W3C home > Mailing lists > Public > public-schemaorg@w3.org > March 2017

Re: Schema addition request

From: Brian Tremblay <schema@btrem.com>
Date: Thu, 23 Mar 2017 11:47:02 -0700
To: public-schemaorg@w3.org
Message-ID: <689646bd-c92e-9f1f-ea23-0b0a6c2b3f7a@btrem.com>
On 3/21/17 1:26 AM, Hans Polak wrote:
> Brian Tremblay wrote:
>> does it make sense to create a new vocab for every imaginable
>> business, every imaginable building, every imaginable service, etc.
>> if [search engines and other tools] are not looking for that level
>> of specificity? I imagine LocalBusiness would be good enough for a
>> huge number of local businesses, be they book stores, hardware
>> stores, bike shops, or groceries.
>
> How can you indicate what type of LocalBusiness a LocalBusiness is?

What difference does it make, if search engines or other tools don't
process anything other than LocalBusiness? We're increasing complexity
with no benefits. And not at 0 cost. If an author changes markup from
LocalBusiness to MedicalSpa before any tools have learned what
MedicalSpa is, then the markup has /lost/ value.

That's my point. I fear that we've gotten so far ahead of the web that
we're talking to nobody.

Take another example, Restaurant. Do we really need FastFoodRestaurant
too? Why? People are desperate to search for fast food restaurants
because they're in a rush, and a search for restaurants just won't do?
If someone wants fast food, I imagine they'll just google "macdonalds"
or "kentucky fried chicken". Or more realistically, they'll just fire up
Yelp and search for restaurants, and filter the results themselves.

You might argue that there will be no progress if we never add new
vocabularies. Ok, but perhaps we should see what kind of traction there
is for Restaurant before adding Bakery, BarOrPub, Brewery,
CafeOrCoffeeShop, FastFoodRestaurant, etc.

http://schema.org/Restaurant reports "Usage: Between 10,000 and 50,000
domains". Given that there are over a billion websites spread over
hundreds of millions of domains, that's infinitesimally small. Then we
find that IceCreamShop's usage is between 100 and 1000 domains. That's
in the range of statistical error.

So when someone comes here and says, "I have a shop where we sell frozen
yogurt, can we add a new type?" The answer, imho, should be, "No,
because there's no evidence that there's enough demand to warrant it,
but we do have IceCreamShop, use that instead."

Or, more realistically, we should answer, "Restaurant is probably the
best choice for most restaurants, but if you want to be edge, you can
use IceCreamShop instead."

> I don't doubt search engines will use semantic data/schema.org
> ontology when they encounter it. It seems like a non-issue to me.

They use a small part of the vocabularies. Most are, afaict, ignored.
Which should not come as a surprise. There's little point in writing
algorithms to account for an ontology used by 1000 domains.

-- 
Brian Tremblay
Received on Thursday, 23 March 2017 18:47:38 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:12:34 UTC