2011 Forum questions with and without notice

When the offline duplicate detection program (DDR) finds two or more potential duplicate records, how does it decide which record to keep in the ANBD?

The offline duplicate detection program keeps the record that has the

  • higher coding in Encoding level Leader 17
  • greater number of links to authority records ($9 subfields)
  • higher coding in Descriptive cataloguing form Leader 18
  • greater number of holdings

An algorithm processes this information and the record with the higher score is the record that is kept. The DDR script only matches and merges a single pair of records at a time.

How does the Relate program, which links access points in bibliographic records to authority records, handle headings that contain subdivisions?

Relate does an exact match on the text.  It disregards what subdivisions may be in the heading, differences in case or punctuation.  The only exceptions are subfields that are treated as notes, e.g. $e relator terms in personal names.

For instance, matching where the text is exactly the same, but there are fewer subdivisions in the bibliographic record heading:

Bibliographic record tag
$aANZAAS. Congress (41st : 1969 : Adelaide, S. Aust.)
matches
Authority record tag
$aANZAAS.$bCongress$n(41st :$d1969 :$cAdelaide, S. Aust.)

Bibliographic record tag
$aNew York University Institute of Philosophy$n4th$d1960$cNew York, N.Y.
matches
Authority record tag
$aNew York University$eInstitute of Philosophy$n4th$d1960$cNew York, N.Y.

In this example the bibliographic record heading has subdivisions the authority lacks, and also differences in case and punctuation:

Bibliographic record tag
$aC.A.C.I., inc.$bFederal.
matches
Authority record tag
$aCACI, Inc. - Federal 

Is there a simple method to identify free e-resources on the NBD that could then be loaded into our local LMS?

There is no simple or exhaustive way of identifying freely available e-resources. There isn’t a single flag that can be used to identify all free resources while reliably excluding all others.

You could consider a combination of 856$u plus 506 (Restriction on access note). However 506 is free text and can be used to indicate that access is either “restricted” or “free”.  You could consider excluding records with terms like ‘licence’ or ‘restricted’ in the Note Keyword. Using 856$u plus Government publication codes could identify online government publications that are more likely to be freely available. You could also exclude records with 856$xNUC because these would usually be for subscription e-resources.

For finding freely available pdfs, consider a combination of Any Keyword equal to pdf and Note Keyword equal to http but not subscription.