Possible deliverable of this XG

Hi all,

I promised to send some thoughts regarding the RDB2RDF update log
mechanism I mentioned in my Triplify talk. Generally, I think the
following deliverables would be very good to have:

1. *Tool and approach review*
  * looks like we started with this already ;-)

2. *Methodology for transforming relational data into RDF*:
  * one time transformation:
    * depending on your environment/requirements use the tool X
  * live transformation, along the lines:
    * create an abstract mapping from your DB schema to RDF(S)
    * select a tool (i.e. D2RQ, Virtuoso, Triplify) depending on your
      environment/requirements
    * translate your abstract mapping into a tool specific one
    * deploy tool with mapping on your DB server

3. *Standard for publishing DB/RDF update logs*:

When RDB data is published on the Web e.g. as LinkedData it is important
to keep track of DB (and hence RDF) updates so crawlers know what has
changed (after the last crawl) and should be re-retrieved from that
endpoint.

Semantic Sitemap Extension [1] does not seem provide a solution.

Possible solutions:
  * ATOM/RSS feeds pointing to the updated data sources
    Advantag: established format for publishing chronological info
    Disadvantage: additional format, might not scale
  * LinkedData update vocabulary:

http://example.com/lod/updates
points via RDF links to:
http://example.com/lod/updates/2008
http://example.com/lod/updates/2007
...
point via RDF links to:
http://example.com/lod/updates/2008/Jan
http://example.com/lod/updates/2008/Feb
...
point via RDF links to:
http://example.com/lod/updates/2008/Jan/01
http://example.com/lod/updates/2008/Jan/02

...

http://example.com/lod/updates/2008/Jan/01/00/00
contain RDF links to the LinkedData documents updated on Jan 1st, 2008
at 00:00, e.g. following triples:

http://example.com/lod/updates/2008/Jan/01/00/00/123
updatedResource http://example.com/lod/users/SoerenAuer .

http://example.com/lod/updates/2008/Jan/01/00/00/123
updatedAt "20080101T00:00:01"^<xsd:dateTime> .

http://example.com/lod/updates/2008/Jan/01/00/00/123
updatedBy http://example.com/lod/users/SoerenAuer .

This mechanism as well as some base update log vocabulary (i.e. the
updatedResource, updatedAt, updatedBy properties) could be standardized
by this XG.

What do you think? Would be nice to discuss this on Friday shortly.

Best,

Sören

[1] http://sw.deri.org/2007/07/sitemapextension/

-- 

--------------------------------------------------------------
Sören Auer, AKSW/Computer Science Dept., University of Leipzig
http://www.informatik.uni-leipzig.de/~auer,  Skype: soerenauer

Received on Wednesday, 21 May 2008 20:50:01 UTC