Re: Generating XHTML/RDFa from Turtle to express mappings between vocabularies?

On May 13, 2012, at 7:25 AM, Thomas Baker wrote:

>> Documenting and publishing mappings
>> 
>>     Antoine has started work on an RDFa representation [1] of the 
>>     mappings in [2].  We will discuss this approach and address
>>     Kirsten's question [3,4] of how best we should incorporate new
>>     mappings into the set of mappings under consideration.
>> 
>>     Off-list, Dan has suggested that we approach mappings in the context of 
>>     usage patterns (application profiles).  He points out that with better 
>>     online documentation of both DCMI Metadata Terms and Schema.org, it should
>>     not be necessary to compile wiki pages such as [2] by hand and suggests that
>>     publication of mappings could therefore be simplified.
>> 
>>     [1] https://github.com/dcmi/schema.org/blob/master/mappings.html
>>     [2] http://wiki.dublincore.org/index.php/Schema.org_Alignment/Mappings_Details
>>     [3] https://github.com/dcmi/schema.org/issues/3
>>     [4] https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1202&L=dc-architecture&F=&S=&P=14738
> 
> Dear all,
> 
> On a telecon of the DCMI Schema.org Alignment Task Group tomorrow, we will 
> discuss how best to publish mappings between Schema.org and DCMI terms.
> 
> Initially, I pulled together a wiki page to show proposed mappings together
> with the definitions of the relevant terms [2, above].  This was a tedious
> process of cutting-and-pasting, and as Dan pointed out, it really highlights a
> need to improve the documentation of both vocabularies -- an observation that
> led to the current push to complete the task of modifying the scripts to
> generate DCMI Metadata Terms with RDFa.
> 
> In chats over the past few days, I have heard it suggested that RDF data could
> be maintained in Turtle and used to generate RDFa.  I wonder whether that could
> be a longer-term goal for the DCMI Metadata Terms work -- i.e., to maintain the
> source data in Turtle instead of in the ad-hoc XML markup we currently use [5];
> however, that would surely require a re-write of the scripts beyond what we
> could realistically undertake in the near term.
> 
> I wonder, though, whether such a workflow might be adopted for producing a
> simple Web page for DC/Schema.org mappings.  Antoine Isaac has started
> converting my tediously assembled wiki page [2] into RDFa [6,7], and this could
> be the basis for a more "presentable" way to publish the mappings, with
> document anchors [8] that would be citable in issue trackers.  As I understand
> it, however, the RDFa document would require no less hand editing than the wiki
> document.
> 
> Ideally, readable documentation for the vocabularies being mapped would be just
> a mouse-click away and it would be unnecessary to cut-and-paste definitions.
> In that case, the mappings could perhaps be maintained in Turtle or N-Triples
> and published in XHTML/RDFa -- as DCMI Metadata Terms will soon be, only with
> perhaps a simpler format, using simpler scripts.  Ideally, of course, we would
> be able to publish the mappings together with provenance information for each
> mapping -- who, when, and status -- and with HTML anchors that could be used
> for citing a particular mapping, for example in an issue tracker.
> 
> If any of you see how we might do this efficiently, please post your ideas.
> You would also be most welcome to participate in discussion on the
> dc-architecture mailing list [9] or join in the discussion on tomorrow's
> telecon (see joining information below).

The RDFa group has had success in maintaining information in Turtle, and using that to generate RDFa through templating mechanisms. I blogged about this when describing how JSON-LD is used to run the RDFa test harness [10].

Basically, we use this in two ways, the first is to take the hand-managed RDFa manifest kept in Turtle format [11] and transform it into a JSON representation, in this case JSON-LD [12], which looses none of the semantic content. This is used to dynamically construct the actual test driver using client-side templating [13] (simple browser-id based authentication required to avoid DDoS spammers).

The second way we use it is to coalate individual EARL reports into a comprehensive JSON-LD representation [14]. This is used with a simple server-side temlating tool to generate the HTML report marked up with RDFa, of course [15].

JSON is a great intermediate representation for working with HTML templates, and it can be automatically generated from most any RDF representation.

The resulting JSON can be easily put into a format that's convenient for iterating through classes and properties in an HTML templating engine. I usually do this using Ruby ERB, or Haml.

I'd be happy to work something up as a demonstration of doing this if there's interest.

Gregg

> Tom
> 
> [5] https://raw.github.com/dublincore/website/master/2012-05-21/xmldata/dcterms-properties.xml
> [6] https://raw.github.com/dcmi/schema.org/master/mappings.html
> [7] http://rdf.greggkellogg.net/distiller?format=turtle&in_fmt=rdfa&uri=https://raw.github.com/dcmi/schema.org/master/mappings.html
> [8] https://raw.github.com/dcmi/schema.org/master/mappings.html#schema:Organization_rdfs:subClassOf_dct:Agent
> [9] http://www.jiscmail.ac.uk/lists/dc-architecture.html
[10] http://greggkellogg.net/2012/03/20/the-use-of-json-ld-in-the-rdfa-test-harness
[11] http://rdfa.info/test-suite/manifest.ttl
[12[ http://rdfa.info/test-suite/manifest.json
[13] http://rdfa.info/test-suite/
[14] http://rdfa.info/earl-reports/earl.jsonld
[15] http://rdfa.info/earl-reports/

> On Sat, May 12, 2012 at 06:51:42PM -0400, Tom Baker wrote:
>> Schema.org Alignment Task Group telecon - 2012-05-14 11:00 EDT
>> 
>> This agenda:    http://wiki.dublincore.org/index.php/Schema.org_Alignment/Telecon_20120514
>> Chair:          Tom
>> Date:           Monday, 2012-05-14
>> Time:           11:00 AM Eastern Daylight Time
>> Dial-in:        +1-218-936-4141, participant Access Code 334034
>> IRC:            irc://irc.freenode.net/#dcmi
>> Mailing list:   http://www.jiscmail.ac.uk/lists/dc-architecture
>> Expected:       http://www.doodle.com/u3bh48x4f3p8db7r
>>                Tom, Antoine, Karen, Dan, Bernard, Kirsten, Corey
> 
> -- 
> Tom Baker <tom@tombaker.org>
> 

Received on Sunday, 13 May 2012 20:34:02 UTC