OCLC’s iCAS Service as a Tool for Collection Managers Richard Ovenden Director of Collections, Edinburgh University Library.
Post on 18-Jan-2016
215 Views
Preview:
Transcript
OCLC’s iCAS Service as a Tool OCLC’s iCAS Service as a Tool for Collection Managersfor Collection Managers
Richard OvendenDirector of Collections,Edinburgh University Library
Outline
OCLC’s iCAS Service (Interactive Collection Analysis System)
Pilot ProjectFindingsRecommendationsPersonal ViewSummary
OCLC’s iCAS Service
OCLC Lacey Product CenterACAS: Automated Collection
Assessment and Analysis ServicesBased on WLN ConspectusIndividual Collection analysesCombined analyses with overlap and
uniqueness measuresEmphasis on publication dates
Pilot Project
Funded by CURL/RSLPPartners: Hull, Imperial College
London, Edinburgh, SOAS, Liverpool, Natural History Museum
Reported February 2002
Process (1)
2, 767, 669 book format bibliographic records submitted
De-dupeMatched No Call Number items
against identical records within CURL
Process (2)
Match against WorldCatThis found an additional 570, 075 records
with call numbersRejected 12, 802 ‘ill-formed’ call numbersTotal number of records analysed 2, 268,
225 (84%)Only 16% of records could not be
processed because they could not be allocated a call-number.
Number/Percent of Records Analysed
Edinburgh 604, 531 78% Hull 451, 590 99.9% Imperial 202, 758 74% Liverpool 536, 816 85% NHM 45, 820 47% SOAS 436, 710 92%
Group Analysis
Edinburgh 784, 933 772, 923 606, 124 269, 468 166, 799
Hull 473, 069 460, 653 460, 647 11 6
Imperial 276, 952 274, 398 203, 137 101, 105 71, 261
Liverpool 661, 084 634, 714 537, 558 131, 757 97, 156
NHM 99, 971 97, 590 45, 873 44, 345 51, 717
SOAS 471, 660 467, 418 427, 688 24, 019 39, 730
TOTAL 2 767 669 2 707 696 2 281 027 570, 705 426, 669
Findings (1)
Data Problems mapping inconsistencies (SOAS mapping to
Celtic/Chinese)Classification
Hull had used wrong country subdivisions in some parts of the LC schedules
Matching against other utilities vs. local classification needs
Methodologies: COPAC Loading variants Liverpool, Edinburgh
Findings (2)
Conspectus The sub-divisions perceived as weak in some
areas (eg biology) Need to update categories: Where is Artificial
Intelligence? Fit with collections not good. Vet Medicine in
Agriculture?Quality & relevance of data
Unique items?Accuracy of Conspectus match at highly
granular level
Findings (3)
Automated matching routines are rough & ready Special topics in Computer Science included
two books 1800-1899Date breakdowns - currency indicationsBased only on OPAC dataSubject groupings don’t fit collections –
compare holdings by site library?Brings out worst in LibrariansBrings out worst in Chief Librarians
Recommendations
Realise limitationsKnow your data (=know your collections)Talk to your colleagues (cataloguers,
systems, as well as collections specialists)Record data decisions (eg CURL uploads)Collections analysis not collection
description (OPAC analysis …)
Personal View
Labour savingObjectivity?Adjunct to other methods
People still matterResource sharing
Compare like with like? Compare against regional centre? Basis for funding collaborative collection
management?
Summary
ICAS has valueICAS has limitationsICAS costs moneyICAS has potential for UK librariesIt does exactly what it says on the tin
FIN
richard.ovenden@ed.ac.uk
top related