<trackbot> Date: 29 June 2010

<mhausenblas> hehe

hey, wait a minute!

<juansequeda> dialing in a few min

<mhausenblas> scribenick: ericP

1. Admin

<mhausenblas> PROPOSAL: Accept the minutes of the F2F meeting, see http://www.w3.org/2010/06/20-rdb2rdf-f2f-minutes.html

<cygri> +1

<mhausenblas> RESOLVED: Accept the minutes of the F2F meeting, see http://www.w3.org/2010/06/20-rdb2rdf-f2f-minutes.html

Ashok: kudos to harry for cleaning minutes

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/track/actions/open

mhausenblas: i cleaned up f2f-related actions and changed the due date for remaining actions to next week

Report on SemTech 2010 panel

Ashok: souri, juansequeda and cygri were at the panel

juansequeda: the public is looking forward to a standard mapping language

mhausenblas: what sort of audience?

souri: good turn-out (more than SPARQL) in a large conference room
... no question about whether a standard is necesary
... heard "how will you do inference?"
... said "the inferred data can sit in a triple store and the asserted data in relational tables"
... that question from a fellow from gallery systems in new york (they manage museums)

juansequeda: audience was mostly industry. i talked with someone from chevron

cygri: after the panel, i heard questions about timeframe for spec and implementations
... indicates (industry) eagerness

mhausenblas: did any new faces show up with implementations or use cases?
... e.g. someone from wordpress giving a requirement

dan: heard a request that UC&R include OLAP application

juansequeda: i asked folks to look over UC&R and email on the comments list

<juansequeda> ACTION: dan to work on UC&R on OLAP application [recorded in http://www.w3.org/2010/06/29-rdb2rdf-minutes.html#action01]

<trackbot> Sorry, couldn't find user - dan

<mhausenblas> ACTION: Dan to work on UC&R on OLAP application [recorded in http://www.w3.org/2010/06/29-rdb2rdf-minutes.html#action02]

<trackbot> Sorry, couldn't find user - Dan

mhausenblas: did we hear from IBM or MS?

[no]

<juansequeda> ACTION: miranker to work on UC&R on OLAP application [recorded in http://www.w3.org/2010/06/29-rdb2rdf-minutes.html#action03]

<trackbot> Sorry, couldn't find user - miranker

mhausenblas: sounds like we're motivated to finish quickly

juansequeda: we heard "will there be write?"

<Ashok> Dan, what's your moniker on Zakim

juansequeda: answered "we're chartered for read-only"

<scribe> ACTION: ddanielm to work on UC&R on OLAP application [recorded in http://www.w3.org/2010/06/29-rdb2rdf-minutes.html#action04]

<trackbot> Created ACTION-69 - Work on UC&R on OLAP application [on Daniel Daniel Miranker - due 2010-07-06].

<mhausenblas> ACTION: Michael to slap Eric very soon [recorded in http://www.w3.org/2010/06/29-rdb2rdf-minutes.html#action05]

<trackbot> Created ACTION-70 - Slap Eric very soon [on Michael Hausenblas - due 2010-07-06].

<mhausenblas> close ACTION-70

<trackbot> ACTION-70 Slap Eric very soon closed

report of F2F meeting

Ashok: would have been better with more attendance, but we did what we could
... started with talk by juansequeda and dan about using the appropriate terminology vis a vis ontologies etc.
... action 68 records an action for them to write that up
... souri introduced SQL approach:
... .. 3 kinds of statements
... .. .. this class can be populated with this SQL view
... .. .. this *data* can be populated with this SQL view
... .. .. constraint statement
... we need to create a syntax and write this up as a starting point
... issue: must the class be expressed? (cygri's question)
... cygri discussed RDF-based approach
... made some progress
... RDF-based approach is really mapping SQL to a default mapping and then default to the desired ontology
... some of the functionality requires SPARQL11 (still in progress)
... we can translate from the SQL statement to the RDF syntax
... if this is possible, we don't need two totally difference approaches
... seemed like a good idea to me
... do we want to write this up?

ericP: not much to write; it's directMapping + perhaps some defined RIF or SPARQL functions to bi-map node transformations

juansequeda: dan and i chatted with Harold Boley
... (of RuleML, RIF)
... i think we're pretty close to done if we finish the direct mapping
... the syntax needs to be done, but the semantics is all done for us
... given Direct_Mapping_in_RIF, we already have the semantics

<MacTed> vindication! :-)

juansequeda: marcelo says we can write the semantics with datalog

cygri: of the semantics and syntax, the semantics is simple
... the syntax is still hard and undone
... whatever approach we come up with, the semantics should be do-able

juansequeda: i'd still like to create a group to write the semantics
... i'll work on it with dan and marcello

ericP: is http://www.w3.org/2001/sw/rdb2rdf/directGraph/ insufficient?

cygri: folks seem to mean two things:
... .. data mapping
... .. schema mapping

<mhausenblas> mute soeren

ericP: how is the generated schema used?

<soeren> sorry

juansequeda: it's the only source we have
... when you decide something is a class, you need to make it explicit
... the directGraph is insufficient 'cause it doens't include the schema

cygri: "insufficient" must mean insufficient for some purpose
... what use cases does the data mapping not meet?

dan: right, we don't need the schema semantics to be expressed in OWL or RDFS
... but if we're going to develop a meaning for this mapping, they must be represented

<cygri> ericP: having the schema explicit is nice because it serves as documentation for querying

<cygri> ericP: another use case could be: machines can do faceted browsing or assisted exploration of the data

<cygri> ericP: but i believe those are secondary to creating the data mapping itself

dan: after the vendors implement the standard, we need to allow folks *not* on the committee to e.g. say "this table name appears here"

<Souri> Without carefully going thru the high volume data (sitting in relational tables), the only other way I can think of writing a SPARQL query against that data is to look at the RDFS/OWL schema that is the basis of the data

dan: we can't do our job without saying how an implementor will map symbols from the data to the interface

cygri: but that's already written down

dan: i'd like ericP to assert some correctness semantics

<Zakim> ericP, you wanted to propose syntax

<soeren> +soeren

<soeren> +q

MacTed: a goal is to cross-join against e.g. SQL dbs
... is field db1.x == db2.y?
... making the schema available is required to enable these cross-joins
... i think the directGraph is uninteresting until we map to an interface graph

<cygri> +1 to what soeren said

http://www.w3.org/2001/sw/rdb2rdf/directGraph/

http://www.w3.org/2001/sw/rdb2rdf/directGraph/schema

<cygri> second one this? http://www.w3.org/2001/sw/rdb2rdf/wiki/Direct_Mapping_in_RIF

http://www.w3.org/2001/sw/rdb2rdf/directGraph/schema no longer 404s

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/track/actions/open

<mhausenblas> ACTION-66?

<trackbot> ACTION-66 -- Richard Cyganiak to document issues on the SQL-based approach due July 8 -- due 2010-07-06 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/66

pending actions

cygri: per 66, souri had some slides which gave an abstraction of the sql approach
... we discussed expressivity. i took an action for writing requirements for this approach

<mhausenblas> ACTION-67?

<trackbot> ACTION-67 -- Souripriya Das to write first draft of SQL-based approach due July 22 -- due 2010-07-06 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/67

Souri: per 67, will work on the SQL-based approach draft

http://www.w3.org/2001/sw/rdb2rdf/directGraph/Overview.xml

<mhausenblas> ACTION-68?

<trackbot> ACTION-68 -- Juan Sequeda to write note on Ways to Leverage Ontologies due July 8 -- due 2010-07-06 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/68

juansequeda: per 68, when we utter "ontology" we mean different things

AOB

<mhausenblas> PROPOSAL: Cancel call on 6 July 2010 due to holiday week in the US

<Souri> +1 to cancelling next week

<cygri> +1 to cancellign

ericP: if i include e.g. FOAF, DCIM, uniprot, in an interface graph, do i have one ontology which includes all these other ontologies?

juansequeda: that will come out with the document

mhausenblas: how shall we proceed this summer?
... i'll be gone mid-july to mid-august
... make up your minds and be ready to discuss next week

<mhausenblas> thanks a lot for scribing ericP !

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)

Found ScribeNick: ericP
Inferring Scribes: ericP
Default Present: EricP, +3539149aaaa, mhausenblas, +49.322.222.0.aabb, +1.603.897.aacc, soeren, Seema, cygri, MacTed, nunolopes, Ashok_Malhotra, juansequeda, angela, Souri
Present: EricP +3539149aaaa mhausenblas +49.322.222.0.aabb +1.603.897.aacc soeren Seema cygri MacTed nunolopes Ashok_Malhotra juansequeda angela Souri
Regrets: LiMa Wolfgang Boris Marcelo
Agenda: http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2010Jun/0063.html
Found Date: 29 Jun 2010
Guessing minutes URL: http://www.w3.org/2010/06/29-rdb2rdf-minutes.html
People with action items: dan ddanielm michael miranker