See also: IRC log
< antoine> Previous: 2011-03-10 http://lists.w3.org/Archives/Public/public-xg-lld/2011Mar/0055.html
RESOLUTION: To accept http://www.w3.org/2005/Incubator/lld/minutes/2011/03/10-lld-minutes.html
antoine:postponed until better conditions
antoine:Want to review the current sections as envisioned
... and the
remaining work that needs to be carried out.
... Report includes 6 items.
...
Nothing to say on executive summary.
... Next in the report is the "Benefits of LD for
libraries"
... will have a discussion on this in two weeks.
emma:Not sure we can commit to this date.
edsu:Haven't created a wiki page for it. As it is currently framed may be
asking a lot.
... (extracting benefits from clusters and use cases)
... a couple
of bullet points might be doable in two weeks.
< TomB> +1 to start with high-level bullet points
antoine:Thinks it is okay to start with that.
< emma> yes @ed, +1 for starting with bullet points
antoine:from a list of bullet points we could have an interesting discussion.
< emma> that may be ok for 31 march
< emma> ross said he would joint
< rsinger> i think i signed up for this, right?
< rsinger> ok, confirmation from emma :)
edsu:If anyone is interested, speak up.
< emma> suggests we record a formal action
< emma> +1
< rsinger> +1
< scribe> ACTION:edsu, rsinger, emma to create a few bullet points on the benefits of linked data in libraries for the call on March 31st [recorded in http://www.w3.org/2005/Incubator/lld/minutes/2011/03/17-lld-minutes.html#action01] [CONTINUES]
< edsu> pmurray: cheers
antoine:Next report item is "Use cases and requirements"
... to be
presented as a separate deliverable as a clean snapshot from the community
... we
don't have an owner for this deliverable
Call for owner of UC deliverable: http://lists.w3.org/Archives/Public/public-xg-lld/2011Mar/0038.html
antoine:this isn't intended to be extremely big.
... no one is
volunteering, so we will leave this open for a couple more weeks.
< monica> I am very sorry but I am over-committed in March-April - this needs to be finished quite soon doesn't it?
antoine:Next report item is "Available data (vocabularies, datasets)"
...
hoping to make progress on this for presentation on April 14th
... Next report section
is "Relevant technologies (as described in #6 in http://lists.w3.org/Archives/Public/public-xg-lld/2011Feb/0034.html)"
< emma> @monica : current schedule is april 7th but could be postponed (will be, anyway, if no owner)
antoine:Jeff has started a page on the wiki
< antoine_> http://www.w3.org/2005/Incubator/lld/wiki/Tools#Tools_List
< jeff_> http://www.w3.org/2005/Incubator/lld/wiki/Tools
antoine:on this topic there is a related ACTION:
< scribe> ACTION:Alex, Jeff, Martin, MichaelP elaborate on general purpose IT architecture for dealing with linked data with caching feature (short sketch for final report) [recorded in http://www.w3.org/2005/Incubator/lld/minutes/2010/10/24-lld-minutes.html#action04]
antoine:sent a message asking the authors if they wanted to reframe it into producing a section of "relevant technology" for the report
< antoine_> http://lists.w3.org/Archives/Public/public-xg-lld/2011Mar/0034.html
jeff:Is concerned that the message got caught in the spam folder.
... has
put bullets that were relevant in there. is looking for a mre relevant way to express
this.
antoine:Skiping over "Problems and limitations" the final section is "Requirements and recommendations"
kcoyle:Requirements and Recommendations follows from issues, so these two follow
< scribe> ACTION:Uldis and Jodi to create social uses cluster [recorded in http://www.w3.org/2005/Incubator/lld/minutes/2010/12/16-lld-minutes.html#action03] [CONTINUES]
< emma> http://www.w3.org/2005/Incubator/lld/wiki/Draft_issues_page
kcoyle:Started by consolidating all of the problems and issues from the use
cases
... distilled them down into categories and rewrote them as paragraphs
...
The important thing for today is to see if these 5 points resonate with everyone else.
... Do they capture the key points that have been brought out?
< emma> +1, this is a great start
kcoyle:the first main area is: "Linked Data is an emerging technology"
... the second point is "Library data is expressed in library-specific formats that
cannot be easily shared outside the library community"
... the third point is "Current
library data practices are expensive (and the true costs are unmeasured)"
... how
could we say that moving to LLD is a good return on investment.
... the fourth point
is "Library ecosystem is designed for stability and resists change"
... the fifth
point is "Library data may be protected by IP rights that prevent open publication"
< rsinger> yeah, that's great, karen
< antoine_> +1
< jeff_> +1
kcoyle:a question is "Are these too broad?"
emma:Doesn't think they are too broad; they are at the right level for the
report.
... hasn't looked in detail at the "Not Used" page, but how do we know we
haven't missed any (such as the quality of the data itself).
... Is everything
captured or do we think there are more details?
kcoyle:The subgroup went through the list and thinks everything has been captured.
< edsu> kcoyle++ # really nice job
kcoyle:we try to make statements about the data. e.g., the data is quality data, but it is text data not data data
antoine:Suggests everyone look at this document and see if they find their own issues in here.
rsinger:Wants to add someing to emerging technology. Libraries are used to
existing techology.
... it is going to be difficult for people to even envision the
cost when we have been using the same technology stack for 20 years.
kcoyle:That is a good point. Some of that may come through in the "library
ecosystem designed for stability" but will look at it to make sure it is in there.
...
previous changes have been minor compared to this change.
< rsinger> yes, definitely!
edsu:It is a nice summary. It is a good detail level.
< rsinger> (both edsu and TomB :)
edsu:in the section on the library ecosystem, did the group think about the library culture and compare it to the web culture?
< TomB> Ed: contrasting library culture and Web culture.
edsu:web culture is about "now".
< jeff_> the clash of terminology between the Web and libraries is part of the problem
kcoyle:Didn't get it in here, but thinks it is important to contrast the two
cultures
... jodi had come up with this, and had talked about how it is a necessary
difference
... it is a tension that wee need to point out.
< emma> seems to me this web vs. libs culture issue is addressed in the "library standards" paaragraph within "library ecosystem" , could be more explicit
< Zakim> TomB, you wanted to wonder whether there is a problem with "models" (e.g., FRBR), e.g., compatibility with the outside world
TomB:Do we have a category that addresses some of the issues of the
compatibility of linked data expressions being developed in the library world around FRBR
and RDA?
... affirming the need for library standards to play well with non-library
data.
... the contrast between traditional culture of defining data formats and the
semantic web culture of making statements bout things.
... there are a bundle of
issues that he is not sure have been captured in these five topics.
kcoyle:Maybe we can broaden the statement of library data being expressed in
library-specific formats.
... there are emerging standards, but they are following the
old model and don't look like linked data.
... we don't know how library functionality
and linked data are compatible. e.g. managing resources in the libraries.
... library
is an institution, it is not information. actual organizations that manage budgets and
inventory
... there is something else there, so we need to talk about what we mean by
library data.
... suggests that what we mean by library data is about the library
delivery functionally and not the data that does acquisitions and accounting
TomB:There are some types of data that sould be in silos for technical reasons.
That needs to be stated early in the report.
... we can make a distinction between
library data in the library and how it is exposed to the internet
kcoyle:Yes, and would like that to be near the top of the report. It introduces
some of the things that come into the report.
... that some data needs to be in silos
could come in here as well.
... maybe instead of having a section about IP rights,
could say there is library data that cannot be shared -- some is bibliographic and some is
not
TomB:Part of the issue belongs there and part belongs in 1.2.
< Zakim> emma, you wanted to speak about linked enterprise data
emma:Quite a restricted definition of linked data as -open- linked data.
... using the principles of linked data as an internal representation within the
enterprise.
< edsu> emma++ # linked data still useful in the enterprise ; sometimes i wonder if it's more useful there :)
< jeff_> emma++
emma:can we use linked data principles to build internal systems?
< TomB> +1 good point about LED - use of LD technology inside the firewall
emma:at the beginning people thought of web data as publishing; now we have websites using the same principles that are for internal only
kcoyle:That would be good to put at the beginning of the report, that our emphasis is on discovery data that could be shared but it can also be used in the enterprise as well.
< edsu> emma: we should remember that for the benefits page ... ah you just said it
emma:It is a benefit that can also improve the efficiency inside the organization.
kcoyle:Brings up the issue of not being able to say anything about
return-on-investement. Doesn't think we will have that.
... perhaps need to add linked
data as a benefit to the larger system.
edsu:Hesitant to add anything like that because we don't have enough practice yet to say something like that.
kcoyle:It starts to sound evangelical and not scientific.
edsu:...or even honest. It doesn't need to be statistically sound.
...
emma's point about linked data versus linked open data needs to be brought out. the two are
commonly thought of together.
... using these technologies doesn't necessarily entail
dumping your data out on the web
kcoyle:Good point. How do we want to say what we are addressing in this report?
... are we addressing linked data or open linked data?
< GordonD> We should address both open and closed LD
edsu:Personal opinion is that this is about linked data and libraries. There
are all kinds of libraries, even those that are closed off to the public. (e.g. corporate,
some gov't)
... we need to talk about the openness part of it.
... a lot of the
benefits accrue by sharing information. Something that libraries typically have done.
< monica> I think you can say that the report mainly addresses open/discovery linked data (because mainly that is what we have discussed) but acknowledge that data does not have to be open and can serve other purposes
< GordonD> Case study: personal identity management in institutions: closed (who borrows what books) vs open (who wrote the paper in the institutional repository)
kcoyle:We haven't really put an emphasis on users.
... one of the things
that FRBR does is put all of the justification on service to users.
... we've been
talking about the advantages of the data. we need to say somewhere early in the report that
we look at this because it improves services to users.
antoine:That is the idea of the use cases. For what the user benefits from.
kcoyle:So in the same way we went through the use cases looking for issues, have someone go through the use cases looking for advantages to users.
< GordonD> Use case clusters should have digested the benefits for users ...
edsu:It is in the same space as the "benefits of LD for libraries"
antoine:Supports TomB's widening suggestion about formats
< TomB> Antoine, I see it as a separate point, but that's a detail to discuss...
antoine:Supports adding Monica's IRC comment into 1.5
kcoyle:yes, and adding the benefits of open if you can do it
antoine:Suggests making it a bit more positive.
< edsu> :)
antoine:can be editorial changes.
... as much as possible, make
recommendations things that can be done, and that helps offset this.
kcoyle:Other than making the changes brought up today, should we move to the recommendations section? What is the next step?
antoine:There are still notes of things to be completed in the issues section.
kcoyle:The issues and recommendations will be an iterative process.
< emma> +1 for brainstorming on recommendations starting next week
< TomB> +1 flesh out the recommendations - agree this is an iterative process
antoine:This makes sense; starting with the recommendations.
< GordonD> Fine with me
< GordonD> I've got some time ...
< scribe> ACTION:everyone to take a look at the issues and make sure that points from the use cases are represented in the document [recorded in http://www.w3.org/2005/Incubator/lld/minutes/2011/03/17-lld-minutes.html#action04]
< Zakim> TomB, you wanted to ask for scribe for next week http://www.w3.org/2005/Incubator/lld/wiki/ScribeDuty
< kcoyle> http://www.w3.org/2005/Incubator/lld/wiki/ScribeDuty
< marcia> :-)
< scribe> ACTION:everyone (on the call and off) to send email message in the next week re brainstorming on important issues [recorded in http://www.w3.org/2005/Incubator/lld/minutes/2011/02/17-lld-minutes.html#action08] [DONE]
< scribe> ACTION:As a future topic for March 10, discuss the open questions in the second half of http://www.w3.org/2005/Incubator/lld/wiki/Library_standards_and_linked_data[recorded in http://www.w3.org/2005/Incubator/lld/minutes/2011/01/27-lld-minutes.html#action03] [CONTINUES]