goes to Reasonator for the corresponding article-like Wikidata
item.
It's something I've found very useful, eg working through the BBC
YourPaintings list of painters on Mix'n'Match, and the corresponding
tracking pages linked from
https://en.wikipedia.org/wiki/Wikipedia:GLAM/Your_paintings/header
to see what article-like Wikidata item (if any) a Commons category may
relate to.
-- James.
On 02/04/2015 14:41, Romaine Wiki wrote:
[snip]
*E. Outside Wikidata*
E1. Commons: show on Commons somewhere when a category, gallery page,
institution page, template, file etc is used in a statement on Wikidata. If
a page is renamed or deleted, this must be changed on Wikidata as well, but
noticing where a page is used is not easy.
If an image is linked in a statement on Wikidata, on the image page this is
shown. Somehow this should also be implemented for categories, gallery
pages, institution pages, templates, and others.
This should be added to pages like Special:WhatLinksHere, Special:MovePage,
Special:GlobalUsage
E2. Wikipedia/other wikis: develop an extension, that communities can
enable, that shows on the bottom of articles, in the style of the category
box, an automatic box with all the identifiers used for authority control
https://www.wikidata.org/wiki/Q18614948 to replace templates like
https://www.wikidata.org/wiki/Q5153934.
___
Wikidata-l mailing list
Wikidata-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata-l
___
Wikidata-l mailing list
Wikidata-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata-l
