I'm hoping the conversations and interest around #metadataquality hashtag: https://twitter.com/hashtag/metadataquality help to move forward some of these conversations from well constructed research projects and academic papers to something that more of us can implement locally in our systems.
There are many different ways that we could look at some of these problems and I think having more of us sharing our ideas and possibly code will be great.
From: Code for Libraries <[log in to unmask]> on behalf of Sergio Letuche <[log in to unmask]>
Sent: Wednesday, May 6, 2015 7:20 AM
To: [log in to unmask]
Subject: Re: [CODE4LIB] How to measure quality of a record
i felt i was missing something, since i could not find some general, "most
used approach", and perhaps some code on github that implements these
2015-05-06 15:08 GMT+03:00 James Morley <[log in to unmask]>:
> I think a key thing is to determine to what extent any definition of
> 'completeness' is actually a representation of 'quality'. As Peter says,
> making sure not just that metadata is present but then checking it conforms
> with rules is a big step towards this. I would also extend this to
> assessing at what level of accuracy things have been set, for example dates
> (a rough range vs a precise day) and geotags (coordinates presenting the
> centre of Paris vs the exact position that a photograph was taken from).
> These sorts of things can make a big difference to both the discoverability
> and practical reusability of records by end users.
> Best, James
> From: Code for Libraries [[log in to unmask]] on behalf of Esmé
> Cowles [[log in to unmask]]
> Sent: 06 May 2015 13:51
> To: [log in to unmask]
> Subject: Re: [CODE4LIB] How to measure quality of a record
> Mark Phillips has a related blog post that I think is an excellent place
> to start, which outlines a system for scoring how complete a record is:
> There was some discussion on twitter recently about this, which you can
> look up on the #metadataquality hashtag:
> I think there was a move to setup a mailing list for this topic or
> something like that, but I'm not sure where that stands now.
> > On 05/06/15, at 7:21 AM, Sergio Letuche <[log in to unmask]> wrote:
> > Hello community,
> > is there a way, any statistical approach, that you are aware of that
> > say, allows one to have an idea of how "complete" a record is, or what
> > the actions you take in order to have an idea of the quality of a record,
> > and eventually a database?
> > Thank you in advance