Skip to Search

 

Summary of discussion of UF authority control staff with UNF catalogers

November 8, 2004

 

UNF -- Angela Randtke, Lee Richardson, Linda Smith, Verna Urbanski

UF-- Gerald Langford, Priscilla Williams

 

I. Advancing SUL authority agenda for Aleph – Authorities Subcommittee (and Nancy Williams (UF), who is mostly working at FCLA) could take lead with this

 

Priorities for change (in order of priority):

 

1) Update LCA10 and continue with regular weekly updates.

(If COR fields are automatically added to authority records when the 1xx has been updated, add a $$5 to the COR field to protect it during future overlays, if possible.  (See first question in Section III below.))

 

2) Produce a report for each SUL for each weekly update

Features needed in the report:

a) indicate the records in LCA10 that have been updated or deleted and are linked to the SUL’s catalog

b) indicate the records in LCA10 that have been updated or deleted and for which there is a matching record in the SUL’s local authority file

c) for the updated records in a) and b) above, indicate:

-- control no. of record

-- type of record (name/subject/series)

-- whether a COR field was created

-- any change to the 1— field

-- any changes in 008 positions 06, 09, 11-16, 33

-- any changes in 645, 646, 781 fields

d) for the deleted records in a) and b) above, indicate:

-- control no. of record

Not an immediate priority but a desirable future enhancement:

Give each SUL options to limit the report to authority changes they regard as significant (i.e., changes in the 1xx field, certain changes in the 008 field, etc.)

 

3) Juvenile (“kiddie”) headings (problem 2 on Angela’s list) –

Possible solutions (if all the SULs are agreeable):

Solution A) -- Step 1) remove juvenile headings from LCA10,  Step 2) Create/maintain a separate authority file for juvenile headings or, at least, continue to not load them in LCA10, Step 3) create a separate index in the bib libraries (01) for juvenile headings; remove them from the SUL (Subjects, All) index

Solution B) -- Add $$9 to the juvenile authorities in LCA10 and to the juvenile headings in the bib libraries (This would require ongoing maintenance as LCA10 is updated and as bib records with juvenile headings are added.)

4) False matches between a normalized cross reference on an authority and a bibliographical heading, even when the heading is correct (problem 3 on Angela’s list) (e.g., “Gainesville, Fla.” appears in the heading index instead of “Gainesville (Fla.)”, even when the bib heading, “Gainesville (Fla.)”, is correct)

Lydia Motyka has turned in a problem report to Ex Libris about this.

 

5) Inability to delink and relink headings to authority records when the system has created incorrect links (or relinking otherwise needed)

A major example of incorrect linking:

False matches between a heading and an authority with another tag -- It appears that a link is made to the first record where the term is encountered, whether the term is in a 4xx or 1xx field (problem 4 on Angela’s list).

There are also linking problems which affect all types of headings.  An example of a linking problem with series headings:

A new series heading entering the bib file is automatically linked to LCA10.  If the series authority is copied from LCA10 to the local authority file and modified to show local treatment, the series heading remains linked to LCA10 and cannot be relinked to the local authority file

Possible solution?:

Let (some of the?) SULs change records in LCA10, then change them back, between weekly updates in order to break incorrect links and allow correct relinking

 

6) Long headings – only the first 90 characters of a heading are indexed in auth-03 (“unauthorized headings”) report.  For long headings, this makes it difficult to identify which heading is being reported.  We should check with FCLA to see if this is a problem known to Ex Libris.

 

7) Global change – Ideally, much of this would be done centrally at FCLA. (This would be contingent on approval of supplementary funding for FCLA by the SUL Directors)

Major problems with Global Changes (“manage-21”) service:

a) Cat levels of bib records being changed are set to blank thus giving them no Cat Level protection (Per Daniel Cromwell, Ex Libris is doing programming to fix this but he did now know a time frame for the fix.)

b) You can’t have commas within parameters.  For example, having $$v Handbooks, manuals, etc. or $$a Smith, John B. as parameters won’t work.  (According to Marie Erdman of Ex Libris, putting single or double quotes around the parameter should solve the problem, but it doesn’t seem to.  Daniel Cromwell was going to look more at the PRBs to if any solution is mentioned there.  Failing that, we would like FCLA to submit a problem report or enhancement request.)

Possible interim work-arounds?  (suggested by Linda Smith):

-- FCLA could restore Cat Levels in batches on bib records which had been changed using “manage-21.”  (We believe this has already been done for UWF.)

-- For changes that can’t be done using “manage-21,” develop batch record editing processes outside of Aleph.  (John Hein has been using Macro Express and GenLoad profiling to make changes to UNF NetLibrary records before loading them.)

-- Identify changes needed on bib records being batch loaded for several/all SULs (e.g., NetLibrary, Serials Solutions records).  Have different SULs (volunteers?) take turns cleaning up records before they are loaded in each SUL’s catalog.  The records could possibly be loaded into a test region linked to LCA10 in order to get a report of heading conflicts and unauthorized headings and to identify the biggest and worst changes.  These changes would be included with other local modifications we want for each group.

 

8) Normalization/punctuation problems -

-- Matches on headings that normalize the same – We’d like these to improve.  SUL staff who work with authorities need to compile a list of proposed solutions to forward to FCLA and Ex Libris

Possible interim solution?: Could FCLA add periods before $t (in author-title entries) and $b (in multi-level corporate author headings)?

-- “Heading doesn’t end with period” problem – result is that sometimes another mark of punctuation is automatically removed from the index entry.

Example found in UF, UNF or UWF OPACs (browse search for author):

Faust, Clarence Henry, 1901- loses his comma before the subfield d in the index entry because the bibliographical heading ends with a dash and not a period.  (The comma is actually there in the bib record.) The heading is not linked to an authority record. You cannot tell from the index if there is no link because no matching authority record exists or because the punctuation is missing in the bib record. 

 

II. Long term basic problem (summarized by Angela Randtke)

 

System unable to parse headings - One of the major problems with Aleph authorities is an inability of the system to break a heading into its elements for comparison with the authority file.

Problems with the SULX mechanism (dropping subdivisions with the same type delimiter together, dropping subdivisions in an order based on the delimiter type instead of the order of the subdivisions in the string) means that SULX does not completely solve the parsing problem.

Without proper parsing, linking and cross-referencing is not complete.  The unverified report (auth-03) does not have information about what part of the unverified heading is unverified. The conflict report (print-05) will not report on conflicts with root headings if those headings are only found with further subdivisions in the bibliographic file.  SULX does not pull out root elements with conflicts nor cause them to show up in the conflict report.

 

Workaround implemented by Angela for UNF -- Angela has written programs that parse the headings, then create MARC bib records composed of the parsed headings.  These dummy MARC records are loaded into Aleph as suppressed records.  Once the parsed headings are in a dummy bib record they won’t show up in subsequent reports as “unverified.”

 

III. Other questions/problems which came up after the meeting

 

What happens to bib records having an earlier form of a heading when the form on the authority record changes? – Is Aleph currently set up, when the 1xx changes in an authority record, to add a COR field (for the earlier form of heading)?  "The COR field is treated by the system as another non-preferred term"--Aleph Web Guide.  (In that case, the bib records with the non-preferred term should come up in the conflict report (print-05)).

Even if a COR field is being added, with weekly LCA10 updates, when there's a later overlay, we think the original COR field would disappear.  If that's what will happen, we would like FCLA to add a $$5 subfield to the COR field to protect it during future overlays.  (See section I, priority 1 above.) If that is not possible, we will probably need to run the conflict report weekly to make sure we catch the new COR fields before they are overlaid.

 

Use of “Search field headings…” to check series, uniform title headings beginning with articles – Catalogers sometimes use the “Search field headings …” function (shortcut key, F3) in the Catalog module to see if a heading in a bib record is linked to an authority record or to search for an appropriate authority record.  When a series or uniform title heading begins with a grammatical article, the article is included in the heading search and the cataloger must rekey the title in order to get to the correct place in the index.

 

Inability of “Search field headings…” to check the NWT index

Possible solution? – When the the “Search field headings …” function is used and the headings index appears, one of the options is to “Jump to” another heading.  When the “Jump to” function is used, it is possible to change the “Scan Code” (browse index label).  Could “NWT” be added as one of the “Scan Code” options?

 

Written by: Gerald Langford

Last revised: December 1, 2004

 

 


 

University of Florida Home Page