Re: Person and fictional Re: VideoGame proposal

I think we could live with the assumption that being fictional is a particular quality of being, so modeling it as a multi-typed entity, i.e. the intersection between the original class and a FictitiousThing type, is both logically valid and practically feasible.

You are right that we should think a bit more about a default mechanism in schema.org. We already have a few parts with weakly defined notions of defaults. For example, if a unit of measurement code is missing in the GoodRelations part of schema.org, one can typically assume that it is "C62" for "no unit" or "piece".

Of course, we do not want fictitious places to show up on Google Maps or Google Places, nor see fictitious products and services in Google Shopping. But since search engines use 200+ signals to judge the meaning and relevance of a Web resource, I do not think that adding a mechanism for fictitious things, in combination with an informally agreed default of an entity being real, will cause any real problems. Plus, fictitious things could become real at any later point in time. 

Best
Martin




On 21 Oct 2014, at 09:27, Peter F. Patel-Schneider <pfpschneider@gmail.com> wrote:

> http://schema.org/Person is exceptional in this regard.
> 
> I would hope that items belonging to nearly all schema.org classes, for example, http://schema.org/TVEpisode, are actual things in the real world.
> 
> Under this hope, the absence of a claim that something is fictional is an indication that it is real.
> 
> peter
> 
> 
> On 10/20/2014 05:30 AM, Dan Brickley wrote:
>> On 20 October 2014 13:14, Peter F. Patel-Schneider
>> <pfpschneider@gmail.com> wrote:
>>> The essence of these proposals is that there is some class or property that
>>> changes the meaning of something else. My worry is that producers and
>>> consumers will need to understand all such classes and properties before
>>> they can use schema.org.
>> 
>> I agree; such mechanisms ought to add knowledge, not change it.
>> 
>> If all you know is that something is a <http://schema.org/Person>, you
>> don't know if they're alive, dead, undead, or fictional. If all you
>> know is that something is a <http://schema.org/Event> or
>> <http://schema.org/Action>, you don't know whether or when it
>> happened. If all you know is that something is a
>> <http://schema.org/Place>, you don't know how long it's been there,
>> whether it's still there, how long it'll be around for, etc., etc.
>> 
>> It would be a mistake to take the absence of a claim that something is
>> fictional as an indication that it is "real", non-fictional etc. (both
>> slippery notions anyway). There are lots of processes by which triples
>> can 'drop off' a graph in some information pipeline, with SPARQL-based
>> extraction being the most obvious.
>> 
>> Dan
>> 
>>> peter
>>> 
>>> 
>>> 
>>> On 10/20/2014 04:43 AM, Dan Brickley wrote:
>>>> 
>>>> On 20 October 2014 10:56, Wallis,Richard <Richard.Wallis@oclc.org> wrote:
>>>>> 
>>>>> +1.
>>>>> 
>>>>> Is it time to resurrect my FictionalThing Type proposal?
>>>>>         http://www.w3.org/wiki/WebSchemas/FictionalThing
>>>>> 
>>>>> It was an attempt to introduce a simple way, through multi-typing, to
>>>>> identify any Thing that could be fictional.  These discussions often
>>>>> centre
>>>>> around people/characters, but fictional-ness spreads way beyond people to
>>>>> organisations, countries, planets, languages and lumps of rock.  It
>>>>> included
>>>>> a property to reference a [real] Thing that the fictional is a
>>>>> representation of.
>>>> 
>>>> 
>>>> Could it make more sense to make this relational - fictionallyAbout or
>>>> similar - so that the relevant CreativeWork is included in the
>>>> description. This might make it easier to handle fictitious accounts
>>>> of real world entities. --Dan
>>>> 
>>> 
> 

Received on Tuesday, 21 October 2014 09:51:16 UTC