[Wikidata] Re: Wikidata Atlas: a geographic view of Wikidata entities [feedback welcome!]

[Shilad Sen]

Good resources, Isaac! For the third paper below, http://cartograph.info is old but still alive.On Dec 20, 2022, at 7:53 AM, Isaac Johnson  wrote:Daniel,A few other research projects that have used Wikipedia/Wikidata to generate "maps" of concepts that are not explicitly spatial and might be relevant to your thinking. I'm only aware of a live demo for the second one (Frankenplace) but the papers have good figures to illustrate:Hecht, Brent, et al. "Explanatory semantic relatedness and explicit spatialization for exploratory search." Proceedings of the 35th international ACM SIGIR conference on Research and development in information retrieval. 2012.Adams, Benjamin, Grant McKenzie, and Mark Gahegan. "Frankenplace: interactive thematic mapping for ad hoc exploratory search." Proceedings of the 24th international conference on world wide web. 2015.Sen, Shilad, et al. "Toward Universal Spatialization Through Wikipedia-Based Semantic Enhancement." ACM Transactions on Interactive Intelligent Systems (TiiS) 9.2-3 (2019): 1-29.Best,IsaacOn Mon, Dec 19, 2022 at 5:37 PM Daniel Mietchen via Wikidata  wrote:Dear Tassos,

thanks for the example - that map is interesting but still arranged in
terms of geocoordinates, and based on Wikipedia data.
What I had in mind is a map that positions Wikidata items generically
(i.e. without the need for geolocation statements via P625) but
somewhat reliably (for a given query and reasonably stable data) in a
2D or 3D space (perhaps even as a function of some additional
parameters) and then allows the user to zoom around inside this system
to explore spatial relationships just as they can explore geospatial
relationships in your gelocated wiki atlases.
The closest thing to this that I have at hand right now is
https://galaxy.opensyllabus.org/ , which clusters syllabi by topic and
allows zooming but is not based on Wikidata.
The Wikidata Query Service has some visualizations that do part of
that but these (i) do not provide zooming, (ii) often time out and
have other problems, e.g. (iii) no reliable position of a given node
or (iv) little to no meaning in adjacency.

Another thing relevant here are Wikidata maps as per
https://commons.wikimedia.org/wiki/Wikidata_map
with their bright and dark areas and in particular their evolution over time
All of these provide for fertile ground to engage relevant
communities, and It would be very helpful to have similar
visualizations (e.g. change as a function of some parameter) for any
part of Wikidata, including but not limited to geodata.

Best,

Daniel



On Sat, Dec 17, 2022 at 10:57 PM Tassos Noulas  wrote:
>
> Hi Daniel,
>
> The project described here may be in line with what you are suggesting:
> https://www.theverge.com/2022/7/29/23283701/wikipediate-notable-people-ranking-map-search-scroll-zoom
>
> But could I be asking: what use case you had in mind? Why would I want to see a bunch of non geo entities on a map and what value would I extract from this aside from pure fun? I am not saying that fun is not worth it btw :), but one of the challenges we have been having with the tool is narrowing down to specific use cases that empower users and hopefully the Wikipedia ecosystem (you can imagine users crowdsourcing info through a cartographic/mobile platform in the future).
>
> The idea of parameterized url has been somewhat developed:
> https://wiki-atlas.org/?wikipage=Stuyvesant_Town%E2%80%93Peter_Cooper_Village=-73.97778153419495=40.731669455258725=en
>
> But it is not serving all purposes in its current form and I think connecting entities based on QIDs as you suggest is a great idea. In addition to linking better the wiki entities with a map, and vice versa, we could exploit Wikidata’s querying functionality to allow for way more complex filtering approaches to those the tool currently offers (based on popularity, categories, keywords).
>
> Best,
> Tassos
>
> On Sat, 17 Dec 2022 at 19:43, Daniel Mietchen  wrote:
>>
>> Dear Diego, Aidan and Benjamin,
>> thanks for working on such functionality - both tools seem to be quite
>> useful already.
>> One way to abstract things out further would be to facilitate a
>> mapping (e.g. heatmaps) of non-geo things - for example basketball
>> players by number of points, perhaps with filters per season or club.
>> Is anyone here thinking in such directions?
>> Another request would be to have parametrized URLs based on QID and
>> perhaps type or language, e.g.
>> http://www.wiki-atlas.org/English/museums/Q7877613 or some such.
>> Best,
>> Daniel
>>
>> On Sat, Dec 17, 2022 at 2:36 AM Aidan Hogan  wrote:
>> >
>> > Hi Diego,
>> >
>> > Thanks for the pointer; this is very cool! We would be happy to share
>> > experiences. (It's very impressive how many points you are able to
>> > render, and how these resize at different scales!)
>> >
>> > Indeed it seems we were not so original with the name. :)
>> >
>> > 

[Wikidata] Re: Wikidata Atlas: a geographic view of Wikidata entities [feedback welcome!]

[Isaac Johnson]

Daniel,
A few other research projects that have used Wikipedia/Wikidata to generate
"maps" of concepts that are not explicitly spatial and might be relevant to
your thinking. I'm only aware of a live demo for the second one (
Frankenplace ) but the papers have good figures
to illustrate:

   - Hecht, Brent, et al. "Explanatory semantic relatedness and explicit
   spatialization for exploratory search." Proceedings of the 35th
   international ACM SIGIR conference on Research and development in
   information retrieval. 2012.
   

   - Adams, Benjamin, Grant McKenzie, and Mark Gahegan. "Frankenplace:
   interactive thematic mapping for ad hoc exploratory search." Proceedings of
   the 24th international conference on world wide web. 2015.
   
   - Sen, Shilad, et al. "Toward Universal Spatialization Through
   Wikipedia-Based Semantic Enhancement." ACM Transactions on Interactive
   Intelligent Systems (TiiS) 9.2-3 (2019): 1-29.
   

Best,
Isaac

On Mon, Dec 19, 2022 at 5:37 PM Daniel Mietchen via Wikidata <
wikidata@lists.wikimedia.org> wrote:

> Dear Tassos,
>
> thanks for the example - that map is interesting but still arranged in
> terms of geocoordinates, and based on Wikipedia data.
> What I had in mind is a map that positions Wikidata items generically
> (i.e. without the need for geolocation statements via P625) but
> somewhat reliably (for a given query and reasonably stable data) in a
> 2D or 3D space (perhaps even as a function of some additional
> parameters) and then allows the user to zoom around inside this system
> to explore spatial relationships just as they can explore geospatial
> relationships in your gelocated wiki atlases.
> The closest thing to this that I have at hand right now is
> https://galaxy.opensyllabus.org/ , which clusters syllabi by topic and
> allows zooming but is not based on Wikidata.
> The Wikidata Query Service has some visualizations that do part of
> that but these (i) do not provide zooming, (ii) often time out and
> have other problems, e.g. (iii) no reliable position of a given node
> or (iv) little to no meaning in adjacency.
>
> Another thing relevant here are Wikidata maps as per
> https://commons.wikimedia.org/wiki/Wikidata_map
> with their bright and dark areas and in particular their evolution over
> time
> All of these provide for fertile ground to engage relevant
> communities, and It would be very helpful to have similar
> visualizations (e.g. change as a function of some parameter) for any
> part of Wikidata, including but not limited to geodata.
>
> Best,
>
> Daniel
>
>
>
> On Sat, Dec 17, 2022 at 10:57 PM Tassos Noulas  wrote:
> >
> > Hi Daniel,
> >
> > The project described here may be in line with what you are suggesting:
> >
> https://www.theverge.com/2022/7/29/23283701/wikipediate-notable-people-ranking-map-search-scroll-zoom
> >
> > But could I be asking: what use case you had in mind? Why would I want
> to see a bunch of non geo entities on a map and what value would I extract
> from this aside from pure fun? I am not saying that fun is not worth it btw
> :), but one of the challenges we have been having with the tool is
> narrowing down to specific use cases that empower users and hopefully the
> Wikipedia ecosystem (you can imagine users crowdsourcing info through a
> cartographic/mobile platform in the future).
> >
> > The idea of parameterized url has been somewhat developed:
> >
> https://wiki-atlas.org/?wikipage=Stuyvesant_Town%E2%80%93Peter_Cooper_Village=-73.97778153419495=40.731669455258725=en
> >
> > But it is not serving all purposes in its current form and I think
> connecting entities based on QIDs as you suggest is a great idea. In
> addition to linking better the wiki entities with a map, and vice versa, we
> could exploit Wikidata’s querying functionality to allow for way more
> complex filtering approaches to those the tool currently offers (based on
> popularity, categories, keywords).
> >
> > Best,
> > Tassos
> >
> > On Sat, 17 Dec 2022 at 19:43, Daniel Mietchen <
> daniel.mietc...@googlemail.com> wrote:
> >>
> >> Dear Diego, Aidan and Benjamin,
> >> thanks for working on such functionality - both tools seem to be quite
> >> useful already.
> >> One way to abstract things out further would be to facilitate a
> >> mapping (e.g. heatmaps) of non-geo things - for example basketball
> >> players by number of points, perhaps with filters per season or club.
> >> Is anyone here thinking in such directions?
> >> Another request would be to have parametrized URLs based on QID and
> >> perhaps type or language, e.g.
> >> http://www.wiki-atlas.org/English/museums/Q7877613 or some such.
> >> Best,
> >> Daniel
> >>
> >> On Sat, Dec 17, 2022 at 2:36 AM Aidan Hogan  wrote:
> >> >
> >> > Hi Diego,
> >> >
> 

[Wikidata] Re: Wikidata Atlas: a geographic view of Wikidata entities [feedback welcome!]

[Daniel Mietchen via Wikidata]

Dear Tassos,

thanks for the example - that map is interesting but still arranged in
terms of geocoordinates, and based on Wikipedia data.
What I had in mind is a map that positions Wikidata items generically
(i.e. without the need for geolocation statements via P625) but
somewhat reliably (for a given query and reasonably stable data) in a
2D or 3D space (perhaps even as a function of some additional
parameters) and then allows the user to zoom around inside this system
to explore spatial relationships just as they can explore geospatial
relationships in your gelocated wiki atlases.
The closest thing to this that I have at hand right now is
https://galaxy.opensyllabus.org/ , which clusters syllabi by topic and
allows zooming but is not based on Wikidata.
The Wikidata Query Service has some visualizations that do part of
that but these (i) do not provide zooming, (ii) often time out and
have other problems, e.g. (iii) no reliable position of a given node
or (iv) little to no meaning in adjacency.

Another thing relevant here are Wikidata maps as per
https://commons.wikimedia.org/wiki/Wikidata_map
with their bright and dark areas and in particular their evolution over time
All of these provide for fertile ground to engage relevant
communities, and It would be very helpful to have similar
visualizations (e.g. change as a function of some parameter) for any
part of Wikidata, including but not limited to geodata.

Best,

Daniel



On Sat, Dec 17, 2022 at 10:57 PM Tassos Noulas  wrote:
>
> Hi Daniel,
>
> The project described here may be in line with what you are suggesting:
> https://www.theverge.com/2022/7/29/23283701/wikipediate-notable-people-ranking-map-search-scroll-zoom
>
> But could I be asking: what use case you had in mind? Why would I want to see 
> a bunch of non geo entities on a map and what value would I extract from this 
> aside from pure fun? I am not saying that fun is not worth it btw :), but one 
> of the challenges we have been having with the tool is narrowing down to 
> specific use cases that empower users and hopefully the Wikipedia ecosystem 
> (you can imagine users crowdsourcing info through a cartographic/mobile 
> platform in the future).
>
> The idea of parameterized url has been somewhat developed:
> https://wiki-atlas.org/?wikipage=Stuyvesant_Town%E2%80%93Peter_Cooper_Village=-73.97778153419495=40.731669455258725=en
>
> But it is not serving all purposes in its current form and I think connecting 
> entities based on QIDs as you suggest is a great idea. In addition to linking 
> better the wiki entities with a map, and vice versa, we could exploit 
> Wikidata’s querying functionality to allow for way more complex filtering 
> approaches to those the tool currently offers (based on popularity, 
> categories, keywords).
>
> Best,
> Tassos
>
> On Sat, 17 Dec 2022 at 19:43, Daniel Mietchen 
>  wrote:
>>
>> Dear Diego, Aidan and Benjamin,
>> thanks for working on such functionality - both tools seem to be quite
>> useful already.
>> One way to abstract things out further would be to facilitate a
>> mapping (e.g. heatmaps) of non-geo things - for example basketball
>> players by number of points, perhaps with filters per season or club.
>> Is anyone here thinking in such directions?
>> Another request would be to have parametrized URLs based on QID and
>> perhaps type or language, e.g.
>> http://www.wiki-atlas.org/English/museums/Q7877613 or some such.
>> Best,
>> Daniel
>>
>> On Sat, Dec 17, 2022 at 2:36 AM Aidan Hogan  wrote:
>> >
>> > Hi Diego,
>> >
>> > Thanks for the pointer; this is very cool! We would be happy to share
>> > experiences. (It's very impressive how many points you are able to
>> > render, and how these resize at different scales!)
>> >
>> > Indeed it seems we were not so original with the name. :)
>> >
>> > It seems both systems offer two different functionalities: one focuses
>> > on the "what's close to here" functionality, while the other focuses on
>> > the "where in the world are there X" functionality, like "where in the
>> > world are there lighthouses [1]", but generalised to all the types in
>> > Wikidata. It would be interesting to see how these two modalities could
>> > be combined in future maybe?
>> >
>> > Best,
>> > Aidan
>> >
>> > [1] https://www.lightphotos.net/photos/map_all.php
>> >
>> > On 2022-12-16 21:45, Diego Saez-Trumper wrote:
>> > > Hi Aidan,
>> > >
>> > > With Tassos and Rossano, we have a similar project (same name in fact).
>> > > You can check-it out here: www.wiki-atlas.org
>> > > , maybe we could exchange some experiences
>> > > about it.
>> > >
>> > > Best,
>> > > Diego
>> > >
>> > > ___
>> > > Wikidata mailing list -- wikidata@lists.wikimedia.org
>> > > Public archives at 
>> > > https://lists.wikimedia.org/hyperkitty/list/wikidata@lists.wikimedia.org/message/4ZE6CWFNHOYQH47DEXJKT7J4P2ASSQVN/
>> > > To unsubscribe send an email to 

[Wikidata] Re: Wikidata Atlas: a geographic view of Wikidata entities [feedback welcome!]

[Daniel Mietchen via Wikidata]

Dear Diego, Aidan and Benjamin,
thanks for working on such functionality - both tools seem to be quite
useful already.
One way to abstract things out further would be to facilitate a
mapping (e.g. heatmaps) of non-geo things - for example basketball
players by number of points, perhaps with filters per season or club.
Is anyone here thinking in such directions?
Another request would be to have parametrized URLs based on QID and
perhaps type or language, e.g.
http://www.wiki-atlas.org/English/museums/Q7877613 or some such.
Best,
Daniel

On Sat, Dec 17, 2022 at 2:36 AM Aidan Hogan  wrote:
>
> Hi Diego,
>
> Thanks for the pointer; this is very cool! We would be happy to share
> experiences. (It's very impressive how many points you are able to
> render, and how these resize at different scales!)
>
> Indeed it seems we were not so original with the name. :)
>
> It seems both systems offer two different functionalities: one focuses
> on the "what's close to here" functionality, while the other focuses on
> the "where in the world are there X" functionality, like "where in the
> world are there lighthouses [1]", but generalised to all the types in
> Wikidata. It would be interesting to see how these two modalities could
> be combined in future maybe?
>
> Best,
> Aidan
>
> [1] https://www.lightphotos.net/photos/map_all.php
>
> On 2022-12-16 21:45, Diego Saez-Trumper wrote:
> > Hi Aidan,
> >
> > With Tassos and Rossano, we have a similar project (same name in fact).
> > You can check-it out here: www.wiki-atlas.org
> > , maybe we could exchange some experiences
> > about it.
> >
> > Best,
> > Diego
> >
> > ___
> > Wikidata mailing list -- wikidata@lists.wikimedia.org
> > Public archives at 
> > https://lists.wikimedia.org/hyperkitty/list/wikidata@lists.wikimedia.org/message/4ZE6CWFNHOYQH47DEXJKT7J4P2ASSQVN/
> > To unsubscribe send an email to wikidata-le...@lists.wikimedia.org
> ___
> Wikidata mailing list -- wikidata@lists.wikimedia.org
> Public archives at 
> https://lists.wikimedia.org/hyperkitty/list/wikidata@lists.wikimedia.org/message/H2JPDXIET5I2ZXZOUXQNJDCAF4FXJAOP/
> To unsubscribe send an email to wikidata-le...@lists.wikimedia.org
___
Wikidata mailing list -- wikidata@lists.wikimedia.org
Public archives at 
https://lists.wikimedia.org/hyperkitty/list/wikidata@lists.wikimedia.org/message/2GKYN6DMPJVZBVBYWVYG2UF6S5WWI76F/
To unsubscribe send an email to wikidata-le...@lists.wikimedia.org

[Wikidata] Re: Wikidata Atlas: a geographic view of Wikidata entities [feedback welcome!]

[Aidan Hogan]

Hi Diego,

Thanks for the pointer; this is very cool! We would be happy to share 
experiences. (It's very impressive how many points you are able to 
render, and how these resize at different scales!)


Indeed it seems we were not so original with the name. :)

It seems both systems offer two different functionalities: one focuses 
on the "what's close to here" functionality, while the other focuses on 
the "where in the world are there X" functionality, like "where in the 
world are there lighthouses [1]", but generalised to all the types in 
Wikidata. It would be interesting to see how these two modalities could 
be combined in future maybe?


Best,
Aidan

[1] https://www.lightphotos.net/photos/map_all.php

On 2022-12-16 21:45, Diego Saez-Trumper wrote:

Hi Aidan,

With Tassos and Rossano, we have a similar project (same name in fact). 
You can check-it out here: www.wiki-atlas.org 
, maybe we could exchange some experiences 
about it.


Best,
Diego

___
Wikidata mailing list -- wikidata@lists.wikimedia.org
Public archives at 
https://lists.wikimedia.org/hyperkitty/list/wikidata@lists.wikimedia.org/message/4ZE6CWFNHOYQH47DEXJKT7J4P2ASSQVN/
To unsubscribe send an email to wikidata-le...@lists.wikimedia.org

___
Wikidata mailing list -- wikidata@lists.wikimedia.org
Public archives at 
https://lists.wikimedia.org/hyperkitty/list/wikidata@lists.wikimedia.org/message/H2JPDXIET5I2ZXZOUXQNJDCAF4FXJAOP/
To unsubscribe send an email to wikidata-le...@lists.wikimedia.org

[Wikidata] Re: Wikidata Atlas: a geographic view of Wikidata entities [feedback welcome!]

[Diego Saez-Trumper]

Hi Aidan,

With Tassos and Rossano, we have a similar project (same name in fact). You
can check-it out here: www.wiki-atlas.org, maybe we could exchange some
experiences about it.

Best,
Diego
___
Wikidata mailing list -- wikidata@lists.wikimedia.org
Public archives at 
https://lists.wikimedia.org/hyperkitty/list/wikidata@lists.wikimedia.org/message/4ZE6CWFNHOYQH47DEXJKT7J4P2ASSQVN/
To unsubscribe send an email to wikidata-le...@lists.wikimedia.org