Message-ID: <199812090103.UAA19760@lists.ufl.edu>
Date: Tue, 8 Dec 1998 17:03:19 PST
Reply-To: "Z39.50 Implementors Workshop" <Z3950IW@LISTS.UFL.EDU>
Sender: "Z39.50 Implementors Workshop" <Z3950IW@LISTS.UFL.EDU>
From: Lennie Stovel <BL.MDS@RLG.ORG>
Subject: Report of October bibliographic attributes meeting
To: Z3950IW@LISTS.UFL.EDU
Status: O

Bibliographic attributes
Report of the October 21, 1998 meeting
Madrid, Spain

About 15 people attended the post-ZIG meeting about library
bibliographic attributes.  It was a good typical ZIG discussion --
lots of smart people with lots of good ideas contributing to
solutions.  Too bad there were not more primarily-bibliographic
people there.

I asked for agreement to the scope statement I had posted to the ZIG
(included below for convenience).  Matthew Dovey asked, "If there
are rules for cataloging it, does it fit within the scope?" I
realized I really had rules for cataloging library materials in
mind.  People agreed on the scope.  No acceptable name made an
appearance, though there was agreement that the outcome would not
necessarily be "bib-2" because it does not aim for generality.

I suggested that the group might develop two attribute sets.  One
would support searching by MARC tags (fields and subfields); the
other would support searching by "abstract attributes".  Joe Zeeman
had previously submitted a suggestion for Use attributes for MARC
tags that could be the starting point for the former.  It was
generic, i.e., not tied to a specific MARC format.  It was noted
that a way is needed to specify which MARC tag set the query uses;
this could be a semantic qualifier.  Alternatively, there could be a
separate attribute set for each format, e.g., USMARC Bibliographic,
USMARC Authorities, UKMARC, etc.

For lack of a better place to start, I suggested we pick "name" (as
a concept) and try to work out the problems.  From bib-1, we have

   name
   personal name
   corporate name
   conference name
   geographic name
   publisher name
   editor name
   author
   subject

After trying out a few combinations of access point, semantic
qualifier, and format/structure attributes, we ended up with this
list:

  access point: name
    semantic qualifiers: personal, corporate, conference,
      geographic, family, forename, surname

  access point: author (main and added entry)
    semantic qualifiers: personal, corporate, conference

  access point: publisher
    semantic qualifiers: none

  access point: subject
    semantic qualifiers: personal, corporate, conference, form,
      chronological, title, topical

The structure attribute could be name + role.

We could develop specific ASN.1 data structures or datatypes for
various kinds of names.

It is probable that the access points for name and subject could
come from the cross-domain set.

We agreed to the following next steps.  Joe Zeeman will
draft or redraft the MARC attribute set.  Lennie Stovel will write
up the notes of the meeting, take a stab at expanding this list, and
make a list of bib-1 attributes that would most likely not appear in
this attribute set.

Organizationally, Lennie will set up a list for discussion of this
attribute set(s) and a Web site.  Clearly another meeting is
indicated.  Its time and location are to be determined.

Notes by Lennie Stovel, 12/8/98

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Bibliographic attributes working group: scope and statement of work

1. Goal

The goal of the working group is to develop Z39.50 attribute set(s)
for searching bibliographic records that conform to the attribute
architecture adopted by the Z39.50 Implementors Group in June 1998
(<http://lcweb.loc.gov/z3950/agency/attrarch/arch.html>).  The
attribute set(s) may be supplemented by profile(s) for their use.

2. Scope

The attribute sets developed by this group are intended for
searching databases of records with cataloging data created
according to accepted cataloging rules.  Typically these records are
found in library bibliographic databases or online public access
catalogs.  Materials described by these records are of any type.
Servers hosting these databases most likely can present records in a
MARC record syntax based on ISO 2709.

Initially the scope of this working group will exclude citation
databases (e.g., from abstracting and indexing services).

3. Statement of work

The first deliverable from the group might well be an attribute set
based on MARC tagging and structure, i.e., ISO 2709 as implemented
by the various MARC formats.  (Two earlier proposals for a MARC
attribute set should form the basis for this work.) The second
deliverable might be a set of abstract attributes not tied to MARC
but closely related to accepted cataloging rules.  A third
deliverable might be a profile for use of these attribute sets.  A
fourth deliverable should be input to the ZIG for use by other
developers of attribute sets.

4. Plan of work

Select a European co-chair.  (done; it's Liv Holm)
Constitute the group from volunteers and recruits.
Establish mechanisms of electronic communication.
Plan for face-to-face meetings. (will meet Oct. 21 in Madrid)
Appoint an editor (if not the co-chairs(s)).
Agree to a fuller scope statement.
Agree on a statement of work.
Get busy on the deliverables.

To:  Z3950IW(Z3950IW@LISTS.UFL.EDU)