[Data-modeling] Wind farm properties and map

[Ed Laurent]

Thanks for the feedback guys. Robert's suggestion to use
"/location/location/containedby" as a property of structures would be a good
fix for stating the location of wind farms, and it sounds appropriate for
all structures. It would allow people to generally describe the location of
a building or windfarm without needing to know the address.

What about adding "Contains structure(s)" and "Contained by structure"
properties to describe building complexes, wind farms, etc.? I've tried
using the "Building complex" type awhile back and it didn't really fit my
needs. I'm not sure if nesting structures would be any better, but maybe?

@Luke: I did not know of http://carma.org/ . I wonder where they get their
data? Their summaries aren't as interesting to me as the underlying data
used to create them.

My working definition of wind power company would be any power company that
supplies power generated by the wind, even if they also supply power
generated from other sources. I know very little about this subject so am
open to other definitions.

@Iain: Contains and contained by properties should work for describing a
collection of wind turbines for my purposes, especially if there was a
location/containedby property on structure. It would even work for offshore
windfarms, although a geolocation would be better in those cases than a
named topic location.

Your suggestion to add a "power generator/turbine" property to "power
station" could work well for me too.

I'm interested in the physical dimensions of wind turbines and those could
be linked to structure or power station. However, the actual turbine could
be mounted on structures of different heights so those two topics should be
described separately I guess. I'm not concerned at all about creating a
separate type to hold properties of wind turbine products that may not be
relevant to other power stations. (I've already done
so!<http://windenergy.freebase.com/type/schema/base/windenergy/wind_turbine>).
While there is a general push to describe Freebase topics in ways that are
relevant to broad audiences, one of the great things about Freebase is that
we can also geek out on the details! By describing wind turbines in more
detail I might also identify properties that could be moved to structure and
power station.

Thanks again! What's the next steps for the proposed changes to structure
and power station?

-Ed


On Mon, May 11, 2009 at 7:05 AM, Robert Cook <robert at metaweb.com> wrote:

>
> On May 11, 2009, at 8:46 AM, Iain Sproat wrote:
>
> Ed,
> 1) Each individual mast in a wind farm is a separate structure.  The
> problem is that the structure<http://www.freebase.com/view/architecture/structure>type currently doesn't have a property which gives a relationship between a
> group of structures and an individual structure - although, for building's building
> complex <http://www.freebase.com/view/architecture/building_complex>groups multiple
> buildings <http://www.freebase.com/view/architecture/building> . We cannot
> make the relationship between groups of structures and the individual
> structures which are not buildings - e.g. radio masts<http://www.freebase.com/view/en/junglinster_longwave_transmitter> ,
> telescope arrays <http://www.freebase.com/view/en/allen_telescope_array> ,
> linked bridges <http://www.freebase.com/view/en/great_belt_fixed_link> , coastal
> defences <http://www.freebase.com/view/en/delta_works> , fortifications<http://www.freebase.com/view/en/maginot_line>etc. etc.
> It is something I had explored in the (draft) structure2<http://architecture2.freebase.com/type/schema/base/architecture2/structure2>type - I'd created a "Part of larger structure or complex" property.  Let me
> know your comments on this.
>
>
> I'm the one who created the building/building complex property (quite
> awhile ago), and I think that generalizing this to structures is a very good
> idea.
>
> I'm wondering, however, if it might make sense to use
> "/location/location/containedby", perhaps for both structures and buildings,
> as it's likely that queries for containment might be done in a very general
> way.
>
> We could still have "Building complex" and "Structure complex" as types (or
> even deprecate Building complex) for the "is-a" definition.
>
> _______________________________________________
> Data-modeling mailing list
> Data-modeling at freebase.com
> http://lists.freebase.com/mailman/listinfo/data-modeling
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freebase.com/pipermail/data-modeling/attachments/20090511/b29658a5/attachment.htm