Hi Talat,
To provide you with some quantitative data - we are running a survey up here in Scotland as part of the ERIS Project, and have asked researchers/academics if they have ever had to make changes post publication.
Its not a huge sample, but out of the 232 responses we’ve had in, 12 (5.2%) have had to make changes.
Note, these results are not specifically asking if they have ever had to make edits to a record held in an institutional repository.
The following reasons for needing to make changes were cited;
1. Additional copyright had to be sought for images for the e-edition of an already published book. This was a rather arduous and, as it turned out, expensive process, which meant that most of the images had to be removed from the electronic publication.
2. I was asked to update an essay I'd contributed for a possible second edition. To date I've not managed to do this.
3. Likely to be revising one bit of data from a previous paper in the MS currently under revision - new experiments have shown a different conclusion. Don't know yet how this will be accepted by editor.
4. Make an amendment to an incorrect table in the results section
5. My MPhil was recently published into a book so I had to reformat according to the criteria of the publisher. There was a little extra effort.
6. Needed to issue an erratum once because of publisher misprints that I missed at proofing stage
7. No, but anticipate needing to revise a textbook to update it.
8. Published correction in journal
9. Systematic review system is designed for regular updates.
10. Upon publication final typesetting errors made tables difficult to read. Journal readily accepted a correction after a handful of email communications. Total effort, not that much
11. a page or two of update for a paperback publication, nothing more complicated
James
________________________________________
From: Repositories discussion list [[log in to unmask]] On Behalf Of Talat Chaudhri [[log in to unmask]]
Sent: 16 February 2010 16:47
To: [log in to unmask]
Subject: Re: Interest in Memento?
I'd expect people to cite the "jump-off" metadata page - although one
presumes that people can and do cite either the jump-off page or the
bitstream URI, perhaps inconsistently - and I would think it might well
be of interest to know how the item changed over time: so I'd tend to
agree with Andy. For instance, if an author cited an item by its
jump-off page and later a new edited version appeared and the metadata
was altered, that might change the meaning of a particular conclusion
and perhaps some fundamental factual metadata that was central to a
point being made. Who knows how marginal a use case this might be,
without quantitative evidence of how often items have been altered after
their original deposit? Who knows what the effects might be in resources
that cited that original resource? This does depend rather on the
repository software being used and how well it supports versioning at
present. It should always be clear what references were made to what,
and when, whether by assigning new URIs or by time stamping or both.
Speaking as a former repository manager who was very likely guilty of
altering items after deposit, I'm pretty sure this happens reasonably
often, although less sure of the real-world seriousness of the effects
upon citations and so forth. Best practice is rather dependent on the
capabilities of the software in question and the staff time available at
a particular institution, though. Let's imagine how much more serious it
might be if an image was replaced and/or metadata changed but the URI
pointed to the same place without a clear versioning history. Any
reference to it could become wholly inaccurate or misleading, since the
resource could be different in its entirety from a critical point of view.
I'd tend towards the view that repository software needs to have
complete versioning functionality shipped with it. One might compare
blog posts, for instance, which have time stamps in their URIs in an
immediately readable and unambiguous manner. Of course, the repository
also has to describe the relationships between those versions. For this
reason, I don't think that those relationships belong in the metadata
for the items themselves, but in a separate body of linked data metadata
(but perhaps that is a side issue for now).
While Memento type approaches (in my limited experience) are fantastic
for the web in general, I personally think that a more integrated
approach is needed in repositories.
Talat
Andy Powell wrote:
> I think I'd approach this from a slightly different perspective by asking two questions:
>
> 1) what resources do the URIs exposed by a typical repository identify?
>
> 2) does Memento serve a useful purpose in the context of any of those resources/URIs?
>
> (Aside: I suspect the first of these isn't particularly easy to answer in a Semantic Web sense, especially given FRBR-like notions of works and items and the fact that there might not be anything that represents 'typical' in this area... but let's ignore that for now?)
>
> The two most obvious kinds of URI are those for the (so-called) jump-off page (e.g. http://eprints.soton.ac.uk/48073/) and those for the individual documents that have been uploaded into the repository (e.g. https://secure.eprints.soton.ac.uk/soton/secure/00048073/01/Rampmetering.pdf).
>
> Individual documents grow by addition (I assume) - so I tend to agree that it is not clear what useful purpose Memento might serve. Each time a new document is added, it is given a new URI. In extreme circumstances, the document at a particular document URI might be deleted but it is never modified or replaced.
>
> The jump-off page presumably does change over time, as new versions of documents are added? Is that right? If so, it strikes me that it might be useful (in a citation sense) to see how the jump-off page has changed over time? I presume that it is the jump-off page that we expect people to cite??
>
> Does that make any kind of sense or am I totally out of touch with how repositories work?
>
> Andy
>
> --
> Andy Powell
> Research Programme Director
> Eduserv
> t: 01225 474319
> m: 07989 476710
> twitter: @andypowe11
> blog: efoundations.typepad.com
>
> www.eduserv.org.uk
>
>
>
>
>> -----Original Message-----
>> From: Repositories discussion list [mailto:JISC-
>> [log in to unmask]] On Behalf Of Chris Rusbridge
>> Sent: 15 February 2010 16:41
>> To: [log in to unmask]
>> Subject: Re: Interest in Memento?
>>
>> I think being able to access the version of a resource as it was at a
>> prior time is an important aspect of the citation process (hence
>> WebCite, for example, or the retained earlier versions in Wikipedia).
>> So i think Memento potentially has a place in the scholarly process.
>>
>> What the role of Memento should be in relation to repositories, I'm
>> less sure. In particular, if the repository grows by addition rather
>> than by change then I don't see a huge advantage. I think I mean here
>> that if the repository's documents are regarded as fixed, then you
>> don't need to show changes to them. (This might not be true in certain
>> preservation cases, of course, in relation to technology changes and
>> obsolescence, but it's not clear how Memento would survive such a world
>> either).
>>
>> I can imagine other kinds of repositories (for example, data
>> repositories) where change might be more likely to occur. These use
>> cases might need a little further analysis before one would be
>> confident that a service like memento would be helpful, but it feels
>> like they/it should be!
>>
>> --
>> Chris Rusbridge
>> Director, Digital Curation Centre
>> Email: [log in to unmask] Phone 0131 6513823
>> University of Edinburgh
>> Appleton Tower, Crichton St, Edinburgh EH8 9LE
>>
>> The University of Edinburgh is a charitable body, registered in
>> Scotland, with registration number SC005336.
>>
>>
>>
>> On 15 Feb 2010, at 11:53, Leslie Carr wrote:
>>
>>
>>> Have you heard of Memento, one of Herbert van de Sompel's recent
>>>
>> projects.
>>
>>> "Have you ever felt frustrated by your inability to get to old
>>>
>> versions of Web pages? Did you bookmark a page last year, and revisited
>> it recently only to find that the current content isn't even remotely
>> related to what caught your interest back then? ... Wouldn't it be much
>> easier if you could just connect to cnn.com, Wikipedia, or
>> news.bbc.co.uk indicating that you are interested in the pages of March
>> 20 2008, not the current ones? If you could activate a time machine in
>> your browser or bot?" (http://www.mementoweb.org/)
>>
>>> We (EPrints) have been asked to think about providing support for
>>>
>> this facility in our repository software, but we'd like to get some
>> feedback from the community.
>>
>>> Would this be useful to anyone? What use cases can you foresee?
>>>
>> Indeed, have you ever felt frustrated by the lack of this facility, as
>> the project page assumes?
>>
>>> --
>>> Les Carr
>>>
>>>
>
>
--
Dr Talat Chaudhri
------------------------------------------------------------
Research Officer
UKOLN, University of Bath, Bath BA2 7AY, Great Britain
Telephone: +44 (0)1225 385105 Fax: +44 (0)1225 386838
E-mail: [log in to unmask] Skype: talat.chaudhri
Web: http://www.ukoln.ac.uk/ukoln/staff/t.chaudhri/
------------------------------------------------------------
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
|