Can You Dig It A Systems Perspective

1
Can You Dig It?A Systems Perspective

• Dave Pattern
• Library Systems ManagerUniversity of
  Huddersfield
• d.c.pattern_at_hud.ac.uk

2

• There comes a time in every rightly constructed
  boys life that he has a raging desire to go
  somewhere and dig for hidden treasure.
• Mark Twain, The Adventures of Tom Sawyer

3
Supermarkets

• Supermarkets gain valuable insights into user
  behaviour by data mining purchases and uncovering
  usage trends. Further insights are gained by
  analysing purchasing histories, facilitated by
  the use of store loyalty cards.

4
Libraries?

• Libraries gain valuable insights into user
  behaviour by data mining borrowing and uncovering
  usage trends. Further insights are gained by
  analysing borrowing histories, facilitated by the
  use of library cards.

5
Libraries?

• Libraries could gain valuable insights into user
  behaviour by data mining borrowing and uncovering
  usage trends. Further insights could be gained
  by analysing borrowing histories, facilitated by
  the use of library cards.so, why arent we
  all busy doing it?

6
Borrowing at Huddersfield

• 13 years of circulation data
• details of over 3 million transactions
• a goldmine of information
• unfortunately, nobody was doing anything with
  the data!!!

7
Borrowing Suggestions
8
Borrowing Suggestions

• suggestions are generated by data mining the
  historical transaction data
• ranking algorithm
• number of people who borrowed both books A and B,
  divided by the total number of loans for book
  Bthis helps avoid the Harry Potter Effect

9
Personalised Suggestions
10
Personalised Suggestions

• generated by aggregating the recommendations for
  items most recently borrowed with a sprinkling
  of randomness!

11
Lending Paths (beta)
12
Lending Paths (beta)

• if were recommending book B, is it normally
  borrowed before or after book A?
• can we use that information to make predictions
  about future borrowing and give more timely
  recommendations?

13
Usage Statistics for Staff
14
Usage Statistics for Staff
15
Usage Statistics for Staff
16
Usage Statistics for Staff
17
OPAC Keyword Searches

• all keyword searches are logged
• the data can be mined to provide
• prompts/suggestions for searches
• trends in keyword usage
• common keywords that fail to produce results
  (e.g. renew)

18
Keyword Suggestions
19
Keyword Cloud
20
Common Zero Result Keywords

• newspapermen
• estimation
• superficial
• ligament
• also asbo, chav, disneyfication, newszak
  and sudoku!
• is this useful information for collection
  development?!

21
Can You Dig It?

• weve found gold in our own data and our services
  are richer for it

Prospectorhttp//www.flickr.com/photos/tooliver
/497511638/
22
Can You Dig It?

• theres gold in your data

Old Gold Mine http//www.flickr.com/photos/spla
tt/1058989589/
23
Can You Dig It?

• what if we collaborate? Ill show you mine (my
  mine?), if youll show me yours!

Norwegian mining plant http//www.flickr.com/phot
os/bzzzt/2458300944/
24
Huddersfields Usage Data

• aggregated usage recommendation data
• Open Data Commons licence
• http//library.hud.ac.uk/usagedata/

25
Huddersfields Usage Data

• usage data for over 80,000 titles and 2,000,000
  transactions, broken down by
• year
• academic school
• academic course (with relevant UCAS codes)
• recommendation data for over 37,000 titles

26
Lets Work Together

• Ive shown you mine now whats stopping you from
  showing me yours?

peek-a-boohttp//www.flickr.com/photos/96221617
_at_N00/51603550/