Hello all,
I was wondering, now that Wikidata breaks the assumption that pages store wikitext, has it been considered that Wikidata concept pages could actually use schema.org markup? Wikidata could possibly read in data from other web pages which use schema.org markup, but what about the Wikidata pages themselves having the markup?
Max Klein
Wikipedia in Residence
kleinm@oclc.org
+17074787023
Hi Max - why do you say "now that Wikidata breaks the assumption that pages store wikitext" ?
Thanks
On 14.09.2012 08:06, Klein,Max wrote:
Hello all,
I was wondering, now that Wikidata breaks
the assumption that pages store wikitext, has it been considered that Wikidata concept pages could actually use schema.org markup? Wikidata could possibly read in data from other web pages which use schema.org markup, but what about the Wikidata pages themselves having the markup?
Max Klein
Wikipedia in Residence
kleinm@oclc.org
+17074787023
On Fri, Sep 14, 2012 at 3:59 PM, jmcclure@hypergrove.com wrote:
Hi Max - why do you say "now that Wikidata breaks the assumption that pages store wikitext" ?
Pages can now have a type and some pages will be wikitext but some will be other types like structured metadata. Rendering, storage format, diff generation, etc. will depend on page type. (AIUI; only 98% certain about all of that. other people can surely elaborate)
-Jeremy
Hi Jeremy and Max,
1. Does the wikidata design not entail template/parser calls which can be inspected by editing the page and seeing *all* data passed for storage to/by the wikidata engine? Are you saying that wikidata's data entry interface is to be the only method to be available to *humanly* see what data is being stored?
2. About "page type" - where is this discussed? I look in the RDF info for Wikidata and all I see is
w:Berlin s:Population Berlin:Statement1 . Berlin:Statement1 rdf:type o:Statement . <snip/>
I see NO
w:Berlin rdf:type some-type-as-asserted-by-jeremy-and-max.
Would you please provide a link or pointer to documentation about these "types"?
Thanks - john
On 14.09.2012 09:02, Jeremy Baron wrote:
On Fri, Sep 14,
2012 at 3:59 PM, jmcclure@hypergrove.com wrote:
Hi Max - why do
you say "now that Wikidata breaks the assumption that pages store wikitext" ?
Pages can now have a type and some pages will be
wikitext but some
will be other types like structured metadata.
Rendering, storage
format, diff generation, etc. will depend on page
type. (AIUI; only
98% certain about all of that. other people can
surely elaborate)
-Jeremy
_______________________________________________
Wikidata-l mailing
list
Wikidata-l@lists.wikimedia.org
On Fri, Sep 14, 2012 at 5:56 PM, jmcclure@hypergrove.com wrote:
Hi Jeremy and Max,
[...]
Would you please provide a link or pointer to documentation about these "types"?
Maybe some of these questions are answered at https://www.mediawiki.org/wiki/ContentHandler
-Jeremy
Thanks for the link. So Max's original question was whether a "content-type'd page" containing (json-? xml-? rdf-?) -encoded data, can have markup conforming to schema.org ontologies? Seems that if xml/rdf is a "supported content type" then an opening does exist for such to be implemented however only by individual wikipedia sites; because it seems clear that wikidata will not process such rdf/xml in order to create wikidata statements that can be displayed within infoboxes.
It does raise the question of whether wikidata defines a content type for its json markup, that is, json markup that conforms to its SNAKs ontology? Is there a list of the content types somewhere? I looked but could not find one in ContentHandler.php. thanks - john
On 14.09.2012 11:00, Jeremy Baron wrote:
On Fri, Sep 14, 2012 at 5:56 PM,
jmcclure@hypergrove.com wrote:
Hi Jeremy and Max,
[...]
Would you please provide a link or pointer to documentation about
these "types"?
Maybe some of these questions are answered at
https://www.mediawiki.org/wiki/ContentHandler
-Jeremy
_______________________________________________
Wikidata-l mailing
list
Wikidata-l@lists.wikimedia.org
On Fri, Sep 14, 2012 at 6:37 PM, jmcclure@hypergrove.com wrote:
I looked but could not find one in ContentHandler.php. thanks - john
The links at the bottom of the page I just linked to work for me.
-Jeremy
Your response doesn't direct me to a list of content types ("supported formats"). But you know I'd suggest the answer to your question wrt schema.org, is found in the published RDF [1]:
w:Berlin s:Population Berlin:Statement1 . Berlin:Statement1 rdf:type o:Statement .
Note that we introduced two new properties, s:Population (s: like statement) and v:Population (v: like value) instead of the original p:Population property (mind the namespaces). The original property, in the p: namespace, is a datatype property that connects items directly with an integer value (i.e. the domain is item). The s: property, on the other hand, is an object property connecting the item with the statement about it." [1]
[1] https://meta.wikimedia.org/wiki/Wikidata/Development/RDF#Statements_with_qua...
The problem is this: "s:Population" is not published that I can find. So is the question will "s:Population" be mapped to some schema.org/Population rdf:Property definition, if there were such a thing defined at schema.org? But looking at schema.org, "population" is usually defined as a p:Population kind-of-thing, not an s:Population kind-of-thing. In wikidata, p:Population is an rdf:Datatype, s:Population an rdf:Property. Schema.org's population is a datatype, not a class in an ontology, so they're not fitting together.
But as I've said, it's a problem that s:Population is not defined. Is "s:Population" not closer to a collection, a bag, that has restricted content that is, a "bag of statements" (see quotation above), than it is to a functional definition of "Population" as something more like a "group of living individuals", "alive" as per some specified context? Where does THIS concept of "population" fit with <rdf:Property rdf:about=wikidata/s:Population/> ?
Hard to avoid that s:Population basically is just short-hand for "Population Statements". Using shorthand certainly has precedence, being that q:Qualifier is short-hand for a category, an owl:Class, and thus "q:Draft" is (very acceptable) short-hand for "Draft Things"... So we need a crisp definition of s:Population please!
Thanks - john
On 14.09.2012 11:47, Jeremy Baron wrote:
On Fri, Sep 14, 2012 at 6:37
PM, jmcclure@hypergrove.com wrote:
I looked but could not find
one in ContentHandler.php. thanks - john
The links at the bottom of
the page I just linked to work for me.
-Jeremy
_______________________________________________
Wikidata-l mailing
list
Wikidata-l@lists.wikimedia.org
On Fri, Sep 14, 2012 at 7:54 PM, jmcclure@hypergrove.com wrote:
Your response doesn't direct me to a list of content types ("supported formats").
Yeah, sorry, I misread as "I can't find a copy of the file called ContentHandler.php". nevermind. ;)
-Jeremy
It does raise the question of whether wikidata defines a content type for its json markup, that is, json markup that conforms to its SNAKs ontology? Is there a list of the content types somewhere? I looked but could not find one in ContentHandler.php. thanks - john
The content *model* Wikidata uses is the one with Statements and Snaks, described by the Wikidata data model spec. The content *format* used for serialization is JSON. The mapping from the internal model to JSON is currently not formally described, the internal JSON serialization is an implementation detail. The external JSON serialization as used by the web API is similar to, but not necessarily the same as the serialization used to represent Wikidata items internally. The JSON form used in the API is documented to some extend on the Extension:Wikibase page on mediawiki.org, but it's still subject to change.
-- daniel