How can I search authority records using newly available attributes such as “field of activity” or “occupation?”

ANSS Subject and Bibliographic Access Committee
Question/Answer on cataloging issues – July 2016

By Isabel del Carmen Quintana, Harvard University

Question: How can I search authority records using newly available attributes such as “field of activity” or “occupation?”

Answer:

Since Resource Description and Access (RDA) became the cataloging rules for most libraries in 2012-2013, catalogers have been able to supply metadata concerning entities in the authority file in machine readable fields. In the past, librarians could put this sort of information into a text field, but it was only keyword searchable and the information had many variants. For example, it might say that an author was born in Wien, instead of Vienna.

In 2012 there was an ANSS Cataloging Q/A on personal name authorities and how they would be able to be searched in the future. This Q/A is here: https://anssacrl.wordpress.com/publications/cataloging-qa/rda-personal-name-authority-records/

This follow up Q/A shows some examples of how information concerning an entity is being coded and used. For the sake of this Q/A, I’ve chosen OCLC as the database in which I’ve conducted the searches. However, any ILS could have similar indexes to allow users to search on this identifying information for persons and corporate bodies.

In MARC the attributes of an entity are coded in 3XX fields. The currently available 3XX fields are:

336 – Content Type
348 – Format of Notated Music
368 – Other Attributes of Person or Corporate Body
370 – Associated Place
371 – Address
372 – Field of Activity
373 – Associated Group
374 – Occupation
375 – Gender
376 – Family Information
377 – Associated Language
378 – Fuller Form of Personal Name
380 – Form of Work
381 – Other Distinguishing Characteristics of Work or Expression
382 – Medium of Performance
383 – Numeric Designation of Musical Work
384 – Key
385 – Audience Characteristics
386 – Creator/Contributor Characteristics
388 – Time Period of Creation

Here is a section of an authority record for a person, with many 3XX fields encoded:

  • 046   ǂf 1937
  • 1001 Smith, Jane I.
  • 372  Christian-Muslim relations
  • 372  Muslim communities in the U.S.
  • 372  Women in Islam
  • 373  Center for the Study of World Relgion ǂs 1964 ǂt 1980
  • 373  Harvard Divinity School ǂs 1973 ǂt 1986
  • 373  Iliff School of Theology ǂs 1986 ǂt 1995
  • 373  Hartford Seminary ǂs 1995 ǂt 2008
  • 373  Macdonald Center for Christian-Muslim Relations ǂt 2008
  • 373  Harvard Divinity School ǂs 2008 ǂt 2012
  • 374  Associate Dean for Faculty and Academic Affairs ǂs 2008 ǂt 2012
  • 374  Senior Lecturer in Divinity ǂs 2008 ǂt 2012
  • 375  female
  • 377  eng

In this example, the MARC 3XX fields are machine readable and tell us that Jane Smith was at the Harvard Divinity School from 2008 to 2012.

If one searches in OCLC for “Jane Smith” in the authority file you get 151 matches. However, if you use the additional metadata, you can get to the one record that you need. In OCLC the search would be constructed as follows:

Keyword search “Jane Smith” in the Personal Names index, add AND as the Boolean operator, and search “Harvard” in the Entity Attributes index.

Isabel image1

This search will bring you directly to the one record you want.

Likewise you can use the “entity attributes” index to search for corporate bodies. For example, let’s say you are searching for Carnegie Hall, but you don’t want the one in New York. You want to find the correct authority record for the one in Pittsburgh.

If you search for Carnegie in the Corporate/Conference Names index, you get 520 matches. If you search for Carnegie in the Corporate/Conference Names index, add AND as the Boolean operator, and search for Pittsburgh in the Entity Attributes index, you get 7 matches, with “Carnegie Hall (North Side, Pittsburgh, Pa.)” as the first one on the list.

It’s clear that the additional metadata available in the 3XX fields can be very helpful for identifying the correct person or corporate body. However, these fields have only been available for a few years, so most of the national authority file records are lacking this data.

For example, look at these records for Carnegie art museums:

  • 1102 Carnegie Art Museum
  • 5102 Carnegie Cultural Arts Center ǂw a
  • 667  Not to be confused with the Carnegie Museum of Art located in Pittsburgh.
  • 670  Coen, R.N. Elsie Palmer Payne, c1987: ǂb t.p. (Carnegie Art Museum, Oxnard, Calif.)
  • 670  Phone call to Museum, 7/13/88 ǂb (Carnegie Art Museum, formerly the Carnegie Cultural Arts Center [no publs. in L.C. data base])
  • 675  Celebrating sixty years of the arts in Oxnard, 1984: verso t.p. (Carnegie Cultural Arts Center)
  • 1102 Carnegie Museum of Art
  • 5102 Carnegie Institute. ǂb Museum of Art ǂw a
  • 670  American landscape video, 1988: ǂb CIP t.p. (Carnegie Museum of Art)
  • 670  Phone call to Museum, 3/14/88 ǂb (Carnegie Museum of Art, renamed 1986; earlier name: Carnegie Institute, Museum of Art)

The Carnegie Art Museum is located in Oxnard, California. The Carnegie Museum of Art is located in Pittsburgh. However neither record has this information encoded in a 3XX field, so this information is not available in the Entity Attributes index. Furthermore, since the two names are unique, neither name is qualified by the city, making it impossible to tell from an index which is the one located in Pittsburgh. Therefore, it’s important to always broaden one’s search if a more specific search yields no results, and it’s still necessary to page through each record in many cases.

Catalogers continue to enhance current authority records with encoded information. They also strive to create new records with as much information as we have available at the time. However the authority database is large, so it will take a long time to encode all the information we have stored in authority records. Still, the 3XX fields, as indexed fields, are a great way to refine and augment our searches.

Furthermore, in the future, linked data will allow us to link our authority data to additional metadata on the web about the entity. For example, our authority data will link to their corporate main web site, or their Wikipedia page. At that point, we’ll be able to search this metadata on the web, and hopefully bring some authority control, or at least identity management, to certain entities on the web.

Advertisements