HILT High Level Thesaurus Project

1
HILT High Level Thesaurus Project

• Report to the JISC/NSF Conference on HILT Phase I
  (completed) and HILT Phase II (just starting)
• Dennis Nicholson
• Centre for Digital Library Research,
  Strathclyde University

2
Overview

• (1) HILT I From Aims to Conclusions to the idea
  of a pilot TeRM for the JISC IE
• (2) What a TeRM or Terminologies Route Map is
• (3) HILT II Aims, participants, approach,
  deliverables, key issues, a summary, how to
  participate

3
HILT 1 Aim

• Look into the problem of cross-searching and
  browsing by subject across UK services in
  Libraries, Museums, Archives, HE, FE, Public
  Libraries, electronic services, clumps projects,
  the DNER, the RDN etc.
• Were they using different subject schemes, or the
  same schemes differently, and did it matter?

4
HILT 1 Aim

• If it did, could a consensus be reached on a
  solution that was affordable, sustainable,
  implementable, politically acceptable, useful,
  future proofed, and the rest

5
HILT 1 Conclusions

• That many different subject schemes and practices
  were in use
• That cross searching by subject was considered
  of value to users and staff.
• That an online terminologies route map or TeRM
  that would map subject schemes to user
  terminologies and to each other was the preferred
  solution

6
HILT 1 Conclusions

• But that things like level and nature of user
  need, design requirements, and costs against
  benefits needed to be examined before we
  committed to a possibly expensive solution
• Finally, there was a strong consensus favouring a
  project to create a pilot TeRM and investigate
  these issues

7
HILT 1 A TeRM for the IE?

• This recommendation was in line with JISCs plans
  for a pilot terminologies server for the JISC IE
• JISC agreed to fund a project to implement and
  evaluate a pilot TeRM, similar to that proposed
  by HILT I

8
What is a pilot TeRM?

• Online interactive service that
• Maps relationships between terms in standard
  schemes and, possibly, local variations (LCSH,
  DDC, AAT, UNESCO, user terms for example)
• Gives service staff some input into the process
• Disambiguates user input terms

9
What is a pilot TeRM?

• Offers suitable terms for cross-searching groups
  of services that use different schemes
• Also offers broader and narrower options - a
  route map
• Also identifies services or collections to search
  based on users subject query (perhaps dealing
  with granularity level differences by searching
  collections database using truncated DDC number)

10
What is a pilot TeRM?

• Other points
• Collections database may also record subject
  scheme used to allow site specific search
  formulations
• System also does M2M

11
HILT Phase II Aims

• Use pilot TeRM to establish terminology service
  requirements for the JISC IE
• Focusing on user collection level needs in the
  DNER, the RDN, and UK HE and FE generally
• Make recommendations on a possible future
  service, taking into account such things as
• Level and nature of user need,
• mapping problems,
• commercial software V. original development

12
Participants

• Similar to HILT 1 with a few additions
• CDLR mda NCA OCLC UKOLN RDN DNER NGfL
  Scotland SLIC SUfI
• FE Representative
• Terminology experts, Alan Gilchrist and Leonard
  Will (external evaluator)

13
HILT II Proposed approach

• Build an Illustrative TeRM with Wordmap
• LCSH, DDC, UNESCO, AAT, Others
• Selective collection level requirements
• Interact with users, staff, experts in a
  simulated IE draw out initial specification,
  keeping M2M in mind
• Improve on initial TeRM
• Agree evaluation criteria evaluate TeRM
  (includes user evaluation)
• Make recommendations

14
HILT II Deliverables

• Better understanding of
• The subject retrieval needs of FE and HE users
• Associated terminology mapping requirements
• A working, limited functionality, demonstrator
  service for the JISC IE
• Requirements, set up and maintenance costs, and
  costs against benefits, for a future service,
  including both user and M2M terminological and
  functional requirements.
• Report and recommendations.

15
Issues (1) A central spine?

• Should one scheme form a central spine against
  which all other schemes and the user terminology
  set are mapped?
• If not, then what? Map user terms to every
  scheme?
• If so, which? DDC? (well-structured,
  hierachical multi-lingual).

16
Issues (2) A single scheme?

• Would the alternative option of a single scheme
  be
• Practical?
• Less expensive to maintain?
• If so, which scheme?
• HILT 1 workshop participants unanimously rejected
  this option

17
Summary

• Research project - outcome to some extent
  uncertain
• But IE calls for terminologies server and HILT II
  should help define requirement
• Consensus from HILT 1 is that mapping is the key
  but we keep an open mind on whether a single
  scheme is better
• Will have to prove it if consensus is to be
  maintained

18
Summary

• Particular points well be looking at include
• Mapping problems and solutions
• Value to users and costs against benefits
• Requirements of ensuring co-ordination of staff
  subject description efforts with one solution or
  the other
• And of maintaining consensus and keeping in step
  with the international community

19
Summary

• Finally, keen to encourage participation in the
  discussion over the next year -
  LIS-HILT-ORG_at_jiscmail.ac.uk will be an open forum
  for debate on project issues

20
Further Information

• Website http//hilt.cdlr.strath.ac.uk/
• e-mail
• d.m.nicholson_at_strath.ac.uk
• emma.mcculloch_at_strath.ac.uk