New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

schema.org/Distillery ? #743

Open
radusi opened this Issue Aug 21, 2015 · 19 comments

Comments

Projects
None yet
8 participants
@radusi

radusi commented Aug 21, 2015

would adding //schema.org/Distillery be acceptable, since neither Winery nor Brewery does not correctly describe a production place for a Bourbon / Scotch ?

Thing > Place > LocalBusiness > FoodEstablishment > Distillery
Thing > Organization > LocalBusiness > FoodEstablishment > Distillery

RichardWallis added a commit that referenced this issue Aug 28, 2015

@danbri

This comment has been minimized.

Show comment
Hide comment
@danbri

danbri Aug 28, 2015

Contributor

Seems useful but perhaps low priority? I'm not entirely convinced it is good to add this to the core - there are probably many more kinds of food establishment -related places that we don't include. In particular we should look into places that are more typically visited, as more urgent than a production-oriented place. Perhaps there is a markup model we can find that allows us to pass off such things to Wikipedia?

Specifically Wikidata has https://www.wikidata.org/wiki/Q1251750 for distillery ...

Contributor

danbri commented Aug 28, 2015

Seems useful but perhaps low priority? I'm not entirely convinced it is good to add this to the core - there are probably many more kinds of food establishment -related places that we don't include. In particular we should look into places that are more typically visited, as more urgent than a production-oriented place. Perhaps there is a markup model we can find that allows us to pass off such things to Wikipedia?

Specifically Wikidata has https://www.wikidata.org/wiki/Q1251750 for distillery ...

@RichardWallis

This comment has been minimized.

Show comment
Hide comment
@RichardWallis

RichardWallis Aug 28, 2015

Contributor

I traded off ease of application against 'part of a bigger concern' against
who might then own it and in this case the first of those won.

Liking the idea of finding a generic way that we can hand off such things,
that don't need to introduce new properties, to WikiData.

On Fri, Aug 28, 2015 at 10:37 AM, Dan Brickley notifications@github.com
wrote:

Seems useful but perhaps low priority? I'm not entirely convinced it is
good to add this to the core - there are probably many more kinds of food
establishment -related places that we don't include. In particular we
should look into places that are more typically visited, as more urgent
than a production-oriented place. Perhaps there is a markup model we can
find that allows us to pass off such things to Wikipedia?

Specifically Wikidata has https://www.wikidata.org/wiki/Q1251750 for
distillery ...


Reply to this email directly or view it on GitHub
#743 (comment)
.

Contributor

RichardWallis commented Aug 28, 2015

I traded off ease of application against 'part of a bigger concern' against
who might then own it and in this case the first of those won.

Liking the idea of finding a generic way that we can hand off such things,
that don't need to introduce new properties, to WikiData.

On Fri, Aug 28, 2015 at 10:37 AM, Dan Brickley notifications@github.com
wrote:

Seems useful but perhaps low priority? I'm not entirely convinced it is
good to add this to the core - there are probably many more kinds of food
establishment -related places that we don't include. In particular we
should look into places that are more typically visited, as more urgent
than a production-oriented place. Perhaps there is a markup model we can
find that allows us to pass off such things to Wikipedia?

Specifically Wikidata has https://www.wikidata.org/wiki/Q1251750 for
distillery ...


Reply to this email directly or view it on GitHub
#743 (comment)
.

@jvandriel

This comment has been minimized.

Show comment
Hide comment
@jvandriel

jvandriel Aug 28, 2015

The way I see it we have 2 options:

Option A - use schema.org/additionalType + Productontology

<div itemscope itemtype="http://schema.org/LocalBusiness">
    <link itemprop="additionalType" href="http://www.productontology.org/id/Distillery" />
    <h1 itemprop="name">Acme Alcohol Inc.</h1>
</div>

Option B - use schema.org/sameAs + Wikidata

<div itemscope itemtype="http://schema.org/LocalBusiness">
    <link itemprop="sameAs" href="http://www.wikidata.org/entity/Q1251750" />
    <h1 itemprop="name">Acme Alcohol Inc.</h1>
</div>

Personally I like the idea of using schema.org/additionalType + Productontology very much because it allows one to turn a generic type into a more specific type..

I feel the downside of using schema.org/sameAs + Wikidata (or DBPedia, Freebase, Wikipedia, etc) is that one ends up comparing a generic type with a more specific type.

But hey, that's just my personal preference. How do you feel about this @RichardWallis and @danbri?

cc: @mfhepp

The way I see it we have 2 options:

Option A - use schema.org/additionalType + Productontology

<div itemscope itemtype="http://schema.org/LocalBusiness">
    <link itemprop="additionalType" href="http://www.productontology.org/id/Distillery" />
    <h1 itemprop="name">Acme Alcohol Inc.</h1>
</div>

Option B - use schema.org/sameAs + Wikidata

<div itemscope itemtype="http://schema.org/LocalBusiness">
    <link itemprop="sameAs" href="http://www.wikidata.org/entity/Q1251750" />
    <h1 itemprop="name">Acme Alcohol Inc.</h1>
</div>

Personally I like the idea of using schema.org/additionalType + Productontology very much because it allows one to turn a generic type into a more specific type..

I feel the downside of using schema.org/sameAs + Wikidata (or DBPedia, Freebase, Wikipedia, etc) is that one ends up comparing a generic type with a more specific type.

But hey, that's just my personal preference. How do you feel about this @RichardWallis and @danbri?

cc: @mfhepp

@mfhepp

This comment has been minimized.

Show comment
Hide comment
@mfhepp

mfhepp Aug 28, 2015

Contributor

Yes, I also think we should use www.productontology.org, which is also semantically fine (note that the notion of "product" is very broad in schema.org and GoodRelations, and that Place is not disjoint with Product (which also makes sense, since you can very well offer a place for sale or rental).

Contributor

mfhepp commented Aug 28, 2015

Yes, I also think we should use www.productontology.org, which is also semantically fine (note that the notion of "product" is very broad in schema.org and GoodRelations, and that Place is not disjoint with Product (which also makes sense, since you can very well offer a place for sale or rental).

@RichardWallis

This comment has been minimized.

Show comment
Hide comment
@RichardWallis

RichardWallis Aug 28, 2015

Contributor

I think that the schema.org/aditionalType option would be a good pragmatic
solution and www.productontology.org as an excellent example to point at.

Wondering though if there should not be a more generic, not so product
focused, solution to sit in front of WikiData - www.thingontology.org ?

Agree with Martin about the technical broadness of 'product', it is more
about perception, and the inevitable questions around is a distillery
really a product. Maybe that should be part of a wider discussion about
representing things with physical characteristics [often associated with
Product - color, weight, etc.] that are not products such as fossils,
rocks, museum exhibits, etc.

~Richard

On Fri, Aug 28, 2015 at 11:43 AM, Martin Hepp notifications@github.com
wrote:

Yes, I also think we should use www.productontology.org, which is also
semantically fine (note that the notion of "product" is very broad in
schema.org and GoodRelations, and that Place is not disjoint with Product
(which also makes sense, since you can very well offer a place for sale or
rental).


Reply to this email directly or view it on GitHub
#743 (comment)
.

Contributor

RichardWallis commented Aug 28, 2015

I think that the schema.org/aditionalType option would be a good pragmatic
solution and www.productontology.org as an excellent example to point at.

Wondering though if there should not be a more generic, not so product
focused, solution to sit in front of WikiData - www.thingontology.org ?

Agree with Martin about the technical broadness of 'product', it is more
about perception, and the inevitable questions around is a distillery
really a product. Maybe that should be part of a wider discussion about
representing things with physical characteristics [often associated with
Product - color, weight, etc.] that are not products such as fossils,
rocks, museum exhibits, etc.

~Richard

On Fri, Aug 28, 2015 at 11:43 AM, Martin Hepp notifications@github.com
wrote:

Yes, I also think we should use www.productontology.org, which is also
semantically fine (note that the notion of "product" is very broad in
schema.org and GoodRelations, and that Place is not disjoint with Product
(which also makes sense, since you can very well offer a place for sale or
rental).


Reply to this email directly or view it on GitHub
#743 (comment)
.

@jvandriel

This comment has been minimized.

Show comment
Hide comment
@jvandriel

jvandriel Aug 28, 2015

"Wondering though if there should not be a more generic, not so product
focused, solution to sit in front of WikiData - www.thingontology.org ?"

For quite some time I was using schema.org/additionalType for referring to Wikidata and DBpedia, thinking, since these have types, I can refer to them this way. That is, until @mfhepp and @thadguidry pointed out to me that in most cases this isn't the right thing to do.

Since then I've read some documents published by @mfhepp and what I got out of that is that things go wrong when resolving the semantics (sorry if can't fully explain better, that's the most I could make of it. It's a bit out of my league still).

What I want to explain with this is that for folks like me, who don't know all technical rules by heart, using schema.org/additionalType + eg, Wikidata seems a very logical solution because:

schema.org type (machine readable data) > additionalType > Wikidata type (machine readable data)

Which is how deep most webmasters I've encountered think about it. Now I'm happy to have learned from others that my thinking was wrong, but I wonder if there's anything that can be done so that one can use schema.org/additionalType (or @id or @typeof) for referring to a second type at locations like Wikidata, DBpedia and Productontology?

Because coming up with a new system or even using my example [a] still forces webmasters to know and think about whether they should use schema.org/sameAs or schema.org/additionalType depending on the external enumeration they want to refer to.

IMHO it would help greatly if we could simplify this for the average webmaster so they'd only have to think about whether they're referring to a machine readable enumeration (productontology, wikidata, etc) or a plain webpage (wikipedia, or any other page on the web).

Now I have no idea whether it's even remotely possible or not but looking at this from this POV it'd be great if we could simplify this by using schema.org/additionalType (@id, @resource) to point to machine readable 'types' and schema.org/sameAs for pointing to things like social media, Wikipedia, Github, 'normal' webpages, etc, etc.

"Wondering though if there should not be a more generic, not so product
focused, solution to sit in front of WikiData - www.thingontology.org ?"

For quite some time I was using schema.org/additionalType for referring to Wikidata and DBpedia, thinking, since these have types, I can refer to them this way. That is, until @mfhepp and @thadguidry pointed out to me that in most cases this isn't the right thing to do.

Since then I've read some documents published by @mfhepp and what I got out of that is that things go wrong when resolving the semantics (sorry if can't fully explain better, that's the most I could make of it. It's a bit out of my league still).

What I want to explain with this is that for folks like me, who don't know all technical rules by heart, using schema.org/additionalType + eg, Wikidata seems a very logical solution because:

schema.org type (machine readable data) > additionalType > Wikidata type (machine readable data)

Which is how deep most webmasters I've encountered think about it. Now I'm happy to have learned from others that my thinking was wrong, but I wonder if there's anything that can be done so that one can use schema.org/additionalType (or @id or @typeof) for referring to a second type at locations like Wikidata, DBpedia and Productontology?

Because coming up with a new system or even using my example [a] still forces webmasters to know and think about whether they should use schema.org/sameAs or schema.org/additionalType depending on the external enumeration they want to refer to.

IMHO it would help greatly if we could simplify this for the average webmaster so they'd only have to think about whether they're referring to a machine readable enumeration (productontology, wikidata, etc) or a plain webpage (wikipedia, or any other page on the web).

Now I have no idea whether it's even remotely possible or not but looking at this from this POV it'd be great if we could simplify this by using schema.org/additionalType (@id, @resource) to point to machine readable 'types' and schema.org/sameAs for pointing to things like social media, Wikipedia, Github, 'normal' webpages, etc, etc.

@mfhepp

This comment has been minimized.

Show comment
Hide comment
@mfhepp

mfhepp Aug 28, 2015

Contributor

FYI: I want to release a broader variant of www.productontology.org that takes away the "product" notion but essentially does the same thing - providing classes for each and every Wikipedia lemma.

I would suggest that, for the moment, let's use www.productontology.org URIs as the default.

Will keep you posted about the new service when available, but it will take a while.

Contributor

mfhepp commented Aug 28, 2015

FYI: I want to release a broader variant of www.productontology.org that takes away the "product" notion but essentially does the same thing - providing classes for each and every Wikipedia lemma.

I would suggest that, for the moment, let's use www.productontology.org URIs as the default.

Will keep you posted about the new service when available, but it will take a while.

@RichardWallis

This comment has been minimized.

Show comment
Hide comment
@RichardWallis

RichardWallis Aug 28, 2015

Contributor

Looking forward to the broader variant of www.productontology.org and
maybe using it as the basis for answering @jvandriel 's proposals.

On Fri, Aug 28, 2015 at 12:50 PM, Martin Hepp notifications@github.com
wrote:

FYI: I want to release a broader variant of www.productontology.org that
takes away the "product" notion but essentially does the same thing -
providing classes for each and every Wikipedia lemma.

I would suggest that, for the moment, let's use www.productontology.org
URIs as the default.

Will keep you posted about the new service when available, but it will
take a while.


Reply to this email directly or view it on GitHub
#743 (comment)
.

Contributor

RichardWallis commented Aug 28, 2015

Looking forward to the broader variant of www.productontology.org and
maybe using it as the basis for answering @jvandriel 's proposals.

On Fri, Aug 28, 2015 at 12:50 PM, Martin Hepp notifications@github.com
wrote:

FYI: I want to release a broader variant of www.productontology.org that
takes away the "product" notion but essentially does the same thing -
providing classes for each and every Wikipedia lemma.

I would suggest that, for the moment, let's use www.productontology.org
URIs as the default.

Will keep you posted about the new service when available, but it will
take a while.


Reply to this email directly or view it on GitHub
#743 (comment)
.

@thadguidry

This comment has been minimized.

Show comment
Hide comment
@thadguidry

thadguidry Aug 28, 2015

Wikidata does NOT have Types or Classes, their model is a bit different.

What they have are properties and statements....tons....that then . So
when you want to basically find a Class or Type or Equivalent Type or
Freebase kinda Type.... you have to look at properties.

There are a few properties that make finding Types or Classes a bit easier
on Wikidata (you can ask the Wikidat's themselves btw)
or
find topics that have Statements with one of these properties:

https://www.wikidata.org/wiki/Property:P1709

https://www.wikidata.org/wiki/Property:P279

https://www.wikidata.org/wiki/Property:P1963

And this tool uses the first 2 above to visualize and navigate the
pseudo-imposed Types/Classes :
https://www.wikidata.org/wiki/Property:P1963

Btw, the subclass statements P279 that are thrown around are really
subjective on some topics, so just know that not everyone is a Librarian or
Ontologist or even an SEO expert. :) (example: Human is a subclass of
Person... https://www.wikidata.org/wiki/Q5 uh, no and yes, but...there is
probably a better way to express it...and I'm a realist anyways)

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

Wikidata does NOT have Types or Classes, their model is a bit different.

What they have are properties and statements....tons....that then . So
when you want to basically find a Class or Type or Equivalent Type or
Freebase kinda Type.... you have to look at properties.

There are a few properties that make finding Types or Classes a bit easier
on Wikidata (you can ask the Wikidat's themselves btw)
or
find topics that have Statements with one of these properties:

https://www.wikidata.org/wiki/Property:P1709

https://www.wikidata.org/wiki/Property:P279

https://www.wikidata.org/wiki/Property:P1963

And this tool uses the first 2 above to visualize and navigate the
pseudo-imposed Types/Classes :
https://www.wikidata.org/wiki/Property:P1963

Btw, the subclass statements P279 that are thrown around are really
subjective on some topics, so just know that not everyone is a Librarian or
Ontologist or even an SEO expert. :) (example: Human is a subclass of
Person... https://www.wikidata.org/wiki/Q5 uh, no and yes, but...there is
probably a better way to express it...and I'm a realist anyways)

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

@danbri

This comment has been minimized.

Show comment
Hide comment
@danbri

danbri Apr 24, 2016

Contributor

I have added Distillery as a "pending" proposal in the new pending extension, see http://pending.webschemas.org/Distillery

I still believe (see #280 for more) that integrating with Wikidata to bring in a bulk of longer-tail vocabulary as an external extension is the way to go. But the Pending mechanism gives us a way to record sensible proposals like Distillery in a way that can be used and reviewed, while we work out the longer term details. Similarly I have just added http://pending.webschemas.org/WorkersUnion which was nearly agreed back in 2013...

Contributor

danbri commented Apr 24, 2016

I have added Distillery as a "pending" proposal in the new pending extension, see http://pending.webschemas.org/Distillery

I still believe (see #280 for more) that integrating with Wikidata to bring in a bulk of longer-tail vocabulary as an external extension is the way to go. But the Pending mechanism gives us a way to record sensible proposals like Distillery in a way that can be used and reviewed, while we work out the longer term details. Similarly I have just added http://pending.webschemas.org/WorkersUnion which was nearly agreed back in 2013...

@danbri

This comment has been minimized.

Show comment
Hide comment
@danbri

danbri Apr 24, 2016

Contributor

@thadguidry is correct btw that Wikidata doesn't have quite our notion of class. Or rather it is a second class citizen, a bit like the way Twitter for a long time didn't have much infrastructure around user-created conventions like @-mentioning people or #-tagging topics. It exists in "Wikidata the dataset" but not in the supporting software infrastructure as such. But in practice it has something rather like types, with supertyping and instantiation, and it shares with us a pretty loose pattern for association types and properties. I think it is close enough that communication is possible :) /cc @vrandezo

Contributor

danbri commented Apr 24, 2016

@thadguidry is correct btw that Wikidata doesn't have quite our notion of class. Or rather it is a second class citizen, a bit like the way Twitter for a long time didn't have much infrastructure around user-created conventions like @-mentioning people or #-tagging topics. It exists in "Wikidata the dataset" but not in the supporting software infrastructure as such. But in practice it has something rather like types, with supertyping and instantiation, and it shares with us a pretty loose pattern for association types and properties. I think it is close enough that communication is possible :) /cc @vrandezo

@thadguidry

This comment has been minimized.

Show comment
Hide comment
@thadguidry

thadguidry Apr 24, 2016

@danbri A few points on Distillery.

  • Looks like we'll also need to add an "owner" property, in distillery terms and within some laws of some countries this is called a "proprietor". This leads to the old Ownership property conundrum that some folks around the internet have voiced concern over in Schema.org in the past...which leads to a comment for Martin and a side discussion below.
  • You'll also want to be more clear on the definition and determine if it is inclusive of a "distilled spirits plant" which is what a distillery is called under USA Code of Federal Regulations http://www.ecfr.gov/cgi-bin/text-idx?SID=e3b4dd0310e0d8d38b225edfff8efc48&node=27:1.0.1.1.15&rgn=div5#27:1.0.1.1.15.1.155.1
  • You will also want to determine and state if a http://pending.webschemas.org/Distillery is inclusive of a "distilled spirits plant that makes fuel rather than beverages". Which is also within laws of some countries such as the USA and what can also be called a "distillery".

@mfhepp The intangible https://schema.org/OwnershipInfo has a way to provide ownership for lots of Things in Schema.org, but it looks like it can only be applied against Products ? I'd like to see the domain expand to also include businesses and organizations. A definition such as "A structured value providing information about when a certain organization or person owned a certain product, business, or organization". An issue needs to be opened against it to discuss this need further.

For anyone else wanting to participate in the long-tail discussion of the http://pending.webschemas.org/Distillery , a good primer for the industry in the USA is located here: https://www.ttb.gov/spirits/index.shtml

@danbri A few points on Distillery.

  • Looks like we'll also need to add an "owner" property, in distillery terms and within some laws of some countries this is called a "proprietor". This leads to the old Ownership property conundrum that some folks around the internet have voiced concern over in Schema.org in the past...which leads to a comment for Martin and a side discussion below.
  • You'll also want to be more clear on the definition and determine if it is inclusive of a "distilled spirits plant" which is what a distillery is called under USA Code of Federal Regulations http://www.ecfr.gov/cgi-bin/text-idx?SID=e3b4dd0310e0d8d38b225edfff8efc48&node=27:1.0.1.1.15&rgn=div5#27:1.0.1.1.15.1.155.1
  • You will also want to determine and state if a http://pending.webschemas.org/Distillery is inclusive of a "distilled spirits plant that makes fuel rather than beverages". Which is also within laws of some countries such as the USA and what can also be called a "distillery".

@mfhepp The intangible https://schema.org/OwnershipInfo has a way to provide ownership for lots of Things in Schema.org, but it looks like it can only be applied against Products ? I'd like to see the domain expand to also include businesses and organizations. A definition such as "A structured value providing information about when a certain organization or person owned a certain product, business, or organization". An issue needs to be opened against it to discuss this need further.

For anyone else wanting to participate in the long-tail discussion of the http://pending.webschemas.org/Distillery , a good primer for the industry in the USA is located here: https://www.ttb.gov/spirits/index.shtml

@thadguidry

This comment has been minimized.

Show comment
Hide comment
@thadguidry

thadguidry Apr 24, 2016

@danbri Another good property, I just found out from a friend might be "Total annual output" for the Distillery, which could be collected in PG (Proof Gallons) where he has to report this amount on his TBB forms to the government.

Notes: PG - A proof gallon is a gallon of liquid at 60 degrees Fahrenheit which contains 50
percent by volume of ethyl alcohol having a specific gravity of 0.7939 at 60 degrees Fahrenheit referred to water at 60 degrees Fahrenheit as unity, or the alcoholic equivalent thereof.

@danbri Another good property, I just found out from a friend might be "Total annual output" for the Distillery, which could be collected in PG (Proof Gallons) where he has to report this amount on his TBB forms to the government.

Notes: PG - A proof gallon is a gallon of liquid at 60 degrees Fahrenheit which contains 50
percent by volume of ethyl alcohol having a specific gravity of 0.7939 at 60 degrees Fahrenheit referred to water at 60 degrees Fahrenheit as unity, or the alcoholic equivalent thereof.

@danbri

This comment has been minimized.

Show comment
Hide comment
@danbri

danbri Apr 27, 2016

Contributor

re "total annual output", @thadguidry - at this point it might be more productive to look at ways of reflecting all of https://www.wikidata.org/wiki/Wikidata:List_of_properties/all into schema.org. Denny and I got some way into this last time we met. This is feeling pretty "long tail"...

Contributor

danbri commented Apr 27, 2016

re "total annual output", @thadguidry - at this point it might be more productive to look at ways of reflecting all of https://www.wikidata.org/wiki/Wikidata:List_of_properties/all into schema.org. Denny and I got some way into this last time we met. This is feeling pretty "long tail"...

@KoenLeemans

This comment has been minimized.

Show comment
Hide comment
@KoenLeemans

KoenLeemans Oct 17, 2016

Hi,

@mfhepp
Is the best solution to still use http://www.productontology.org/ URIs for more detailled descriptions of an organization (Type)? Or are there better options available at this time?

KoenLeemans commented Oct 17, 2016

Hi,

@mfhepp
Is the best solution to still use http://www.productontology.org/ URIs for more detailled descriptions of an organization (Type)? Or are there better options available at this time?

@kaushiknemani

This comment has been minimized.

Show comment
Hide comment
@kaushiknemani

kaushiknemani Mar 1, 2017

For any word http://www.productontology.org/ is giving error “ERROR (404): No respective entry available”. Try http://www.productontology.org/id/Apple . How should I give wikipedia reference ?

For any word http://www.productontology.org/ is giving error “ERROR (404): No respective entry available”. Try http://www.productontology.org/id/Apple . How should I give wikipedia reference ?

@radusi

This comment has been minimized.

Show comment
Hide comment
@mfhepp

This comment has been minimized.

Show comment
Hide comment
@mfhepp

mfhepp Mar 1, 2017

Contributor
Contributor

mfhepp commented Mar 1, 2017

@kaushiknemani

This comment has been minimized.

Show comment
Hide comment
@kaushiknemani

kaushiknemani Mar 1, 2017

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment