LISTSERV mailing list manager LISTSERV 16.0

Help for EUROPEANA-TECH Archives


EUROPEANA-TECH Archives

EUROPEANA-TECH Archives


EUROPEANA-TECH@LIST.ECOMPASS.NL


View:

Message:

[

First

|

Previous

|

Next

|

Last

]

By Topic:

[

First

|

Previous

|

Next

|

Last

]

By Author:

[

First

|

Previous

|

Next

|

Last

]

Font:

Proportional Font

LISTSERV Archives

LISTSERV Archives

EUROPEANA-TECH Home

EUROPEANA-TECH Home

EUROPEANA-TECH  August 2014

EUROPEANA-TECH August 2014

Subject:

Re: "Lossless" EDM - erring towards data quality

From:

Antoine Isaac <[log in to unmask]>

Reply-To:

Antoine Isaac <[log in to unmask]>

Date:

Fri, 1 Aug 2014 10:29:20 +0200

Content-Type:

text/plain

Parts/Attachments:

Parts/Attachments

text/plain (94 lines)

Hi Dominic,

Yes, these are very good points, and a great part of Europeana's quality efforts should be to find ways to win heart and minds as you suggest.

And I foresee that to a great extent, if we create 'data checking rules' they would be more conceived as a motivational factor. We could also give more boosting to good quality metadata records, which will give more context to the users - trying to ensure that users first see something that give them information on the rest.

Still we will have to enforce some basic rules, such as access to digitized representations of good enough quality (see all the objects for which we can't even generate a thumbnail), persistent access (dead links are still plaguing our ecosystem). And the minimal descriptive metadata which would allow the matching of instances as you mention - or at least some form of clustering, as projects like PATHS or even ourselves have experimented with [1].

By the way I also sense that Christian-Emil's point goes in the right direction: it's acceptable to have title-less object. But there should still be some other information that would allow to identify it for a user (e.g. a researcher), such as the 'Hon. III. ann. XI. ep. 367'.
Actually the rule in EDM now is that there should be either a dc:title or a dc:description always present in a metadata record. It is perhaps a bit strict, but finding an alternative would require quite some research.

Cheers,

Antoine

On 7/31/14 3:54 PM, Dominic Oldman wrote:
>
> Dear Antoine,
>
> Data aggregation systems need to benefit the data providers. If they provide benefits to the data providers then the users (who should include the data providers) will benefit as a result. This is at the heart of the problem. How do you make sure that data providers are not just suppliers of information but that they actually get significant benefits from the aggregation so that it not only encourages them to send more data but the resource is valuable enough to them to become users themselves, and are therefore are interested in its sustainability? How do you make the data aggregation essential and an extension of the provider's own infrastructure in which they can produce better services?
>
> This argument works within institutions themselves. How do you get curatorial staff interested in entering more data on the institution's objects. They already do a lot of digisation for standalone research projects. You ensure that what comes out of the other end is invaluable for their own work. If it is just an inventory list then the incentive is not there. If it is of research quality then, Bobs your uncle <http://en.wikipedia.org/wiki/Bob%27s_your_uncle>. Therefore, we now offer curators at the BM data that can be used as the basis for research - and interest has increased.
>
> The fundamental thing is this;
>
> You don't need all providers to offer the same level of information. That's the great benefit of big data and aggregation - "The whole is greater than the sum of its parts". You simply need to concentrate on the information that is being submitted being capable of a high degree of semantic integration with other datasets.  It doesn't matter so much whether one institution provides a little and one provides a lot, or indeed if some organisation provide nothing. The point is that together - as long as the data has been given the correct level of context and semantics - the resource as a whole is far richer and therefore far more useful for reuse. Those who provided nothing initially and who simply use the data others provide (because it is good data) may, as a result, offer data back in the future produced from the data they imported and subsequently enriched.
>
> I have heard the argument before from some institutions that it is not worth them transferring data to CRM because they don't have the same level of data (fields) as the BM. This completely misses the point. What they do have might be a crucial piece of information, or the piece of a puzzle that resolves, for example, a matching of instances. The Tate, for example, do not record as much as the BM but it is clear from their collection online that their data (perhaps just one field) makes significant connections to BM data, connections that are never exposed. It matters not that they don't have as many fields as the BM. Combined with many other datasets the connections can be made that create a more valuable resource. In this view a smaller resource would provide greater re-use benefits, and a service with greater benefits will end up being a larger resource in the end because people will want to use it.
>
> Cheers,
>
> Dominic
>
> PS. Here is a link to the latest CRM Primer v1.1 - Would welcome your (anyone else's) comments.
> <http://www.researchspace.org/file-cabinet/CRMPrimer_v1.1.pdf?attredirects=0&d=1>http://www.researchspace.org/file-cabinet/CRMPrimer_v1.1.pdf?attredirects=0&d=1
>
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
> *From:* Antoine Isaac <[log in to unmask]>
> *To:* [log in to unmask]
> *Sent:* Thursday, July 31, 2014 10:59 AM
> *Subject:* Re: "Lossless" EDM - erring towards data quality
>
> Hi Martin,
>
>
>  >>
>  >> I have to say that Europeana's main problem is not really the deliberate reduction of data, it is that the data is just not present. How about sending thousands of items with the same non-informative title (like 'picture')?
>  > That's another side of the same medal: Many things have very few data, few things have many data.
>  >> I am quite sympathetic to your stance of making no property recommended, but I still believe the value of an aggregator is also to try and offer to data re-users access to a data ecosystem that has some minimum coherence.
>  > Sure! How do we know in which scope such coherence exists? What is the size of the ecosystem(s) ?;-)
>
>
> A trick point indeed, as ou know the sort of scale we're talking about ;-)  I'd say that a minimum scenario would be able to find the object by one relatively simple way (a decent combination of who what where when criteria). I have seen some items that are really not findable at all. Very interesting objects, but there's just not enough description.
>
>
>  >> and there are many cases that do require basic metadata, like title or subject. Actually if you have an example of application that happily mixes title-less objects with objects with titles, please send it, I am curious!
>  > All archaeological collections: Some famous statues have titles, most have not.  The titles are relatively irrelevant, the find-spot/site, approximate(!) date and excavation event being the most important thing.
>
>
> Yes, I know that such data exist. My question was about the sort of application that could exploit them (e.g., a portal for general audience or researchers) and mixed with data that has different profiles. I can imagine fairly well that statues without titles won't be an issue in a system where archaeologists would search based on the place, the time and the type of objects (and even then , what if the type is missing?)
>
> Best,
>
> Antoine
>
>  >>
>  >> On 7/31/14 11:00 AM, martin wrote:
>  >>> Dear Antoine,
>  >>>
>  >>> I believe we have to put an end to the idea that metadata at the aggregator level are supposed to have
>  >>> a completeness to be functional. With the CRM we are taking strictly the position that NO property is
>  >>> recommended. ONLY in a particular environment/application completeness makes any sense.
>  >>>
>  >>> For primary evidence (museum objects etc.), the completeness is restricted to what is known.
>  >>> For application, completeness is a research program, and again not a data prerequisite.
>  >>>
>  >>> Particular communities should develop concepts of what are relevant data in their domain, not
>  >>> Europeana. The current catastrophy of Europeana is that data are deliberately reduced by providers
>  >>> to fit central recommendations, which cannot differentiate adequately.
>  >>>
>  >>> Relevance cannot be expressed in terms of properties. It is a question of reason.
>  >>>
>  >>> Librarians are a particular domain. They typically (but not always) have "minimal data" about all objects, which do not
>  >>> make any sense for archaeology for example.
>  >>>
>  >>> I think this is an educational question for the user communities, otherwise Europeana will hardly
>  >>> steer free of this mess ;-) . Who will start the education? Who will make the theory to be taught?:-(
>  >>>
>  >>> Best,
>  >>>
>  >>> Martin
>  >>>
>
>

Top of Message | Previous Page | Permalink

Advanced Options


Options

Log In

Log In

Get Password

Get Password


Search Archives

Search Archives


Subscribe or Unsubscribe

Subscribe or Unsubscribe


Archives

September 2021
August 2021
July 2021
June 2021
May 2021
April 2021
March 2021
February 2021
January 2021
December 2020
November 2020
October 2020
September 2020
August 2020
July 2020
June 2020
May 2020
April 2020
March 2020
February 2020
January 2020
December 2019
November 2019
October 2019
September 2019
August 2019
July 2019
June 2019
May 2019
April 2019
March 2019
February 2019
January 2019
December 2018
November 2018
October 2018
September 2018
August 2018
July 2018
June 2018
April 2018
March 2018
February 2018
January 2018
December 2017
November 2017
October 2017
September 2017
August 2017
July 2017
June 2017
May 2017
April 2017
March 2017
February 2017
January 2017
December 2016
November 2016
October 2016
September 2016
August 2016
July 2016
June 2016
May 2016
April 2016
March 2016
February 2016
January 2016
December 2015
November 2015
October 2015
September 2015
August 2015
July 2015
June 2015
May 2015
April 2015
March 2015
February 2015
January 2015
December 2014
November 2014
October 2014
September 2014
August 2014
July 2014
June 2014
May 2014
April 2014
March 2014
February 2014
January 2014
December 2013
November 2013
October 2013
September 2013
August 2013
July 2013
June 2013
May 2013
April 2013
March 2013
February 2013
January 2013
December 2012
November 2012
October 2012
September 2012
August 2012
July 2012
June 2012
May 2012
April 2012
March 2012
February 2012
January 2012
December 2011
November 2011
October 2011
September 2011
August 2011
July 2011
June 2011
April 2011
March 2011
February 2011
January 2011

ATOM RSS1 RSS2



LIST.ECOMPASS.NL

CataList Email List Search Powered by the LISTSERV Email List Manager