Sheffield

Big Brotherin the library

Dave Pattern, Library Systems ManagerUniversity of Huddersfield

[email protected]

Preamble

• Presentation available at:– www.slideshare.net/daveyp/

• Please remix and reuse this presentation– creativecommons.org/licenses/by-nc-

sa/3.0

Contents

• Supermarkets• Virtual footprints• Data collected at Huddersfield• Using data analysis to improve

services• Sharing data

A long, longtime ago…

www.flickr.com/photos/hippie/2476662922/

tinyurl.com/c8qm8h

www.guardian.co.uk/money/2005/sep/27/ethicalmoney.lifeandhealth

www.timesonline.co.uk/tol/news/uk/article1572626.ece

www.flickr.com/photos/mikewarren/48237822/

www.flickr.com/photos/jay-chilli/2178457837/

in a digital world, we leave virtual

footprints wherever we go

University ofHuddersfield

Facts and figures

Student population• about 20,000 (full-time and part-

time)

University Library• just under 400,000 physical items• over 700,000 loans per year

User data collected by the library

• Turnstile System– when you entered the library

• Books– which items you borrowed and when

• Computers– which PCs/Macs you used and for how

long– which web sites you looked at

Anonymous data collected

• Library Catalogue web site– keywords used for searches and the

number of results found– which books were looked at– which features and tools were used

What the data is used for

• Library Catalogue web site– improving the search facility– analysing search trends

• Turnstile System– find out how well the library is being

used (e.g. weekends and holidays)– has the refurbishment of the library

made any difference to the number of people using the library?

What the data is used for

• Books– which books do we need extra copies

of?– which books can we get rid of?– which books are being borrowed by

specific groups of students?– create borrowing suggestions for

students in the Library Catalogue web site

library non-usageanalysis

Library Catalogueweb site

Keyword cloud

Keyword suggestions

Borrowing suggestions

Personalised suggestions

how is it done?

1) find out who’s borrowed the book

Alice

Brian

Cathy

Alice

Brian

Cathy

2) find all the other books they borrowed…

Alice

Brian

Cathy

3) find the common titles…

people who borrowed

also borrowed

so, was it worth it?

Range of stock being borrowed

Books per active borrower

sharing usage data

Usage data

• Nov 2008: Released of aggregated and anonymised book usage data and book recommendation data (link)– 2 million book loans and 80,000 book titles

• Apr 2009: Aggregated keyword search data and linked keywords (link)– 3 million searches on the library catalogue

• Open Data Commons Licence– no restrictions on how the data can be

used

what next?

What if…

• …every library shared its usage data?

• …could we use that data to improve the “student experience”?

• can we encourage people to play with that data and create cool stuff?

Thank you!

www.slideshare.net/daveyp/www.daveyp.com/blog/