Index 653 contents as keywords

[escowles]

Index MARC field 653 (uncontrolled index terms) contents as keywords

Concrete example

Re-index required?

• [x] Full re-index
• [ ] Partial re-index
• [ ] No re-indexing required

Implementation notes, if any

Approved by DSSG in April 2021

[sandbergja]

@escowles a few questions about this one:

• Do we want these indexed for display, search, or both?
• If for display, what is the desired display? Underneath the subject headings? Should they be links or not? If a 653 term is the same as a subject heading in the record, do we still want to display the 653 separately?
• Are there any specific example records that DSSG had in mind?

[escowles]

@sandbergja Sorry I didn't see your questions. We do want them indexed for both search and display, displaying them below subjects with the heading "Keywords". I'll ask DSSG for examples.

[escowles]

Apparently we have been stripping this field because it isn't being indexed, so we don't have sample records. So we can probably use the example values from the MARC docs.

[kevinreiss]

Display these values as regular text not links. Make a ticket in OL to do thi after indexing.

[maxkadel]

Starting in February 2023, the 653 fields were no longer being stripped. @mzelesky is finding examples to share for testing.

[mzelesky]

Record with a 653 field that has multiple subfield a: 9925459873506421

Record with multiple 653 fields: 9925493943506421

Miscellaneous records with 653 fields: 9925508433506421 9925505643506421 9925496713506421 9925496093506421 9925494973506421 9925494093506421 9925493823506421 9925492163506421 9925474453506421 9925473353506421 9925466353506421

[mzelesky]

Here is a report of all 653 fields in PUL records, including the number of bibs that contain each field. https://docs.google.com/spreadsheets/d/1VgWYcZ1yW1EdlxN4G2tX5chIMOvZohXfg84tT6wwgBw/edit?usp=sharing

[maxkadel]

Per Esmé, whether / how this field should be indexed is still under discussion by DSSG.

[escowles]

Jennifer reported back that she looked for email or notes on why she requested this and couldn't find anything. So I'm going to close this issue and we can revisit if she remembers where this request came from or someone else asks about it.

[escowles]

Reopening as new use cases for indexing 653 have surfaced. Minjie noted that non-Roman items because other fields included Romanized values, often without translation, but 653 includes English-language topic/format terms. A sample record is: https://catalog.princeton.edu/catalog/9951301943506421 (which includes "toy making" in a 653 which is otherwise not included in the record).

DSSG decided that these should be indexed but not displayed.

[sandbergja]

We can discuss this in the next Orangelight meeting.

[maxkadel]

@mzelesky is following up with stakeholders and will come up with a plan. May take a few weeks.

[mzelesky]

A meeting is scheduled for next week.

[kevinreiss]

No final decision until Jennifer returns from vacation at the end of Sept.

[mzelesky]

I met with Minjie Chen, Don Thornbury, and Beck Davis today about the 653 fields in Cotsen records.

The agreed plan of action is to compare all the 653 fields in the Cotsen records with all LCSH authorities (authorized headings and references) to see which ones can be easily mapped to LCSH with no further intervention.

After that process, we will evaluate what to do next. One potential solution raised by Minjie was to add 500 fields for the remaining Cotsen records with 653 fields similar to the following format:

500 __ $a Uncontrolled keywords: Toy-making; Toys, wooden.

[maxkadel]

@mzelesky - Moved to backlog. Let us know whether further action is needed from the DACS team from this, or whether it can be closed.

[mzelesky]

I believe this can be closed. Other avenues are being pursued for the 653 fields in Cotsen.

[kevinreiss]

Discussed at Alma-DACS 1/9. CaMS is pursuing other avenues to make this data usable using LCSH subject headings added to the records.