2015-03-10 15:08 GMT+01:00 Daniel Kinzler daniel.kinzler@wikimedia.de:

Am 10.03.2015 um 14:31 schrieb Amir E. Aharoni:

...

If it's way too much, how can this be bridged, or federated, or whatever the current popular word is? I don't even know exactly how does OSM store labels and translations now, but it sounds like another instance of Wikibase, if not Wikidata itself, can be used for it.

A separate Wikibase instance, federated with Wikidata sounds ideal (but the federation bit will need work). Scalability is an issue though. Wikibase is designed to scale to a few hundred million items. I suspect that OSM would need a couple of orders of magnitude more...

A separate Wikibase instance that could serve both OSM and Wikivoyage, maybe? Might be useful for both to keep the data about shops, restaurants, hotels, commercial whatelses and whatnots in the same instance, but keeping it separated by the "official" Wikidata...

2015-03-10 16:28 GMT+01:00 Janko Mihelić janjko@gmail.com:

What would this new Wikibase have that OpenStreetMap doesn't already have?

The possibility of "talking" with WMF projects, as Wikidata "talks" with all the other projects...

2015-03-10 16:46 GMT+01:00 Daniel Kinzler daniel.kinzler@wikimedia.de:

Am 10.03.2015 um 16:32 schrieb Yaroslav M. Blanter:

...

Hi Amir,

anything which can be remotely considered as a tourist attraction, as well as shops, hotels, reataurants and such are withing the scope of Wikivoyage and thus of Wikidata. For streets, we have now an approved bot task adding all Dutch streets on Wikidata, and I do not see why any other country could be different - provided we have good sources.

I fear doing this is going to kill Wikidata. Neither the software nor the community scales to managing entries for every street in the world.

...and basically avoiding this for Wikidata.

2015-03-10 17:10 GMT+01:00 Daniel Kinzler daniel.kinzler@wikimedia.de:

Am 10.03.2015 um 16:54 schrieb Luca Martinelli:

...

2015-03-10 16:28 GMT+01:00 Janko Mihelić janjko@gmail.com:

...

What would this new Wikibase have that OpenStreetMap doesn't already have?

The possibility of "talking" with WMF projects, as Wikidata "talks" with all the other projects...

Only if it's also hosted on the WMF cluster. Or we implement http based federation (planned, but a lot of work, and waaay down there on the prio list).

Of course, it was just an idea.

BTW, maybe related: https://meta.wikimedia.org/wiki/Grants:IdeaLab/OSMdata:_a_Wikidata-like_edit...

L.

Think... BIGGER.

Jo has the right idea... Linked Data.

It sounds to me like the right way forward for domain specific interest data (like OSM) ...is,

Instead of 1 source of data (Wikidata)...and throwing domain specific interest data into Wikidata (not all data needs to live inside it).

Just use >1 source of data, i.e. Linked Data.

Wikidata does not have to be the 1 sole source... and besides, I am sure that eventually someone could just allow federated queries / joins to Linked Data.

Perhaps Wikidata can try to be and eventually help build... the 1 sole QUERY source for Linked Data. :)

(Select OSM_ROADS contained_in "Ardennes")

Thad +ThadGuidry https://www.google.com/+ThadGuidry

What about bus stops?

Jo

2015-03-10 15:37 GMT+01:00 aude aude.wiki@gmail.com:

On Tue, Mar 10, 2015 at 2:31 PM, Amir E. Aharoni < amir.aharoni@mail.huji.ac.il> wrote:

...

Hi,

[ Aude and Christian Consonni, this should especially interest you. ]

I was throwing around ideas with a friend about how OpenStreetMap could be integrated with Wikidata.

The thing that I care the most in any software is internationalization. Having a map in which all labels of towns, streets and everything else is translated to all languages sounds like a super-wonderful thing.

Wikidata allows labeling everything, translating everything, and attaching properties to everything, so it sounds like it could be a good match.

But then the question of "what IS everything" came up. Wikidata was created mostly with Wikipedia in mind, so Wikipedia's notability policies influenced Wikidata. Roughly, Wikidata has items for every thing about which there is, or can be, a Wikipedia article and for things that are useful, or if it "fulfills some structural need https://www.wikidata.org/wiki/Wikidata:Notability".

Towns obviously have or can a Wikipedia article about them, but probably not every street or shop.

At minimum, supporting Wikivoyage is within scope of Wikidata, so imho items for shops would be ok. (at least the ones that would be used in Wikivoyage)

I am not sure about streets... maybe, it depends and is up to the community (what extent we want them) and scalability (technical and community) might be considerations. As well, in Wikidata, we want references.

For the OSM use case, for streets, I think the osm name tags are sufficient. Transliteration of street names definitely makes sense, but maybe can be autogenerated?

In some cases multilingual (e.g. in Brussels), actual translations might be desired, and think osm supports that adequately.

Katie

...

But do they fulfill a structural need or is it way too much?

If it's way too much, how can this be bridged, or federated, or whatever the current popular word is? I don't even know exactly how does OSM store labels and translations now, but it sounds like another instance of Wikibase, if not Wikidata itself, can be used for it.

I don't have much to add, but I'd love to hear ideas from people who do (again, Aude and Christian Consonni, I'm looking at you :) ).

-- Amir Elisha Aharoni · אָמִיר אֱלִישָׁע אַהֲרוֹנִי http://aharoni.wordpress.com ‪“We're living in pieces, I want to live in peace.” – T. Moore‬

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

-- @wikimediadc / @wikidata

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

I am an avid OpenStreetMap user, and I like the Wikidata project very much. That's why I proposed a OSM tag wikidata=* and I wasn't the only one to think of this[1].

Making a wikidata element for every or some OSM elements is not quite possible, at least for now. OSM objects don't have constant IDs like Wikidata elements do. If you delete a city node, and make a new one with all the same tags, nobody cares. That's why connecting Wikidata and OSM can be made only one way, from OSM to Wikidata. The other direction can be made only with querys like "a city within 10km from this coordinate", or "a city with this name within borders of a country with this name". A project "Permanent ID" is trying to do this[2]

First use of Wikidata in OpenStreetMap is a 1 to 1 mapping. A town element in Wikidata is a place=town node in OSM. There can be only one in either database. We tag this in OSM with a simple wikidata=Q1435.

But then there are more interesting uses, for example "artist of this sculpture is Q313828", "architect of this building is Q12633257", "this street is named after Q9036". I think this is the best marriage of both projects so far. For example in OSM we can see all sculptures made by a specific artist:

http://overpass-turbo.eu/s/86X

It would be great to see this information in Wikipedia.

[1] http://wiki.openstreetmap.org/wiki/Proposed_features/Wikidata [2] http://wiki.openstreetmap.org/wiki/Overpass_API/Permanent_ID

2015-03-10 15:37 GMT+01:00 aude aude.wiki@gmail.com:

On Tue, Mar 10, 2015 at 2:31 PM, Amir E. Aharoni < amir.aharoni@mail.huji.ac.il> wrote:

...

Hi,

[ Aude and Christian Consonni, this should especially interest you. ]

I was throwing around ideas with a friend about how OpenStreetMap could be integrated with Wikidata.

The thing that I care the most in any software is internationalization. Having a map in which all labels of towns, streets and everything else is translated to all languages sounds like a super-wonderful thing.

Wikidata allows labeling everything, translating everything, and attaching properties to everything, so it sounds like it could be a good match.

But then the question of "what IS everything" came up. Wikidata was created mostly with Wikipedia in mind, so Wikipedia's notability policies influenced Wikidata. Roughly, Wikidata has items for every thing about which there is, or can be, a Wikipedia article and for things that are useful, or if it "fulfills some structural need https://www.wikidata.org/wiki/Wikidata:Notability".

Towns obviously have or can a Wikipedia article about them, but probably not every street or shop.

At minimum, supporting Wikivoyage is within scope of Wikidata, so imho items for shops would be ok. (at least the ones that would be used in Wikivoyage)

I am not sure about streets... maybe, it depends and is up to the community (what extent we want them) and scalability (technical and community) might be considerations. As well, in Wikidata, we want references.

For the OSM use case, for streets, I think the osm name tags are sufficient. Transliteration of street names definitely makes sense, but maybe can be autogenerated?

In some cases multilingual (e.g. in Brussels), actual translations might be desired, and think osm supports that adequately.

Katie

...

But do they fulfill a structural need or is it way too much?

If it's way too much, how can this be bridged, or federated, or whatever the current popular word is? I don't even know exactly how does OSM store labels and translations now, but it sounds like another instance of Wikibase, if not Wikidata itself, can be used for it.

I don't have much to add, but I'd love to hear ideas from people who do (again, Aude and Christian Consonni, I'm looking at you :) ).

-- Amir Elisha Aharoni · אָמִיר אֱלִישָׁע אַהֲרוֹנִי http://aharoni.wordpress.com ‪“We're living in pieces, I want to live in peace.” – T. Moore‬

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

-- @wikimediadc / @wikidata

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

I helped with the Lighthouses schema in Freebase.

Some of which is based on List of Lights (NGA) USA.

I have DB conversion data for the PDFs...just never got around to loading them all in.

Let me know if I can help.

Thad +ThadGuidry https://www.google.com/+ThadGuidry

On Tue, Mar 10, 2015 at 10:19 AM, Markus Bärlocher < markus.baerlocher@lau-net.de> wrote:

It would be fine to have all this together:

_OSM_ seamark:type=light*

_Wikidata_ https://www.wikidata.org/wiki/Q44782

_Wikipedia_ https://de.wikipedia.org/wiki/Kategorie:Liste_(Leuchtt%C3%BCrme) https://de.wikipedia.org/wiki/Liste_von_Leuchtt%C3%BCrmen

_Commons_ commons:Lighthouse category:Lighthouses

_Example_ Wikipedia: https://de.wikipedia.org/wiki/Roter_Sand Commons: https://commons.wikimedia.org/wiki/Category:Leuchtturm_Roter_Sand Wikidata: https://www.wikidata.org/wiki/Q220034 Resonator: http://tools.wmflabs.org/reasonator/?q=Q220034 OSM-DB: https://www.openstreetmap.org/node/635484478 OpenSeaMap: http://map.openseamap.org/?zoom=13&mlat=53.85557&mlon=8. 08198&mtext=Roter%20Sand&layers=BFTFFFTFFTT0FFFFFTFF

Best regards, Markus

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

Markus,

Here's the basic schema that we put together in Freebase that is semi-populated with data now.

Whomever whats to build the structure for Wikidata can just setup like it:

http://www.freebase.com/architecture/lighthouse?schema=&lang=en

Thad +ThadGuidry https://www.google.com/+ThadGuidry

On 10.03.2015 17:31, Thad Guidry wrote:

I helped with the Lighthouses schema in Freebase.

For your personal enjoyment:

https://tools.wmflabs.org/wikidata-exports/miga/?lighthouses (use Google Chrome or any other non-IE, non-FF browser to view).

Just a quick hack, many of the data are incomplete (no English labels or images, for example), but the map is pretty informative. For example, you can see that Portugal has a lot of lighthouses while Spain has almost none -- maybe we need to look at our data there ;-) The data is not auto-updated (e.g. I already fixed the issue with the lighthouse in rural Kasachstan on Wikidata -- a typo that changed W for E in coordinates). But it is true that we have a lighthouse close to Dresden (far from the sea). Anyway, maps are a great way to spot data quality issues.

Cheers

Markus

On 11.03.2015 05:40, Tom Morris wrote:

On Tue, Mar 10, 2015 at 6:41 PM, Markus Krötzsch <markus@semantic-mediawiki.org mailto:markus@semantic-mediawiki.org> wrote:

For example, you can see that Portugal has a lot of lighthouses
while Spain has almost none -- maybe we need to look at our data
there ;-)

Perhaps it's a language confusion issue, but does Spain really have few lighthouses? That would seem VERY unusual for a territory with an extensive coastline.

No, you are right: this is of course an issue in the completeness of our data. If you zoom in to Europe, you can see that some countries have costs full of lighthouses, while others seem to lack them almost completely. I think it clearly shows that a lot of our data comes from Wikipedias (in some specific language).

Where does Wikidata sit in that mix? How does it compare with DBpedia, Freebase, Wikipedia, or even *real* data sources like official governmental lists of navigational aids for mariners at sea?

Good question. I guess that we could have a much better coverage for some of the obvious holes in our data, e.g., by adding classificaiton information based on Spanish Wikipedia categories.

Independent of where it sits now, what where does it aspire to sit?

The uncontested target would be:

* Every lighthouse that is found in any Wikipedia (hence in Wikidata) should be "instance of lighthouse" and have coordinates and country defined.

A possible target to discuss would be to add:

* Every lighthouse should be in Wikidata.

Whether or not this makes sense depends on how many lighthouses there really are. Maybe we are not so far from completeness. Lighthouses are prominent landmarks of historic, nautic, and touristic interest, so should be valid Wikipedia topics anyway. Moreover, they tend to change very little over time, so data maintenance is relatively easy.

Note that this is quite different from streets, which change names all the time, get created, demolished, and merged, and are much larger in number. From these properties, I think OSM is much better suited for managing this data for now.

In view of the current developments towards Wikidata query support, a tangible goal would be to set up integrated query services that provide a joint view of the data from OSM and Wikidata without physically moving large quantities of data from one to the other.

Markus

Openstreetmap has many lighthouses on the Spanish Coast:

http://overpass-turbo.eu/s/87R

You can change the bbox and investigate the other coast lines easily with that query.

Jo

2015-03-11 10:37 GMT+01:00 Magnus Manske magnusmanske@googlemail.com:

On Wed, Mar 11, 2015 at 9:24 AM Markus Krötzsch < markus@semantic-mediawiki.org> wrote:

...

No, you are right: this is of course an issue in the completeness of our data. If you zoom in to Europe, you can see that some countries have costs full of lighthouses, while others seem to lack them almost completely. I think it clearly shows that a lot of our data comes from Wikipedias (in some specific language).

In this instance, the issue appears to be that the existing lists on

Wikipedia have not been touched, e.g.: https://en.wikipedia.org/wiki/List_of_lighthouses_in_Spain

These, including redlinks, could be imported into Wikidata rather easily. Some already have images.

Ideally, we'd want some "official" (e.g. national, UN) source to cross-check.

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

Sorry the tags was wrong... Right is:

_in OpenSeaMap_ you can find 40'000 Lights by a query on: seamark:type=light_float seamark:type=light_vessel seamark:type=light_minor seamark:type=light_major seamark:type=light

_in WP for Spain_ http://en.wikipedia.org/wiki/de:Liste_von_Leuchtt%C3%BCrmen_in_Spanien

_in WD for world_ 2135 lights today List: http://tools.wmflabs.org/autolist/autolist1.html?q=CLAIM%5B31:39715] Chart: http://tools.wmflabs.org/autolist/autolist1.html?q=CLAIM%5B31:39715]

best regards, Markus

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

Note that OSM is likely to be unusable for copyright reasons

-- J.

On 11/03/2015 11:20, Magnus Manske wrote:

Started scraping Wikipedia lists: https://en.wikipedia.org/wiki/List_of_lighthouses_and_lightvessels

Got to Canada: https://docs.google.com/spreadsheets/d/12UaLKolLl5hbOFP9vztShQsiKkr5AZo_TbNr...

Continue this (someone help!), or start anew from another source?

On Wed, Mar 11, 2015 at 10:59 AM Markus Bärlocher < markus.baerlocher@lau-net.de> wrote:

...

Sorry the tags was wrong... Right is:

_in OpenSeaMap_ you can find 40'000 Lights by a query on: seamark:type=light_float seamark:type=light_vessel seamark:type=light_minor seamark:type=light_major seamark:type=light

...

_in WP for Spain_ http://en.wikipedia.org/wiki/de:Liste_von_Leuchtt%C3%BCrmen_in_Spanien

_in WD for world_ 2135 lights today List: http://tools.wmflabs.org/autolist/autolist1.html?q=CLAIM%5B31:39715] Chart: http://tools.wmflabs.org/autolist/autolist1.html?q=

CLAIM[31:39715]

...

best regards, Markus

---

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

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

Am 11.03.2015 um 12:39 schrieb Jo Walsh:

http://wiki.openstreetmap.org/wiki/Wikidata#Importing_data *Copying data to Wikidata from OSM* (or even from other Wikimedia projects) *is not allowed* because Wikidata uses the public-domain style Creative Commons CC0 license which does not contain any attribution or share-alike provisions. Conversely, data may be copied from Wikidata without restriction.

(OSM is licensed under the Open Database License)

This applies to any copyrightable material. Facts ("this is a lighthouse") are not copyrightable.

On 03/11/2015 12:58 PM, Daniel Kinzler wrote:

Am 11.03.2015 um 12:39 schrieb Jo Walsh:

...

http://wiki.openstreetmap.org/wiki/Wikidata#Importing_data *Copying data to Wikidata from OSM* (or even from other Wikimedia projects) *is not allowed* because Wikidata uses the public-domain style Creative Commons CC0 license which does not contain any attribution or share-alike provisions. Conversely, data may be copied from Wikidata without restriction.

(OSM is licensed under the Open Database License)

This applies to any copyrightable material. Facts ("this is a lighthouse") are not copyrightable.

Facts are not copyrightable, but in EU a lot of facts are when they are assembled in a database due to sui generis database right.

I suppose that OSM falls under UK law and thereby EU sui generis database right, so systematic extraction of OSM data into Wikidata will constitute a violation in EU of the Open Database License terms.

https://en.wikipedia.org/wiki/Sui_generis_database_right

/Finn

On 11 March 2015 at 09:23, Markus Krötzsch markus@semantic-mediawiki.org wrote:

...

Independent of where it sits now, what where does it aspire to sit?

The uncontested target would be:

• Every lighthouse that is found in any Wikipedia (hence in Wikidata) should

be "instance of lighthouse" and have coordinates and country defined.

A possible target to discuss would be to add:

• Every lighthouse should be in Wikidata.

Whether or not this makes sense depends on how many lighthouses there really are. Maybe we are not so far from completeness. Lighthouses are prominent landmarks of historic, nautic, and touristic interest, so should be valid Wikipedia topics anyway. Moreover, they tend to change very little over time, so data maintenance is relatively easy.

Note that this is quite different from streets, which change names all the time, get created, demolished, and merged, and are much larger in number. From these properties, I think OSM is much better suited for managing this data for now.

I wonder if it would be useful to have a centralised list of "classes of things in Wikidata". For example:

Things entirely in Wikidata

* MEPs * County-level administrative divisions of all countries * All artworks by the following people (list) * Cultural heritage sites in the following countries (list) * All people listed in the following biographical databases (list) * (etc)

Things not yet entirely in Wikidata (but probably will be eventually)

* All national-level elected representatives * All species * Lighthouses * All artworks by the following people (list) * Cultural heritage sites in the following countries (list) * All people listed in the following biographical databases (list)

Things which will never be complete in Wikidata

* All local politicians * Streets worldwide * All businesses

This would be a very useful adjunct to the notability page, as it would give concrete examples to work from for the sort of things we feel are appropriate.

I can offer https://tools.wmflabs.org/mix-n-match/

This (or a future successor) could serve as the "list keeper" for, say, list of lighthouses, or graded buildings in the UK.

One thing that IMHO would be required for this to work would be a (semi-)automatic sync of the "list keeper" with authoritative external lists (unless we want to sync those to Wikidata directly).

On Wed, Mar 11, 2015 at 2:09 PM Markus Krötzsch < markus@semantic-mediawiki.org> wrote:

Hi Andrew,

This is a great idea! It would help data consumers to know what to expect and community members to know what to put in (or where help with imports would be appreciated). Moreover, the discussion about this list would be a great way to structure our work in general (have documented discussions about our goals for certain types of data). I feel that the bot right approval process is not the best place to decide if we strive to have all streets or all lighthouses in.

For things that are not complete in Wikidata (yet or ever), it would further help to provide pointers to other, more complete data sources (and the properties we might have to link to them).

The question is how to best organise this list. Your initial example setup already shows that this tends to become very diverse (not to say: chaotic). One could link this from the related class items (e.g., lighthouses or paintings), but having this as another extra load on the talk page would maybe not so ideal either. After all, this could be one of the first things that newbies to Wikidata want to get an idea about.

Cheers,

Markus

On 11.03.2015 14:07, Andrew Gray wrote: ...

...

I wonder if it would be useful to have a centralised list of "classes of things in Wikidata". For example:

Things entirely in Wikidata

• MEPs
• County-level administrative divisions of all countries
• All artworks by the following people (list)
• Cultural heritage sites in the following countries (list)
• All people listed in the following biographical databases (list)
• (etc)

Things not yet entirely in Wikidata (but probably will be eventually)

• All national-level elected representatives
• All species
• Lighthouses
• All artworks by the following people (list)
• Cultural heritage sites in the following countries (list)
• All people listed in the following biographical databases (list)

Things which will never be complete in Wikidata

• All local politicians
• Streets worldwide
• All businesses

This would be a very useful adjunct to the notability page, as it would give concrete examples to work from for the sort of things we feel are appropriate.

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

Thanks both. I'll try hacking out an aggregated list tonight and send the link around, but for now if there's anything you know for sure we cover comprehensively, fire away with your suggestions :-)

Mix-and-match was in the back of my head, as it has a nice example of all three groups:

* Already completely included - MEPs (at least until the next election), ODNB

* Going to be included but not there yet - Dictionary of Welsh Biography

* Will never be completely included - ACAD

Andrew.

On 11 March 2015 at 14:31, Magnus Manske magnusmanske@googlemail.com wrote:

I can offer https://tools.wmflabs.org/mix-n-match/

This (or a future successor) could serve as the "list keeper" for, say, list of lighthouses, or graded buildings in the UK.

One thing that IMHO would be required for this to work would be a (semi-)automatic sync of the "list keeper" with authoritative external lists (unless we want to sync those to Wikidata directly).

On Wed, Mar 11, 2015 at 2:09 PM Markus Krötzsch markus@semantic-mediawiki.org wrote:

...

Hi Andrew,

This is a great idea! It would help data consumers to know what to expect and community members to know what to put in (or where help with imports would be appreciated). Moreover, the discussion about this list would be a great way to structure our work in general (have documented discussions about our goals for certain types of data). I feel that the bot right approval process is not the best place to decide if we strive to have all streets or all lighthouses in.

For things that are not complete in Wikidata (yet or ever), it would further help to provide pointers to other, more complete data sources (and the properties we might have to link to them).

The question is how to best organise this list. Your initial example setup already shows that this tends to become very diverse (not to say: chaotic). One could link this from the related class items (e.g., lighthouses or paintings), but having this as another extra load on the talk page would maybe not so ideal either. After all, this could be one of the first things that newbies to Wikidata want to get an idea about.

Cheers,

Markus

On 11.03.2015 14:07, Andrew Gray wrote: ...

...

I wonder if it would be useful to have a centralised list of "classes of things in Wikidata". For example:

Things entirely in Wikidata

• MEPs
• County-level administrative divisions of all countries
• All artworks by the following people (list)
• Cultural heritage sites in the following countries (list)
• All people listed in the following biographical databases (list)
• (etc)

Things not yet entirely in Wikidata (but probably will be eventually)

• All national-level elected representatives
• All species
• Lighthouses
• All artworks by the following people (list)
• Cultural heritage sites in the following countries (list)
• All people listed in the following biographical databases (list)

Things which will never be complete in Wikidata

• All local politicians
• Streets worldwide
• All businesses

This would be a very useful adjunct to the notability page, as it would give concrete examples to work from for the sort of things we feel are appropriate.

---

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

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1

I'm a bit confused as to why some of entries are in the "never" list.

For example, there are published lists of musical artists (roughly defined as people (or groups?) who have made a commercial recording). This would seem to make then notable under criteria 2. in https://www.wikidata.org/wiki/Wikidata:Notability. Why then are musicians in the "never" list?

I can understand that these large groups might have a "Not right now" status, but that seems to be different from "never". What am I misunderstanding?

peter

On 03/11/2015 01:28 PM, Andrew Gray wrote:

Here's a very rough first attempt:

https://www.wikidata.org/wiki/Wikidata:Comprehensive_groups_of_items

The main problem is that other than the areas I've been working in, I really don't know what's out there :-). Does the Wiki Loves Monuments work mean that we have complete monument coverage in certain countries, for example?

Andrew.

On 11 March 2015 at 15:01, Andrew Gray andrew.gray@dunelm.org.uk wrote:

...

Thanks both. I'll try hacking out an aggregated list tonight and send the link around, but for now if there's anything you know for sure we cover comprehensively, fire away with your suggestions :-)

Mix-and-match was in the back of my head, as it has a nice example of all three groups:

• Already completely included - MEPs (at least until the next

election), ODNB

• Going to be included but not there yet - Dictionary of Welsh

Biography

• Will never be completely included - ACAD

Andrew.

On 11 March 2015 at 14:31, Magnus Manske magnusmanske@googlemail.com wrote:

...

I can offer https://tools.wmflabs.org/mix-n-match/

This (or a future successor) could serve as the "list keeper" for, say, list of lighthouses, or graded buildings in the UK.

One thing that IMHO would be required for this to work would be a (semi-)automatic sync of the "list keeper" with authoritative external lists (unless we want to sync those to Wikidata directly).

On Wed, Mar 11, 2015 at 2:09 PM Markus Krötzsch markus@semantic-mediawiki.org wrote:

...

Hi Andrew,

This is a great idea! It would help data consumers to know what to expect and community members to know what to put in (or where help with imports would be appreciated). Moreover, the discussion about this list would be a great way to structure our work in general (have documented discussions about our goals for certain types of data). I feel that the bot right approval process is not the best place to decide if we strive to have all streets or all lighthouses in.

For things that are not complete in Wikidata (yet or ever), it would further help to provide pointers to other, more complete data sources (and the properties we might have to link to them).

The question is how to best organise this list. Your initial example setup already shows that this tends to become very diverse (not to say: chaotic). One could link this from the related class items (e.g., lighthouses or paintings), but having this as another extra load on the talk page would maybe not so ideal either. After all, this could be one of the first things that newbies to Wikidata want to get an idea about.

Cheers,

Markus

On 11.03.2015 14:07, Andrew Gray wrote: ...

...

I wonder if it would be useful to have a centralised list of "classes of things in Wikidata". For example:

Things entirely in Wikidata

• MEPs * County-level administrative divisions of all countries *

All artworks by the following people (list) * Cultural heritage sites in the following countries (list) * All people listed in the following biographical databases (list) * (etc)

Things not yet entirely in Wikidata (but probably will be eventually)

• All national-level elected representatives * All species *

Lighthouses * All artworks by the following people (list) * Cultural heritage sites in the following countries (list) * All people listed in the following biographical databases (list)

Things which will never be complete in Wikidata

• All local politicians * Streets worldwide * All businesses

This would be a very useful adjunct to the notability page, as it would give concrete examples to work from for the sort of things we feel are appropriate.

_______________________________________________ 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.gray@dunelm.org.uk

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1

But then just what is the current scope of the project? I, too, think that the scope of Wikidata is at least provide support for things that could be on Wikipedia. Perhaps the current scope of Wikidata then should not be "every demo tape ever produced", but it seems to me that "every generally available recording" (perhaps something like every recording in MusicBrainz) is in scope right now.

peter

On 03/11/2015 04:48 PM, Magnus Manske wrote:

On Wed, Mar 11, 2015 at 8:52 PM Peter F. Patel-Schneider <pfpschneider@gmail.com mailto:pfpschneider@gmail.com> wrote:

I can understand that these large groups might have a "Not right now" status, but that seems to be different from "never". What am I misunderstanding?

I'd say it's "never within the current scope of the project".

I remember when Wikipedia had a two-digit article count. The plan was 100K articles. We all thought "yeah, right", but went to work anyway. en.wp is ~4.7M articles at the moment, and many topics we would never have imagined are covered.

So right now, the scope of Wikidata is roughly "what could be on Wikipedia, plus glue". A musician would not be notable if all he did was send off a single tape to a record company; neither would a self-published e-book with a readership of five.

But in ten years time, maybe we get auto-generated and seeded items from every publisher or national library for every new book. Maybe we've gone from 14M items to 700M (the same 50x factor as en.wp), and then these "never" groups will be "on their way", and it's "never" for astronomical objects and molecules (because for those, 700M won't even scratch the surface).

_

Thad Guidry thadguidry@gmail.com wrote:

I helped with the Lighthouses schema in Freebase.

Some of which is based on List of Lights (NGA) USA.

I have DB conversion data for the PDFs...just never got around to loading them all in.

Let me know if I can help.

I was able to match 407 lighthouses on OSM and Wikidata:

http://edwardbetts.com/osm-wikidata/2015-04-18/match_results/Lighthouses.htm...

Tom,

Split tool to the rescue ? :)

Looks like the GeoNames ID is for the islet itself...so probably needs another entity for the Vinga Lighthouse itself sitting on that islet.

Thad +ThadGuidry https://www.google.com/+ThadGuidry

On Thu, Apr 23, 2015 at 1:02 PM, Tom Morris tfmorris@gmail.com wrote:

On Thu, Apr 23, 2015 at 10:16 AM, Edward Betts edward@4angle.com wrote:

...

Thad Guidry thadguidry@gmail.com wrote:

...

I helped with the Lighthouses schema in Freebase.

Some of which is based on List of Lights (NGA) USA.

I have DB conversion data for the PDFs...just never got around to

loading

...

them all in.

Let me know if I can help.

I was able to match 407 lighthouses on OSM and Wikidata:

http://edwardbetts.com/osm-wikidata/2015-04-18/match_results/Lighthouses.htm...

How can something be both a islet, a nature preserve, and a lighthouse?

http://www.wikidata.org/wiki/Q3372089

Tom

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

James Heald j.heald@ucl.ac.uk wrote:

The post last autumn said you had 176,000 items matched to OSM, a very large proportion of which (at least at that time) had no P31 "instance of", but you had been able to deduce this from Wikipedia.

Would it be possible to start loading Wikidata with that information -- ie complete the P31's that were missing?

The code that generates the mappings uses Wikipedia categories, and it makes some mistakes. Take beaches for example:

http://edwardbetts.com/osm-wikidata/2015-04-18/match_results/Beaches.html

Some of the Wikidata items are about a beach, some about a settlement next to a beach. All of them are in within a beaches category on Wikipedia. It would be wrong to say that a settle with a beach is an instance of a beach.

For this reason it probably doesn't make sense to add 'instance of' statements automatically.

On 2015-03-10 16:46, Daniel Kinzler wrote:

Am 10.03.2015 um 16:32 schrieb Yaroslav M. Blanter:

...

Hi Amir,

anything which can be remotely considered as a tourist attraction, as well as shops, hotels, reataurants and such are withing the scope of Wikivoyage and thus of Wikidata. For streets, we have now an approved bot task adding all Dutch streets on Wikidata, and I do not see why any other country could be different - provided we have good sources.

I fear doing this is going to kill Wikidata. Neither the software nor the community scales to managing entries for every street in the world.

We have an ambition to have a separate entry for every species of insects, and the number of insect species is comparable to the number of streets.

(I personally would not add streets, but if there is a good database under an appropriate license, and an enthusiastic bot owner - why not?)

Cheers Yaroslav

On 10 March 2015 at 15:32, Yaroslav M. Blanter putevod@mccme.ru wrote:

For streets, we have now an approved bot task adding all Dutch streets on Wikidata

Approved by whom? I foresee the majority of those being deleted as non- notable.

On 10 March 2015 at 20:20, Yaroslav M. Blanter putevod@mccme.ru wrote:

On 2015-03-10 20:47, Andy Mabbett wrote:

...

On 10 March 2015 at 15:32, Yaroslav M. Blanter putevod@mccme.ru wrote:

...

For streets, we have now an approved bot task adding all Dutch streets on Wikidata

Approved by whom? I foresee the majority of those being deleted as non- notable.

By me as a crat.

You mean this:

https://www.wikidata.org/wiki/Wikidata:Requests_for_permissions/Bot/RobotMic...

which was approved after two days, and with only comments - not a !vote of support - from just one editor other than the proposer?

That's not much discussion, much less evidence of consensus for a 240,000-item import and a precedent for several magnitudes more than that.

In the future, if you want your opinion on bot tasks to be taken into account, please comment on bot permission requests. I usually let them pend for several days, which obviously can be extended if there are constructive comments.

If this is considered notable, I'd consider doing this for all streets in the Brussels region. I would add wikidata tags to all OSM objects involved (highway ways and associatedStreet relations) at the same time though. So if they'd get deleted once again on Wikidata, that would be quite a useless endeavour... and a lot of junk would remain behind in OSM. In fact, I hadn't realised wikidata items were prone to deletion once again. I thought of them as stable 'entities'.

Jo

2015-03-10 20:47 GMT+01:00 Andy Mabbett andy@pigsonthewing.org.uk:

On 10 March 2015 at 15:32, Yaroslav M. Blanter putevod@mccme.ru wrote:

...

For streets, we have now an approved bot task adding all Dutch streets on Wikidata

Approved by whom? I foresee the majority of those being deleted as non- notable.

-- Andy Mabbett @pigsonthewing http://pigsonthewing.org.uk

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

On 23 April 2015 at 15:20, Serge Wroclawski emacsen@gmail.com wrote:

I am not sure how I missed this discussion, but adding information from OSM into Wikidata en mass like this is a violation of the OSM license.

Nobody is prosing to add information from OSM into Wikidata en mass.

The best way to link objects from Wikidata and Openstreetmap is by adding wikidata tags to the OSM objects. That way those Q-numbers can follow along when an OSM object gets converted from a node to a way (POI to building for example) or from a way to a relation. (area to a multipolygon for example).

The best way to figure out which objects are connected is the Overpass API. It would be a good idea to have a poperty on the wikidata side where Overpass Queries can be stored.

Polyglot

2015-04-23 17:08 GMT+02:00 Peter F. Patel-Schneider pfpschneider@gmail.com :

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256

How so? Oh, because Wikidata is CC0 and the Open Street Map database is ODbL, which is somewhat like CC BY-SA. I don't think that that follows, though, as what is being put into Wikidata is contents, which appear to me to be covered under the DbCL, which is like CC0.

Peter F. Patel-Schneider, speaking as an individual

On 04/23/2015 07:20 AM, Serge Wroclawski wrote:

...

I am not sure how I missed this discussion, but adding information from OSM into Wikidata en mass like this is a violation of the OSM license.

• Serge

On Tue, Mar 10, 2015 at 11:32 AM, Yaroslav M. Blanter putevod@mccme.ru wrote:

...

On 2015-03-10 14:31, Amir E. Aharoni wrote:

...

Hi,

[ Aude and Christian Consonni, this should especially interest you. ]

I was throwing around ideas with a friend about how OpenStreetMap could be integrated with Wikidata.

...

...

Towns obviously have or can a Wikipedia article about them, but probably not every street or shop. But do they fulfill a structural need or is it way too much?

Hi Amir,

anything which can be remotely considered as a tourist attraction, as well as shops, hotels, reataurants and such are withing the scope of Wikivoyage and thus of Wikidata. For streets, we have now an approved bot task adding all Dutch streets on Wikidata, and I do not see why any other country could be different - provided we have good sources.

Cheers Yaroslav

_______________________________________________ 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

-----BEGIN PGP SIGNATURE----- Version: GnuPG v2

iQEcBAEBCAAGBQJVOQsGAAoJECjN6+QThfjziAYIAJ6uEmHK2HF9bAnTCo97wUf9 o1OPUCUfdaURiRCSTwZKHnxZVPjNS0pRKQMKzZS6+XMNVqy8jq5wCjDQsk7j+4cj LmRlkZohglxpn9eNqKvpO+5m7gl7Pj+w0eWzb09lE3irMyNN9+C5yBGxUaV+XCAB AIOqhPmBC5tJARHag0CLC5U0ocKb7C/eDX17f8NyeXFDhy34ejB1xywbtxbjAgxS eEp+l/KO3CkoKGabPQNDg9Cko/uFQlldPidvgFIHrfaMMdiccwwIMkKq0CLubvlL CT1J/y8/zd5cMTuiX6S1/HMhIPru+hZMsHlOe2Sx6XRH1MXARn0iX9WXzHqR7eA= =6EWJ -----END PGP SIGNATURE-----

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l

I'm under the impression people on this mailing list are more interested in this, than most of us mappers, so it belongs here as well. I agree with you as far as the import goes, if you can call it that. We should have way to easily add wikidata tags to OSM elements which already have wikipedia tags.

Knowing how the wikidata related information is stored in OSM, how to retrieve it and how to work with it, is something that definitely belongs on this list though.

Polyglot

2015-04-23 18:42 GMT+02:00 Serge Wroclawski emacsen@gmail.com:

I entirely agree and have been looking at a way we could start this in OSM in a systematic way.

There are a ton of things OSM could benefit from with this.

But this is, IMHO an OSM issue (an OSM import) and not something the Wikipedia folks here need to worry about so I think it belongs on an OSM mailing list.

• Serge

---

Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l