THURSDAY-June 24
Back to the Future (FRBR) --
The Cataloging and
Classification Section's Committee on
Cataloging: Description
& Access and MARBI present Back to the
Future:
Understanding the
Functional Requirements of Bibliographic Records Model
(FRBR)
and its Impact on Users, OPACS, and Knowledge Organization.
This one and a half day
pre-conference was designed to demonstrate how FRBR will
influence the future development of information standards
within the
library community, and acquaint technical services and IT
professionals
with the implications of the FRBR model for cataloging
rules, MARC formats
and other emerging standards for electronic
technologies.
Topics and Speakers :
The FRBR Model--Allyson
Carlyle,
FRBR Basics--Barbara Tillett, Library of Congress
Why FRBR?--Glenn Patton,
OCLC
The FRBR Future--Tom Delsey, Independent Consultant
FRBR and the Fiction Finder
Project--Diane Visine-Goetz, OCLC
FRBR Applications in an
LMS/OPAC--Vinod Chachra,
VTLS
FRBR and MARC--Sally
McCallum, Library of Congress
How FRBR Will Affect
Library Users and Staff--Jennifer Bowen, University
of
FRBR and Library
Services--Olivia Madison,
Barbara
Tillett
Allyson
Carlyle
See
http://www.ischool.washington.edu/acarlyle/FRBR2004Carlyle_files/frame.htm
(until Sept. 15) [saved at home]
Glenn
Patton
Tom
Delsey
Function
a)
Cutter: find (search author, title, subject), collocate (what the library has),
assist in choosing (select)
b)
Paris Principles: includes nothing about description, only access points
c)
FRBR: find, identify, select, obtain. Identify should apply to all - work,
expression, manifestation whereas obtain applies only to manifestation and item
Data set/core record
a)
ISBD mandatory elements
b)
AACR 1,2,3, levels of description but with little
instruction
c)
find, identify, select, obtain -> the work, expression, manifestation, so:
1) to find a manifestation, you need a title, parallel title, etc.; 2) to find
all manifestations it needs to reflect relationships to other works, concepts
(subject); 3) to identify, we need form, language, more descriptors on bibs
like edition, publisher, etc.; 4) to select, we need form, relationships like
formerly and continued by, physical medium; 5) obtain - source for
acquisitions, license rights
Class of materials
a)
class: book, map, music, sound record, electronic
resource, graphic, etc.
b)
type of publication: finite or continuing
c)
problem: type of material in 008 isn't the same as
AACR
d)
then 007 - give a category of material that also
doesn't match AACR, e.g. tactile material, notated music
e) FRBR - we need to realign
class of materials and look at content and carrier
|
CONTENT |
CARRIER |
|
WORK
-- EXPRESSION |
MANIFESTATION |
|
Category
of work: literary, musical graphic |
Type
of medium: print, graphic, 3-dimensional |
|
Mode
of expression: notation, sound, fixed image |
Status:
published or unpublished |
|
|
Access:
remote or direct |
|
|
Issuance:
simultaneous or successive |
|
|
Termination:
predetermined or undetermined |
Record type
Now
|
AUTHORITY |
BIBLIOGRAPHIC
RECORD |
HOLDINGS |
|
Control
access points in bibs and give cross references |
Bib
contains: author, title, description, subjects, added entries |
Local
call number, copy, etc. |
|
Right
now we view these as adjunct to the bib |
|
|
FRBR
|
AUTHORITY |
BIBLIOGRAPHIC RECORD |
HOLDINGS |
|
English (language) expression of the work Illiad by person Home |
Manifestation - only the attributes for THIS
manifestation are needed |
Item with call number and copy info |
|
Person is Homer |
Title on this particular publication |
|
|
Work is Illiad |
Publisher name |
|
|
Expression is Illiad |
Illustrator |
|
|
Person is translator |
|
|
|
Concept/event is Trojan War |
|
|
Syndetic structure
We should be making full use of both bib and
authority records
Bibs: main and added entries, notes, editors,
sponsoring bodies, variant titles, series, analytics, relations (alternate
form, continued by, related editions, etc); unless you use relator
codes whenever you can, the user has no idea why you put all that data and
names on the record - we don't usually tell the user why we put what we did on
the record
Authorities: see refs - variant names, variant
titles; see also - related persons and corporate bodies; related subjects
You should be able to move up and down FRBR
relationships
Work <- -> Expression <- ->
Manifestation
Person
authority
Person or corporate name Work
Corporate
authority Expression
Manifestation
Sally
McCallum
Vinod Chakra and John Esply - VTLS
Jennifer
Bowen
RedLightGreen
Diana
Vizine-Goetz on FictionFinder
FRIDAY--June 25
Back to the Future (FRBR) -- LEAVE
EARLY !!!
SAC Subcommittee on Semantic
Interoperability meeting, be early
Members present: Shelby E. Harken,
Chair; Lois M. Chan, Anton J. (Tony) Olson, Ruth Bogan,
Giles S. Martin, Shannon Hoffman, Bonnie Dede
Guests: Marie Whited, Yale;
John Maier, Institute of Fine Arts, NYU;
Bryan Baldus, Quality Books; Sherman Clarke,
NYU; Pat Kuhr, H.W. Wilson
The schedule/timeline was
discussed.
Tony reported on the work
they intend to do beginning in August with the list of projects. Giles and Ruth
will be working with him and others are welcome. He suggested AAT and Mesh may
not belong on the list. Some of the information is skimpy. Some projects are
part of something else.
We want to ask ALCTS if
publishing would be possible in LRTS - they get the right of first refusal.
David Miller is working on getting the Reference Structures report published
there.
The next meeting will
presumably be at the same time (Friday,
Meeting adjourned.
Note:
Draft documents on the web, see: http://www.und.nodak.edu/dept/library/Departments/abc/subjhead.htm#SACSEM
MARBI and CC:DA reports were
given by John Attig. A relevant MARBI proposal is to
enhance the 752 to expand country to country or larger region and add subunit
of a city. CC:DA will be charged with work on AACR3.
The revision will be in phases. Part 1 will be first, between Nov. and
Mar. CC:DA
needs to hear from various communities. Part 2 will be access points but this
part still needs a good conceptual design. The plan is to publish in 2007. John
thinks we are unlikely to see much rule revision in the mean time. The new rules for Rare Books is planned for 2007 but will refer to
AACR2
Non-human access points - robots, puppets, animals, etc.
in TV, movies, etc. What rules do you use? Where do you do authority work? The
proposal has several options: personal names in name file, animals in name and
imaginary in subject, or all in name file.
It was decided the "expert panel" is a moot
topic. The FAQ/Best practices project will take a lot of maintenance and how
does it fit with Q & A? There was a brief discussion on physical
description for remote electronic resources and source of title note.
E-resource management DG - Digital Library
Federation Electronic Resource Management Initiative Project http://www.diglib.org/forums/Spring2004/robertson0404.htm
; etc.
SATURDAY--June 26
OCLC - workflow - bib-note, PromptCat, etc
Cataloging
services - see handout.
Bibliographic notification (bib note). You set up a profile for one of the 3 options. If
there is a change to the record, we get the full record and can also receive a
report. If we choose the option to put in save file, there is no charge for
export in Connexion. We will be able to flag save
file bibs and do one command. Can apply Constant Data,
update, export. Can use with PromptCat. You
could accept a minimum level bib with PromptCat and
receive bib note of update. Cataloging Partners adds more customized cataloging
and includes original cataloging with TechPro. To do PromptCat, we can say which account at the vendor we want
processed. We complete one form for each vendor. Certain fields can be added to
bibs. If there is no bib, you can get a data record. You get a file of bibs and
a report of what is in your file. Labels can be included with processing. It
can be set up with standing orders. You can set it up to receive an email when
a file is ready - it is usually based on shipments. PromptCat
is employed after you order - it is not used to create an order record.
Batchload - this can be used to upload a file of local bibs to OCLC to set
holdings. If you did original cataloging but another record is input in the
meantime, your holdings are set on that record. Can do this for LDRs - Bobby Botham at
Connexion - a personal name that is not qualified will never be a match. Can
lock and replace to save controlled headings as long as all the headings were a
full match. Otherwise all rules about lock and replace apply. Steps: call up a
bib, control all - if perfect replace, if need to control single and it is a
pure match - replace, do local edits, export/update. OCLC knows there are
punctuation problems with partially controlled headings. LCCN displays without
a hyphen; they will change searching for with or without hyphen.
Review
files - use action menu to save automatically as a special status. You chose
who you want to send the record to review to. Email that library. They have
read only access and you edit while they read or they can just email you back.
The reviewer's save file will have regular save menu at the top and below will
be records for review.
Upcoming
enhancements - will allow string search in keyword box. Will allow you click to
export then get a file of all you did in the day in one file. Three months will
be given for each upgrade. Connexion 1.2 will be late
2004 with authority file searching, NACO, local files, text strings (assign to
a keystroke and it can be added wherever the cursor is). CatME
and Passport will probably end at the same time.
Connexion import - OCLC needs to find an OCLC number in the 001 to upload. If it
doesn't find it, it will create a workform to edit.
To upload a record that has OCLC number in 035, Aleph will need to change it to
001.
Connexion Express is great for deleting. It has full keyword searching. Connexion client will allow offline work in ver. 1.2. You will also be able to save records at the
point of acquisitions and set holdings late. Spellchecker won't be included til at least 1.3.
MARBI
2004-6.
Indicator to define new fields and suppress display labels copyright or legal
deposit number. Approved with changes.
2004-5.
Music incipits. Approved with changes to make it less centered on
"RISM".
2004-7. Revision of 752 field to add indicators or define a new field of similar
structure as a subject. Will come back at Midwinter.
2004-DP04. Discussion favored option 1 which would add a $y
for 010 020 022, clarify definitions for #a, $y $z. Other control number fields
should be considered too.
2004-8 approved.
Adam Schiff will become MARBI chair.
ALCTS Electronic Resources IG -- Tracking electronic resources at the Library of Congress
SACSEM PROGRAM be early
Report to be published in Technical Services Quarterly
<
CC:DA Liaison EMAILED MARY ABOUT
CONFLICT
http://www.libraries.psu.edu/iasweb/personal/jca/ccda/
(see Monday meeting time)
Scholarship
Bash
SUNDAY--June 27
OCLC breakfast -- Cataloging
Product pricing will probably go up ca. 3%. 50,000 libraries
in 84 countries use OCLC.
Some current research includes: OAI, FRBR algorithm,
bookkmarklets, ISBN to match other eds of the work, FictionFinder.
Amazoogle effect - everyone seems happy with it so what needs
to be done it get them to qualitative information.
New preservation center in
Will be integrating FirstSearch and
WebJunction being used by 8000 libraries
QuestionPoint is being used by 3000 libraries
Goal for complete move to oracle - July/Oct 2005
WorldCat Pilot with Google now also includes MSN and Yahoo
Connexion: local editing - 1.2; spellcheck, NACO - 1.3
MeSH - not til 2005
Full unicode - 2005
PromptCat - is ISBN mismatch, vendor doesn't do label
ISBN 13-digit for now put in 024 indicator 3 for EAN
Can request specific trainer for SCCTP SAC Subject - maybe
NDLA pre-conference - Linda Gabel would be good
SAC - General Meeting
MARBI report was given (ANN/04). LC report
(see handout) items included: job status - all vacancies will be filled from
within and no new outside people; acquisitions and bibliographic services are
merged; web cat desktop includes correlations enhanced with Dewey-LCSH; Unicode
is delayed another year (Voyager not ready); LC will soon handle 13-digit ISBN;
looking at adding biographical information and more call numbers in 035 for
name autho records; religious law KB will be out later this year; LC is using
Gary Strawn's software to fix subject headings.
DDC report:
3 publications: Dewey Decimal Classification: Principles and Application (3rd
ed), DD22 and DD22 Abridged; there is a proposal on how to classify graphic novels
http://www.oclc.org/dewey/discussion/default.htm
; Editorial Policy Committee will put recent work on a web page - looks SACSEM
related
IFLA report:
IFLA working on a Virtual Clearinghouse for Subject Access Tools; many subject
thesauri are not available on the web, but they inted to develop a clearing
house for subject access tools worldwide and use descriptions/attributes
defined in a glosary -- "subject access tools" to include classification
systems, subject heading systems, and thesauri.
Music
Library Association: MLA has a new standards group to develop proposals for
standards; Music Decisions now available only online (for fee)
Law report:
a committee is looking at how to determine whether a subject heading is a legal
term
Discussion
period: FAST - beta should be ready during the summer; geographic is really
ready; creating records in MARC21 format; have 2 million records not including
name/title or conference headings; FAST is upward compatible - any LCSH can be
made into a FAST heading but can't necessarily put it back together; SAC will
form a committee for comment
Reference
structures: David Miller and Tony Olson said they are working on a paper for
LRTS for the whole 10 years worth of work; how the 5 major ILS systems address
them; he felt the best way to get recommendations to vendors is through user
groups - he did III and Tony will try with Voyager
Conference 2000 Action Plan Forum - Bibliographic Control of Web Resources: A Library of Congress
Action Plan - Sandy Roe reported at SAC Monday
LITA/ALCTS Authority Control in the Online Environment -- Fast, Slow,and Z39.19: Emerging Trends
in Subject Authority Control
VIAF
- there are now technological capabilities to allow a linked system of
independent files. Unicode will allow viewing records in original script. OAI
protocols could be used to create a union authority file. In
the Semantic Web, ontologies equal controlled
vocabularies.
NISO
Thesaurus Standard (report by Stephen Hearn). The standard is being revised
because it is too focused on print lists. Things like synonym rings (similar to
Marcia Bates' proposal), taxonomies, thesauri are
being addressed. It can be used to expand search not just to the preferred form
but allow navigation among linked/related/cross-reference terms - all
within/amongst controlled thesauri/lists. Semantics is an issue. They are not
addressing subdivision or classification systems, but are looking at
relationships as is (broader, narrower, etc.) and keeping semantics in mind.
Arlene
Taylor - what have we got to lose? The question: what proportion of records
retrieved by a keyword has that keyword only in the subject field and as a result
bibliographic records would not be found? In one method of searching, 36% would
not have been found by a searcher. In another dealing
with large result sets - 47.2% first found - further work was done to eliminate
misspellings etc and got down to 35.4%. It appears that multi-keyword searches
actually increase the number of bibs not found. It appears that adding TOC
reduces the number not found but precision is reduced.
Ed
O'Neil & Lois M. Chain - FAST: developed to provide a new approach to
subject access for digital resources, etc. Vocabulary should be simple, usable,
optimal, semantic interoperable, compatible, easy to maintain. Advantages of
using LCSH for FAST: rich vocabulary, defacto
standard, has homograph and synonym control, extensive hierarchical relations,
etc. The hierarchy is retained in facets - there are 8: topical, geography,
form, chronological, person, corporate, conference/meeting, uniform titles.
They decided to use MARC21 format but needed a chronological field. Chronology
will only be created when given in a cross-refereces
150, 688 # of occurrences, 750 links to LCS. Decided to do geographics only in
indirect order. First level geographic names will be limited to names
from geographic area codes; other names will be entered as subdivisions under
the smallest first level name within which it is fully contained. They are
qualifying only when otherwise two would be exactly the same. E.g.
151
670
GNIS whenever possible
751
Rowan Bay (
Genre
headings are problematic because many lose meaning when separated from their
topic. Person and corporate had to appear as a subject in the database (not
replacing NACO). Chronological (period) consists of only a date or date range. Limited to a single chronological heading per record.
Establish if in chronological cross-reference. Problem: topic and period
joined, e.g. Civil War, 1861-1865 - decided to split into topical for the event
and place/time for the dates. In LCSH many headings are established, but many
more are synthesized by catalogers based on the Subject Manual. In FAST (except
chrono) all headings are established. Many music
headings are established - ended up with 50,000. Expansion of headings using
patterns ends up with established headings that are not necessarily logical:
Burns
and scalds
Burns
and scalds - Patients (class of person)
Burns
and scalds - Patients - Family relationships (back to topic again)
Future:
resynchronize with LCSH, work on conference/meeting facet and uniform titles;
expand headings with GNIS; expand geographic based on usage; revise and expand
form (genre) facet
Fast
is faceted, hierarchical, fully established, and compatible with LCSH
http://www.oclc.org/research/projects/fast/
John Attig -
FRBR and authority records. Why does it appeal worldwide?
Librarians and catalogers form a view of what needs to be done. System
designers and vendors know it without knowing cataloging rules. It brings focus
to the OPAC, not just the records.
How
does it apply to subject authority control? Primary requirement: to create a
structured framework for relating data in bib and authority records to users
needs. FRBR is basic for bibs. FRAR clarify? identifiers
or standard numbers. User tasks for bib: find, identify, select, obtain. User
tasks for authority records: find identify conceptualize,
justify (document decisions and usage for literary warrant). FRBR: work is realized
in expression, expression is embodied in manifestation, manifestation
is exemplified in item.
Subjectness: "work
has as subject"
Work
has: work, expression, manifestation, item
Work
as: person, family, corporate body
Work
subject: concept, object event, place
Here
is where authority control comes in. Each entity is known by a name. Name is
basis for access point. Access point is governed by rules. Access point is created/modified by agency.
Pre-coordinated
subjects: each concept is independently related to a work
Post-coordinated
subjects: concepts or facets are organized into a structured access point,
which in turn is related to the work
New
entities Tom Delsey added: form, function, objective,
time, physical characteristics
For
sharing authority records: need to know what they agree on in the rules, where
they don't agree, and need to reduce disagreement
IFLA
Meeting on Cataloging Experts discussed an international cataloging code.
Sharing names is not too problematic, except US is only place with option that
one person writing under different names have separate NACO records. Subject
sharing: 1) subject terminology is language determined,
b) subject thesauri are self-contained and operate as a whole
FRBR
- new ways of thinking about bibliographic and authority control sharing
globally.
http://www.libraries.psu.edu/tas/jca/frbr-acig.pdf
PCC --
CONSER summit, ISSN, OpenURLS & DOIs in CONSER bibs - watch for summary of
this meeting after
(see
below)
Standards. ..conference on
monographic/serial standards: working on making a major/minor list for ch. 12,
sponsoring bodies that change, call numbers and classed together series. There
is concern about the direction ISSN is going. Access-level record is in
discussion.
CONSER summit: They discussed a
number of ideas (see CONSER web page for Summit): title level print journal is
still useful; lot of publishers are digitizing backfiles; most researchers
don't care where they get data if the data is okay; interoperability among
systems is needed - if it doesn't work users don't get the content
ISSN (Regina Reynolds): ISSN
revision group has been meeting. ISO 3297 is up for 5-year review. Committee
includes: CrossRef, Bowker, ISBN, major publishers, etc. There are 4 ways to
address ISSN:: a) status quo - ISSN as a manifestation ID, b) change to
title-level ID, c) use one base ISSN plus a suffix, d) hierarchical - title
(top) with an ISSN and manifestations with "lower" ISSNs. None of
these was acceptable as a concensus. Suffix for how many different formats? Too
confusing to gell different between numbers; "D" was the most
unpopular, "B" title level didn't address manifestation. Need:
title-level, manifestation or product-level, and access to authoritative ISSN
metadata. Three point approach: functional granularity, title-level, ISSN
register - better access. Functional granularity: ex. 1) group of formats for 1
title content, ex. 2) 1 format of 1 title.
Ownership of ISSN: ISSN blocks could be given to major publishers; they
would be required to supply metadata, publishers would be
"independent" if assigsning it right, there would be fewer blocks of
numbers they never use. Problem: what if they don't assign right at the level
that works for us? Title-level ID is needed! It needs to be recognizable. Maybe
a URN, DOI, ISTC, in a 024 in bibs and ISSN's records. Why a title-level ISSN?:
need to link to content, need to use open URL, need collocation in catalogs,
need to address all physical forms, editions, and languages. In essence the
problem is FRBR-ish - work or expression = title-level. We need lookup ability
of ISSN database - goal is to be done in 2006.
URLS (Valerie Bross) - linking
matters !! Can we accommodate open URL's? Background developments and work:
CONSER module 31, aggregator-neutral reocrd, CONSER Summit. Possible areas of
work: a) PURL - it is not OpenURL
compliant (http://www.bibpurl.oclc.org),
b) DOI - must be register with CrossRef (ex: embedded in URL: http://dx.doi.org/---/ISSN), c) OPEN
URI.0 is a developing standard. Problem: bibs increasingly have many URLs, some
taking you to same site. So, task force will continue to figure out how to
determing "best", "right" URIs / URLs in bibs
MONDAY--June 28
CC:DA -- Liaison
The first meeting included
the report on JSC meeting, 19-22 April 2004 [CC:DA/JSC
Rep/MLB/2004/1 (
The NISO report included
information on a number of standards up for review: price indexes, holdings, OpenURL, and information system metrics. The 13-digit ISBN
wasn't due to be used for several years but publishers are already using so
both LC and OCLC have given statements of what to do in the meantime. The ISSN
is undergoing scrutiny. Four scenarios have been discussed: 1 ISSN for each
medium but not each format (e.g. 1 for electronic journal, but not one for PDF,
one for HTML, etc.); title level - like a work ID; base number at title level,
then add suffix for each format; or, master ISSN at the title level with
secondary ISSNs for each format and medium. Three
message formats for ONIX have been developed so far: a) online holdings (SOH),
serial product and subscription (SBS), and serial release notification (SRN).
The MetaSearch Initiative has 3 subcommittees for
different components: BA for access managmetn, BB for
collection description (Dublin Core), BC for search an
retrieve.
Rule revisions discussed
included Early Printed Monographs, corrections to rules identified while
preparing Cartographic Materials: a manual of interpretation for AACR2,
work from the Task Force on Consistency across Part I of AACR2, and the Task
Force on SMDs. Despite all the work the Task Force
has done, JSC has decided to work in a different direction. A report at the end
of July will contain rules for area 5 and related notes in general chapters and
special chapters. There will be supplemental chapters based on carriers to
address extent and dimension. In these categories:; books, atlases, aggregated
works bound (scores); book things - loose, not bound; photographs, flash cards,
postcards, misc. graphic; slides, transparencies, projected single format;
3-dimensional; filmstrips, fiche, frames aggregated together on a container;
disk cylinders (medium is integral part of container); reels, cartographic
(medium in container); chips; no carrier. There will be a second set of
supplemental chapters based on medium for other physical description: books,
microforms; tactile; 3-D; projected, still, moving; sound; digital. All this is
an effort to separate carrier and content. Discussion was held on how to handle
e-discussion amongst
members and with non-committee experts.
CC:DA
will sponsor a program for 2005, Cataloging Cultural Objects with the Visual
Resources Association (VRA). They are working toward something like making
their manual an AACR3-like set of rules for visual materials. This could be
followed in 2006 by training. It is a "content" standard.
summer
only OCLC
Luncheon
Sat with NAL cataloger and compared notes on ILS
SAC -- General meeting
LC Action report
(Sandy Roe): reviewed Marcia Bates' report and work expected in next few
months; Reeser presented the LC modes of cataloging report: 1) web guides,
pathfineders are to assure guide or guides are represented by subjects in OPAC,
2) MODS records are up for Minerva and I hear America singing, 3) MARC/AACR2:
a) access level record approach emphasizes content and subject access points,
task based assessment research was contracted for with Tom Delsey, they want
use statistics, and a way to check links on records; b) Greenberg's metadata
generator project - a comparions study was done extracting words in text and harvesting
metatags similar to dc.dot - results show catalogers create best subjects,
researcher/creator next, and machine last
Graphic
novels discussion took place with input from two publishers
Subject
training - the content of the pre-conference will become a SCCTP training
institute; if it works well LC/ALCTS may try other areas of cataloging; PCC
will develop a web page for the course; training fees include buying the manual
and paying for trainer travel
SACSEM
report - I gave
SACO
program - now more like NACO; current members are grandfathered in, i.e. NACO
people are also default SACO members; new members must do 10-12 records/year,
have institutional commitment, encourage LCSH/SACO training, use newly revised
subject proposal form on SACO page; new FAQ says how to join; need good
understanding of LC Subject Manual
ALCTS/SS
Committee to Study Serials Cataloging - watch for reports
The
Committee to Study Serials Cataloging,
Serials Section, ALCTS, has changed its name to Continue Resources
Cataloging Committee.