This house believes that A

1
This house believes that AI is RIP

• JIBS Workshop
• 13 November 2009

2
Disclaimer!

• This presentation does not reflect the views of
  the University of Huddersfield
• Nor does it necessarily mean ALL AI databases!

3
Context

• University of Huddersfield
• 50,000 e-journals
• 50,000 e-books
• 1,000 full text repository items
• 200,000 library records
• 80 AIs
• First European customer for Serial Solutions
  Summon

4
Are we fighting a losing battle?

• By relying on AIs
• Do we support our users in the way they expect
  and demand?
• Do we overload them with different systems or
  lack of intuitive interfaces?
• Is Google actually a problem?
• Can we afford to cut journals and keep AIs?

5
Quality vs. Cost (1)

• Article abstracts many AI databases are still
  only indexes, which is often frustrating to the
  user
• Cover to cover indexing An AI with a low
  proportion of core content or a high proportion
  of tertiary content is at risk
• Duplication of content Two resources with
  similar content are not economically justifiable
• Full-text linking an AI that does not link does
  not promote resource discovery

6
Quality vs. Cost (2)

• Date coverage unless specifically covering an
  archive period, AI databases that purport to be
  current, but index a high number of ceased titles
  are not relevant
• Geographical Coverage if specific geographical
  areas are not adequately covered then the AI is
  not valuable for research

7
Quality vs. Cost (3)

• Publisher coverage libraries want to use AI
  databases to achieve a good spread of publishers
  if this is not the case, then the resource is
  little better than searching a publishers
  platform
• Intuitive interface is the resource as easy to
  use as Google?
• Shibboleth authentcation, EZProxy access as
  standard any AI database that relies on
  individual usernames and passwords for access is
  creating a barrier to use

8
Quality vs. Cost (4)

• Unrestricted access analysis of turnaways and
  usage data at the University of Huddersfield
  shows that resources that restrict access by
  number of simultaneous users often leads to
  dramatic drop in usage over a period of time as
  users become frustrated by turnaway messages
• COUNTER compliant usage data a lack of COUNTER
  compliance means that accurate comparisons cannot
  be made

9
Primary material

• In a digital environment this information is
  becoming more retrievable AIs only cover a
  small proportion of what is out there
• JISC Digitisation Programme
• Sound, images, journals, moving pictures,
  newspapers etc.
• UK Research Data Service Feasibility Study final
  report
• Research data has remained a substantially
  untapped resource and that it is often
  unstructured and inaccessible to others.

10
Federated (Meta) Searching (1)

• Many librarians do not recommend federated search
• federated systems remain controversial because
  they focus on what we think users want, at the
  expense of functionality, precision, and finesse.
  They are still a long way from providing a
  single, simple solution to information
  retrieval.
• Tenopir, C, Online databases Can Johnny search?
  Library Journal, 2007, 132(2), 30.
• http//www.libraryjournal.com/article/CA6407784.ht
  ml?industryid47130

11
Federated (Meta) Searching (2)

• Research at Stockholm University shows that
  students were not enthusiastic about Google
  Scholar or MetaLib however, they agreed that
  Google Scholar was easy to use
• Nygren, E, Haya, G and Widmark, W, Students
  experience of Metalib and Google Scholar, 2006,
  Stockholm, universitetsbiblioteket
• Is federated searching a transient technology
  like the CD-ROM and is the real Holy Grail just
  over the horizon?

12
Pre harvested searchthe true one-stop shop?

• The key for the medium term is to provide
  Google-like interfaces with Google-like results
• There is a race to provide this through systems
  that use preharvested data rather than federated
  searching
• Primo (Ex Libris)
• Summon (Serials Solutions)
• WorldCat Local (OCLC)

13
This house believes that AI is RIP

• Why is Google so easy and the library so hard?
• Duddy, C, A student perspective on accessing
  academic information in the Google era, 32nd UKSG
  Annual Conference and Exhibition, 30 March - 1
  April 2009, Riviera International Conference
  Centre, Torquay
• Why do we want to teach our users to be
  librarians?
• Pattern, D, OPAC 2.0 and beyond. 32nd UKSG Annual
  Conference and Exhibition, 30 March - 1 April
  2009, Riviera International Conference Centre,
  Torquay. http//eprints.hud.ac.uk/4143/

14
and finally Visualization