RDB2RDF Working Group Teleconference -- 06 Apr 2010

<trackbot> Date: 06 April 2010

minutes:

http://www.w3.org/2010/03/30-rdb2rdf-minutes.html

APPROVED: minutes from 30th of March.

<scribe> scribe: hhalpin

<scribe> ScribeNick: hhalpin

Going over use-cases

See e-mail from Dan/Juan textified by EricP

<MacTed> +1

http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2010Apr/0004.html

<MacTed> 3 case primitives still not in there

<MacTed> ?

Ashok: I am concerned this use-case and requirement doc is too large!

Ahmed: I do tend to agree

Daniel: It's about 26 pages printed

ericP: Can we just remove redundant use cases

MacTed: There are 3 case primitives that have not been delimited, and we can pick one of each of them to make document short and sweet

ericP: there are a number of reasons why one would want relational data on the rdb2rdf and a number of expressivity requirements

<juansequeda> There is a XLS doc attached that shows how all the use cases are categorized per the taxonomy that we presented

ericP: the objections are that its too long.

<juansequeda> I just put the XLS on a google doc: http://spreadsheets.google.com/pub?key=tIJFfWcie4XSod7jcnMaBaQ&output=html

<MacTed> hhalpin - I'm guessing that's your keyboard banging ... can you mute that or move your mic or something?

<MacTed> hhalpin - it went quiet as soon as you asked him to listen :-/

ericP: that we are not motivating the project as much
... what do we try to communicate to people?
... if we have illustrative text and some analysis, we are community a service

Ahmed: Should we still requirement?

Daniel: Requirements need to be cover the issues

http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2010Apr/0000.html

<juansequeda> http://spreadsheets.google.com/pub?key=tIJFfWcie4XSod7jcnMaBaQ&output=html

+1 going through this taxonomy

daniel: this is a potential way to structure the use-case document
... this taxonomy is our technical requirement
... which use-case cover which requirement
... and of course more than one use-case per requirement needs to be deleted
... but we need to cover these requirements

MacTed: I'm looking at the spreadsheet and seeing 5 examples of schema joining
... one semi-structured to mashup
... nothing showing structured to semi-structured
... let's go through this primitive use-cases
... and make sure we have those as headings

ericP: not sure these are more important than expressivity requirements.

MacTed: but the expressivity flow from the use-cases, not lead to them.

daniel: would like to see the use-cases connected to pieces of expressivity

soeren: do we have to focus on unstructured information?
... isn't that beyond scope?

unstructured?

Ahmed: But it is part of the motivation
... just structured is a bit short-sighted

MacTed: it's not a matter of focussing
... they elaborate on the motivation
... my motivation is to join structured and unstructured
... my use-case is I have Oracle database and a website data.
... but there's ways we could say that there's a way to do that.

ericP: there's a formulaic way to create node identifiers, but we need to explain why there are less formulaic ways that are needed
... we need to explain the combination of these two tables create a new kinda table
... and microparsing or not
... we need these kinds of requirements

MacTed: But aren't these implementation details?
... in one case, the thing I want to join to is more RDB data, or RDF already, or something that could be RDF.

Ahmed: if it's a use-case where RDB2RDF is used for integration of structured and unstructured data.
... the main effort is moving structured and unstructured data.
... this is the main problem, structured data and structured data we already do in the database world
... so I'd like to see the motivation, even though it's larger than the charter
... is fine.
... even if the unstructured does not part of the charter

MacTed: not on same page
... degree to which we want to push away the use-case to reduce to move to RDB data to RDF
... but requirements should be limited

Ashok: not motivating the use-case, just requirement

MacTed: we are not focussing on how unstructured data to RDF.

Soeren: How do you distinguish this

Ahmed: We have a relational database
... and we have HTML pages and PDF pages
... and we have a newsfeed that's real-time
... and we want to combine these in real-time
... the main motivation for me is to deal with RDF
... is that one.
... we can highlight this motivation the way we want
... but we don't need a separate requirement for this.

I'd say its a continuum between structured data and unstructured data

and we can just assume

Ahmed: 15 percent relational data, 85 percent unstructured, unstrucutred growing faster
... and so we must integrate over this, and that could be motivating
... and we have a small piece of this larger problem

juan: our numbers are kinda backwards

Ahmed: I'm referring to well-understood numbers

MacTed: +1 Ahmed

let's send an e-mail about these numbers out

Ahmed: the question is unstructured in the universe as opposed to just on Websites, I think your numbers are from websites juan
... Are we converging here?

MacTed: From my part we've converged.

Ashok: I think we have converged.
... but would like a better feeling for what people are talking about unstructured data

text, XML, etc?

RDF+XML = semi-structured

relational data = structured

text and tags (including text inside documents) = unstructured

and unstructured text inside semi-structured data.

Ahmed: so it seems we have consensus here

daniel: we need to finish out that spreadsheet

and write the preambles

the use-cases are of vastly different sizes

Ahmed: Angela would do motivation section

can we repeat whose working on what?

ericP: we can use tables in existing wiki
... do you want a template?

MacTed: wiki tables are a nightmare.

we could do tables by hand?

in HTML

ericP: someone could make a template for wiki?

MacTed: we can keep in spreadsheet

ericP: access control in google spreadsheet

juan: if you want to edit it, send me an e-mail

ericP: that works for me.

<scribe> ACTION: ericP to structure document motivation, use-cases, and requirements [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action01]

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

<scribe> ACTION: eric to structure document motivation, use-cases, and requirements [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action02]

<trackbot> Created ACTION-41 - Structure document motivation, use-cases, and requirements [on Eric Prud'hommeaux - due 2010-04-13].

<scribe> ACTION: angela to draft motivation text for use-case document [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action03]

<trackbot> Created ACTION-42 - Draft motivation text for use-case document [on Angela Fogarolli - due 2010-04-13].

<scribe> ACTION: angela and daniel to work on use-case section of document [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action04]

<trackbot> Created ACTION-43 - And daniel to work on use-case section of document [on Angela Fogarolli - due 2010-04-13].

<scribe> ACTION: eric to make sure spreadsheet is complete for their use-cases [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action05]

<trackbot> Created ACTION-44 - Make sure spreadsheet is complete for their use-cases [on Eric Prud'hommeaux - due 2010-04-13].

<scribe> ACTION: soeren to make sure spreadsheet is complete for their use-cases [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action06]

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

<scribe> ACTION: soren to make sure spreadsheet is complete for their use-cases [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action07]

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

<scribe> ACTION: sauer to make sure spreadsheet is complete for their use-cases [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action08]

<trackbot> Created ACTION-45 - Make sure spreadsheet is complete for their use-cases [on Sören Auer - due 2010-04-13].

<scribe> ACTION: jsequada to make sure spreadsheet is complete for their use-cases [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action09]

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

<scribe> ACTION: jsequeda to make sure spreadsheet is complete for their use-cases [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action10]

<trackbot> Created ACTION-46 - Make sure spreadsheet is complete for their use-cases [on Juan Sequeda - due 2010-04-13].

<scribe> ACTION: MacTed to help with use-case restructuring and motivation text [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action11]

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

<MacTed> my w3 handle is tthibodeau (I believe)

MacTed: would like to work on the motivation and use-case structuring

<scribe> ACTION: angela to make sure spreadsheet is complete for their use-cases [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action12]

<trackbot> Created ACTION-47 - Make sure spreadsheet is complete for their use-cases [on Angela Fogarolli - due 2010-04-13].

<scribe> ACTION: leef to make sure spreadsheet is complete for their use-cases [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action13]

<trackbot> Created ACTION-48 - Make sure spreadsheet is complete for their use-cases [on Lee Feigenbaum - due 2010-04-13].

<scribe> ACTION: thibodeau to work on motivations and use-case restructuring [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action14]

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

<scribe> ACTION: tthibodeau to work on motivations and use-case restructuring [recorded in http://www.w3.org/2010/04/06-rdb2rdf-minutes.html#action15]

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

trackbot, meeting adjourned

<trackbot> Sorry, hhalpin, I don't understand 'trackbot, meeting adjourned'. Please refer to http://www.w3.org/2005/06/tracker/irc for help

trackbot, end meeting

- DRAFT -

RDB2RDF Working Group Teleconference

06 Apr 2010

Attendees

Contents

Going over use-cases

Summary of Action Items

Scribe.perl diagnostic output