See also: IRC log
<trackbot> Date: 17 November 2009
<scribe> scribenick: mhausenblas
Previous: http://www.w3.org/2009/11/10-RDB2RDF-minutes.html
<Ahmed> I am back...
Ahmed: roll call
PROPOSAL: accept minutes from last meeting http://www.w3.org/2009/11/10-RDB2RDF-minutes.html
<hhalpin> +1
<jsequeda> +1
<Ahmed> +1
<Marcelo> +1
<angela_UNITN> +1
RESOLUTION: approved minutes from last meeting
Ahmed: scribe - shall we rotate?
<hhalpin> The command is "Zakim, pick a scribe"
<hhalpin> I'm happy to make a scribelist.
<hhalpin> Or we could do volunteering...
<hhalpin> ACTION: hhalpin to make a draft scribelist on wiki [recorded in http://www.w3.org/2009/11/17-RDB2RDF-minutes.html#action01]
<trackbot> Created ACTION-15 - Make a draft scribelist on wiki [on Harry Halpin - due 2009-11-24].
http://www.w3.org/2001/sw/rdb2rdf/wiki/Main_Page#Liaisons
sorry, http://www.w3.org/2001/sw/rdb2rdf/wiki/Liaisons
Michael: please review and add your name there
http://www.w3.org/2001/sw/rdb2rdf/wiki/Requirements/DDLCoverage
<scribe> ACTION: Jsequeda to init DDL coverage on Wiki [recorded in http://www.w3.org/2009/11/17-RDB2RDF-minutes.html#action02]
<trackbot> Created ACTION-16 - Init DDL coverage on Wiki [on Juan Sequeda - due 2009-11-24].
<Souri> We would like to present our approach/proposal to the WG on 01-Dec-2009 if the slot is available.
Souri, please put it on the Wiki at http://www.w3.org/2001/sw/rdb2rdf/wiki/Initial_Round_of_Presentations and send out to the group
<hhalpin> Souri, that date looks open
<hhalpin> http://www.w3.org/2001/sw/rdb2rdf/wiki/Initial_Round_of_Presentations
ISSUE-2?
<trackbot> ISSUE-2 -- R2RML serializations -- OPEN
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/2
<Souri> Thanks. I'll add our name to the schedule. We'll put the slides later.
thanks, Souri
<hhalpin> Just edit the wiki-page.
<hhalpin> And I think EricP also has a presentation he wants to give
<ericP> latest databinding draft
<hhalpin> http://www.w3.org/2002/ws/databinding/
<hhalpin> http://www.w3.org/TR/xmlschema-2/
Ahmed: supporting vendor-specific data types
<scribe> ACTION: mhausenb to create an issue for vendor-specific data types [recorded in http://www.w3.org/2009/11/17-RDB2RDF-minutes.html#action03]
<trackbot> Created ACTION-17 - Create an issue for vendor-specific data types [on Michael Hausenblas - due 2009-11-24].
<hhalpin> http://www.w3.org/2001/sw/rdb2rdf/Mtx-RDB2RDF.pdf
<hhalpin> That's the slides!
<hhalpin> Is it batla?
balta: we're a privately-held
company producing semantic technology
... sectors: financial services, public sector, manufacturing,
life sciences
slide 4
batla: the Metatomix
approach
... embrace the distributed enterprise landscape
... providing a way to generalise schema
... esp. important re distributing query (leveraging rules and
semantics)
slide 5
batla: listing some tools we have
(Eclipse-based)
... editor for ontology, RDF, rules/policy
... mapping tool data sources to RDF and other way round
slide 6
batla: three RDB2RDF specific
tools
... Ontology Editor Import Tool, Eclipse Mapping Tool, and the
Mapping in Mtx Discovery
slide 7
batla: allows for seeding an ontology with a representative relational schema
slide 8
batla: example
... seeding the ontology
Orri: how do you, for example, handle consolidation throughout various CRM systems?
batla: we would specific
ontologies for each CRM then
... and the create subclass/superclasses to connect different
classes in the specific ontologies
Ahmed: so, you need mapping from local to other local ontologies, right?
batla: yes
slide 9
batla: our tool set runs in
Eclipse
... the first step in “importing” an ontology from a relational
schema is to connect to the data source. In this case we are
connecting to a MySQL database.
slide 10
batla: After connecting to the data source, the wizard displays the schema, allowing the user to select the tables and columns to be included in the ontology.
<ericP> batla: demo ontology uses the northwind database
slide 11
batla: Inevitably, the ontology
will need to be augmented - remember the step of “importing”
the ontology from the database is a way to “seed” the
ontology.
... here, we are selecting an entity, NwCustomers and adding
new properties – contactLastName and contactFirstName as
datatype properties
... we have also a graph viewer in the tools
slide 12
batla: This slide displays the new ontology in a graph visualization. The first image shows how the application created a relationship between the nwOrders and nwCustomers entities through the customerID object property (nwOrders customerID nwCustomers). The second image displays the new properties we added in the previous step.
<ericP> slide 13, i suspect
slide 13
batla: Eclipse-based Mapping Tool allows mapping from relational schemas to an ontology in a click-and-drag paradigm and generates xml-based mapping file, executed at runtime for real-time RDF generation
slide 14
batla: 1. Configure the
Input
... The end goal is to create RDF from RDB. In order to that we
must create a mapping from the database to the ontology we just
created/augmented.
... this mapping is a design time step and when run, generates
RDF that “fits” the ontology.
... The first step in defining this mapping is to configure an
input ; this involves the identification of the input
source
slide 15
batla: 2. configure output
slide 16
batla: 3. Create the Mappings
jsequeda: first you create the ontology and then you augment it again?
batla: yes, on purpose
... key point is that while you seed at the beginning, the
runtime stuff happens dynamically (automatically)
... using regex, etc the splitting functions can be
implemented
slide 17
batla: Final mapping (merge/split)
slide 18
batla: an JDBC MSG example
slide 19
batla: resulting RDF/XML
slide 20
<Zakim> ericP, you wanted to ask how this mapped view is queried
batla: A graph representation of
the new RDF resource generated from our mapping.
... idea is to have bi-directional data flow, yes
slide 21
batla: discovery - bridges the
gap between structured and unstructured data
... rest of the slide deck is mainly about Metatomix discovery
tools and their usage
jsequeda: how complex can the mapping be?
batla: we have some sophisticated function, there's an API as well to extend it
<hhalpin> is thinking what would be useful would be to focus on how the various tools would also benefit (or not) from a RDB2RDF standard, and what features of this proposed standard they would already support.
<hhalpin> the thought was provoked by all the features that Metamotix has.
<jsequeda> +1
<hhalpin> thanks!
[adjourned]
This is scribe.perl Revision: 1.135 of Date: 2009/03/02 03:52:20 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/Souru/Souri/ Succeeded: s/rest of the/ batla: rest of the/ Succeeded: s/extend ti/extend it/ Found ScribeNick: mhausenblas Inferring Scribes: mhausenblas Default Present: angela_UNITN, whalb, mhausenblas, metatomix, jsequeda, EricP, +1.562.686.aaaa, [IPcaller], Souri, orri, Ashok_Malhotra, hhalpin, BenS, Seema, nunolopes Present: angela_UNITN whalb mhausenblas metatomix jsequeda EricP +1.562.686.aaaa [IPcaller] Souri orri Ashok_Malhotra hhalpin BenS Seema nunolopes Regrets: Richard Cyganiak Agenda: http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2009Nov/0036.html Found Date: 17 Nov 2009 Guessing minutes URL: http://www.w3.org/2009/11/17-RDB2RDF-minutes.html People with action items: hhalpin jsequeda mhausenb[End of scribe.perl diagnostic output]