There is an important wider issue of 'big data/analytics' that Mathew
alluded to in his original email..and that seems to me to be one of the key
underlying points. 'Big Data' and analytics is one of the current key
technology trends along with the 'cloud' and as that trend manifests itself
in library technologies I'm sure we'll see recommender services (and much
else) become more common.
Libraries and library system vendors have been very slow to make use of data
to deliver new services such as recommender services. In some ways it seems
astonishing that libraries lag so far behind in this regard. As vendors move
their systems to the cloud and aggregate, on a global scale, various kinds
of data (that was hidden away in local systems) we'll see more recommender
type services emerging. ExLibris bX is a pioneering example of that. It
might be worth noting too that OCLC Research and the Information School,
University of Sheffield have a project on Library centric recommender
systems
http://www.oclc.org/research/activities/recommender.html
If you don't want to wait there are some things you can do right now:-
(1) Follow Dave Pattern's example from the University of Huddersfield and
use your local circ data using to develop your own local recommender system.
I'm sure Dave would advise
(http://library.hud.ac.uk/blogs/projects/lidp/meet-the-project-team/
(2) Use the BibTip service to deliver a recommender service
http://www.bibtip.com/en BibTip doesn't use *circ* data but rather the
search behaviour of catalogue users. The University of East London reported
trialling it a while back and they might be able to share their experience.
(3) implement a 'Reading List' system--this is another way to direct users
to 'recommended' resources. A very different approach to using local circ
data of course but is can be implemented now. It's a pretty good bet these
reading list systems themselves will implement some kind of recommender type
service based on aggregated user data in due course. There are several new
systems coming to market at the moment. For more info see HELibTech
http://helibtech.com/Reading_Resource+lists
Good luck!
Ken
Ken Chad Consulting Ltd
Tel +44 (0)7788 727 845. Email: [log in to unmask]
www.kenchadconsulting.com
Skype: kenchadconsulting Twitter: @KenChad LinkedIn:
http://www.linkedin.com/in/kenchad
Higher Education Library Technology wiki: http://helibtech.com/
My presentations on slideshare: http://www.slideshare.net/kenchad
-----Original Message-----
From: A general Library and Information Science list for news and
discussion. [mailto:[log in to unmask]] On Behalf Of Lizz Jennings
Sent: 19 December 2012 10:11
To: [log in to unmask]
Subject: Re: Using circulation history data to inform library catalogue
search results
Robert,
bX is definitely useful in terms of subject coverage, but the advantage of
local data would be that you could theoretically promote heavily used items
to the top of search results pages, which would be a massive help to
students looking for recommended reading on their courses etc.
especially where the title is very generic.
Kind regards,
Lizz
On 19/12/2012 09:51, Robert Bley wrote:
> Dear All
>
> Apologies for responding to the list with a blatant "plug", but I think
it's relevant (albeit tangentially) to this topic.
>
> We at Ex Libris have been mining usage data for quite a while now, to
provide a scholarly article recommender ("people who read this article also
read...") service called "bX".
>
> Please let me know if you would like more information about it.
>
> Kind regards,
>
> Robert
>
>
>
> -----Original Message-----
> From: A general Library and Information Science list for news and
> discussion. [mailto:[log in to unmask]] On Behalf Of Claire
> Eskriett
> Sent: 18 December 2012 16:01
> To: [log in to unmask]
> Subject: Re: Using circulation history data to inform library
> catalogue search results
>
> Hi
>
> Capita are developing a recommendations service for Prism based on circ
data which I understand is pretty imminent, a kind of 'people who borrowed
this borrowed that' idea. I'm sure someone with more details will be along
soon!
>
> Claire
>
>
> Claire Eskriett, Systems Librarian
> (Days of work are Mon, Tues, Thurs& Fri) Information Services
> Cockcroft Building University of Brighton Lewes Road Brighton BN2 4GJ
> T: 01273 642766
> E: [log in to unmask]
>
>
>
> -----Original Message-----
> From: A general Library and Information Science list for news and
> discussion. [mailto:[log in to unmask]] On Behalf Of Paul
> Stainthorp
> Sent: 18 December 2012 15:42
> To: [log in to unmask]
> Subject: Re: Using circulation history data to inform library
> catalogue search results
>
> Hello Mathew,
>
> Absolutely.
>
> I'd point at a couple of JISC-funded projects in HE library data that
explored doing just that (disclaimer: my own institution contributed
circulation data to both projects):
>
> [1] MOSAIC ("Making Our Shared Activity Intelligence Count"):
> http://www.sero.co.uk/jisc-mosaic.html
>
> [2] SALT2 / Copac Activity Data:
> http://copac.ac.uk/innovations/activity-data/
>
> Other people will probably have examples of other projects.
>
> Unfortunately this re-use of data (locally, or shared across
> institutions) still doesn't seem to have made it into the mainstream of
commercial LMS provision in Higher Education (if anywhere), nor does it seem
likely to in the near future.
>
> Regards,
>
> Paul
>
> ------
>
> Paul Stainthorp MSc
> Electronic Resources Librarian
>
> University of Lincoln
> The Great Central Warehouse University Library Brayford Pool Lincoln
> LN6 7TS
> United Kingdom
>
> Tel: +44 (0)1522 88 6193
> Email: [log in to unmask]
> Web: www.library.lincoln.ac.uk
>
>
>
>
> -----Original Message-----
> From: A general Library and Information Science list for news and
> discussion. [mailto:[log in to unmask]] On Behalf Of Mathew Stone
> Sent: 18 December 2012 15:15
> To: [log in to unmask]
> Subject: Using circulation history data to inform library catalogue
> search results
>
> Good afternoon everyone
>
> When library users search for items via the library catalogue, would it
not make for a superior experience if the retrieved results were informed by
the vast amounts of circulation history data that library management systems
collect? So for instance, results could be ranked according popularity
rather than publication date. Circulation history data could also identify
relationships between items that may not be apparent from catalogue records
alone, so this too could inform the results retrieved.
>
> I work in a small hospital library and I can't claim to be much of an
expert when it comes to library management systems, so I was wondering if
anybody on here knew if Talis, Heritage, etc. are developing their products
with such issues in mind?
>
> This issue was raised in a book I recently read (The Atlas of new
librarianship by R. David Lankes - kindly lent to me by the Library&
Knowledge Service @ Rotherham NHS Foundation Trust) which proceeds to make
the wider argument that since the success of companies like Google, Facebook
and Amazon has been founded on their ability to make use of the data
supplied by their own customers, that libraries need to do the same.
>
> Kind regards
> Mathew Stone
> Assistant Librarian | Health Library& Information Service |
> Fieldhouse Education Centre, BRI Bradford Teaching Hospitals NHS
> Foundation Trust
> T: 01274 36 4122
> F: 4704
> http://www.bradfordhospitals.nhs.uk
>
|