W3C home > Mailing lists > Public > public-vocabs@w3.org > January 2014

Re: makesOffer should accept Service

From: Martin Hepp <martin.hepp@ebusiness-unibw.org>
Date: Tue, 7 Jan 2014 21:57:08 +0100
Cc: public-vocabs@w3.org
Message-Id: <71F0AB1E-6289-41BD-BAF5-EF92E5C4BF6E@ebusiness-unibw.org>
To: kcoyle@kcoyle.net
Hi Karen:

I think the only problem in here is the textual definition of http://schema.org/Product. But also keep in mind that the target audience of the schema.org specification are Web developers who want to find the most appropriate type and property identifiers for their content. So we should be careful with too generic wording. In the end, what matters is the quality of type information in Web data, at scale. The conceptual clarify of the schema is only one of many factors in this challenge.

A product is essentially a thing that is the subject of an offer to transfer rights on this object, or an activity that one promises to carry out. Being a product makes many properties useful that are not typical for things per se, e.g. having a manufacturer or being associated with a brand. Again, being a product is a role that a thing can take. 

In the case of creative works, I think the main ambiguity is that there are (1) abstract pieces of creative works and (2) manifestations thereof. For instance, if I offered to write poems for money, we would have the abstract poem (the sequence of words) and a print-out, and we could argue whether my offer includes the former or the latter, or both.

But this issue is in my opinion independent from the distinction between product and offer.

As for your proposal to rearrange

>  agent - object - promise - location
> 
> to:
> 
>  agent - promise - object - location

Yes, the order in my example was sub-optimal. In essence, the offer (the promise) is a higher-arity relation that links agent, object, location, and the various terms and conditions (expected compensation, temporal and geographical validity of the offer, ...), and the location.


Martin


On Jan 7, 2014, at 9:26 PM, Karen Coyle wrote:

> 
> 
> On 1/7/14, 11:21 AM, Martin Hepp wrote:
> 
>> In theory, we would not need a distinct type for "Product" (nor Service); we could simply say we have an agent, and offer, and a thing. But it felt unnatural to not have a Product class in an e-commerce ontology.
>> 
>> But I disagree that the product type is the same as the offer. Being a product is a role that a thing can have, and this type adds properties that are relevant only in this context (e.g. a stock keeping unit). The offer is the promise to transfer some or all rights on the thing.
>> 
> 
> Ah, but that isn't what I meant. I meant that in schema.org "product" is a thing being offered. Which I think would modify your scenario from:
>  agent - object - promise - location
> 
> to:
> 
>  agent - promise - object - location
> 
> The schema Product *definition* includes "made available" (we'll ignore "for sale") in the definition of product. So offer and product appear to be closely entwined, as a product is defined as something that is offered. To separate them, one would need to define product as "something made" with perhaps "eventually intended for sale/exchange". That would be reminiscent of the indecs "people create stuff, people make deals about stuff."
> 
> Product can be contrasted with CreativeWork, whose definition is less than enlightening: "The most generic kind of creative work, including books, movies, photographs, software programs, etc." CreativeWork is not by definition "made available" in the way that Product is. And of course some creative works are products, so like the separation of product and service we have fuzzy definitional boundaries. What is the difference between a film as a creative work and a film on DVD for sale?
> 
> I'm not suggesting that we should or even can resolve these, but I do see the problem as going beyond the product/offer issue.
> 
> kc
> 
> 
>> The beauty of this model is that
>> 
>> - it fits a huge range of scenarios and
>> - with http://schema.org/Demand, we have a type that is the exact mirror of the offer and can be used to model the demand side in a structurally identical fashion.
>> 
>> For instance, if you sell the same product at different conditions in different countries, or at different times (e.g. happy hours for drinks), you simply use multiple offers for the same object.
>> 
>> Martin
>> 
>> On Jan 7, 2014, at 8:08 PM, Karen Coyle wrote:
>> 
>>> There is quite a bit of semantic ambiguity between schema:product, service, and offer. Product seems to match Martin's GoodRelations definition of "product or service", but I could also think of product and service as "product offered" "service offered", thus making them not fully distinct from Offer, and possibly subtypes of offer. The only reason I could see for a separation between product and service would be if they have some distinct properties that cannot be shared.
>>> 
>>> kc
>>> 
>>> Thing > Product
>>> A product is anything that is made available for sale—for example, a pair of shoes, a concert ticket, or a car. Commodity services, like haircuts, can also be represented using this type.
>>> 
>>> Thing > Intangible > Service
>>> A service provided by an organization, e.g. delivery service, print services, etc.
>>> 
>>> Thing > Intangible > Offer
>>> An offer to sell an item—for example, an offer to sell a product, the DVD of a movie, or tickets to an event.
>>> 
>>> 
>>> 
>>> 
>>> On 1/7/14, 10:21 AM, Martin Hepp wrote:
>>>> The properties of http://schema.org/Product are naturally not sufficient for all possible details of every possible object or activity that can be offered. There are two solutions for that:
>>>> 
>>>> 1. Propose additional properties for http://schema.org/Service.
>>>> 2. Wait for the generic property-value proposal that I am working on; more to follow on http://www.w3.org/wiki/WebSchemas/PropertyValuePairs.
>>>> 
>>>> As a bit of background: In GoodRelations, there is one joint class for products and services, because many data sources (e.g. Web shops) are not able to signal whether an item is a product or a service (e.g. because the backend-database does not store this distinction). Thus, we need a common abstraction. When integrating GoodRelations into schema.org, we merged that into http://schema.org/Product, since this covers the majority of the use-cases.
>>>> 
>>>> Martin
>>>> 
>>>> On Jan 2, 2014, at 1:55 PM, Robert Kost wrote:
>>>> 
>>>>> Hi Martin —
>>>>> 
>>>>> with regard to …
>>>>>> 
>>>>>> 2. As for modeling services, schema:Product fits; it can be constrained by combining it with www.productontology.org types or other schema.org types without problems (in general; some Google infrastructure does not yet fully support multi-typed entities).
>>>>>> 
>>>>> 
>>>>> … I don’t understand.  How would one use Product to model, say: the services of a product liability attorney who works on contingency: the SLA of an internet service provider offering to provide specific bandwidth rates, availabilities, etc. over a specific period of time;  a real estate broker specializing in distressed properties; or a hospital that specializes in orthopedic surgeries and that wants to cite certain procedures or success rates?  None of the Product properties seem to support these notions of performance, conditionality, rate, quality, technique, delivery time and method, etc.  And, conversely, many of the Product properties do not pertain.  It is possible, I suppose, to stuff of all of the salient differentiators into “description,” but the utility of Schema is undermined.  In any event, Product seems to be a tortured use of the word to describe what is being offered.
>>>>> 
>>>>> best,
>>>>> 
>>>>> Rob
>>>> 
>>>> --------------------------------------------------------
>>>> martin hepp
>>>> e-business & web science research group
>>>> universitaet der bundeswehr muenchen
>>>> 
>>>> e-mail:  hepp@ebusiness-unibw.org
>>>> phone:   +49-(0)89-6004-4217
>>>> fax:     +49-(0)89-6004-4620
>>>> www:     http://www.unibw.de/ebusiness/ (group)
>>>>          http://www.heppnetz.de/ (personal)
>>>> skype:   mfhepp
>>>> twitter: mfhepp
>>>> 
>>>> Check out GoodRelations for E-Commerce on the Web of Linked Data!
>>>> =================================================================
>>>> * Project Main Page: http://purl.org/goodrelations/
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>> 
>>> --
>>> Karen Coyle
>>> kcoyle@kcoyle.net http://kcoyle.net
>>> m: 1-510-435-8234
>>> skype: kcoylenet
>>> 
>> 
>> --------------------------------------------------------
>> martin hepp
>> e-business & web science research group
>> universitaet der bundeswehr muenchen
>> 
>> e-mail:  hepp@ebusiness-unibw.org
>> phone:   +49-(0)89-6004-4217
>> fax:     +49-(0)89-6004-4620
>> www:     http://www.unibw.de/ebusiness/ (group)
>>          http://www.heppnetz.de/ (personal)
>> skype:   mfhepp
>> twitter: mfhepp
>> 
>> Check out GoodRelations for E-Commerce on the Web of Linked Data!
>> =================================================================
>> * Project Main Page: http://purl.org/goodrelations/
>> 
>> 
>> 
>> 
> 
> -- 
> Karen Coyle
> kcoyle@kcoyle.net http://kcoyle.net
> m: 1-510-435-8234
> skype: kcoylenet

--------------------------------------------------------
martin hepp
e-business & web science research group
universitaet der bundeswehr muenchen

e-mail:  hepp@ebusiness-unibw.org
phone:   +49-(0)89-6004-4217
fax:     +49-(0)89-6004-4620
www:     http://www.unibw.de/ebusiness/ (group)
         http://www.heppnetz.de/ (personal)
skype:   mfhepp 
twitter: mfhepp

Check out GoodRelations for E-Commerce on the Web of Linked Data!
=================================================================
* Project Main Page: http://purl.org/goodrelations/
Received on Tuesday, 7 January 2014 20:57:33 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:49:20 UTC