SERP (Search Engine Results Page) Results Order

[sourslaw]

Descriptive Summary

• On current Umedia the SERP display is in alphabetical order; however, it would be useful for the display to be numerical and follow the order of boxes and folder numbers

Expected Behavior

• Ultimately, the desire would be for the SERP to display items in a collection in-line with the finding aid, as so

Examples

• Example of SERP display order on current Umedia
• Example of finding aid order desired

[jasonoroy]

sorting on DLS identifier might be the easiest way (in most cases) to sort by box and folder since the file naming follows this sort of ordering.

…

On Tue, Sep 30, 2025 at 12:56 PM Scott Lawan ***@***.***> wrote: *sourslaw* created an issue (UMNLibraries/cdm-blacklightify#126) <#126> Screenshot.2025-09-30.at.12.52.11.PM.png (view on web) <https://github.com/user-attachments/assets/4bafd4db-844a-4705-ae50-b25052cf509a> Screenshot.2025-09-30.at.12.53.15.PM.png (view on web) <https://github.com/user-attachments/assets/3b60b60a-c644-441f-a53e-47f79cd35f92> Descriptive Summary - On current Umedia the SERP display is in alphabetical order <https://umedia.lib.umn.edu/search?utf8=%E2%9C%93&sort=&q=y.usa.40>; however, it would be useful for the display to be numerical and follow the order of boxes and folder numbers Expected Behavior - Ultimately, the desire would be for the SERP to display items in a collection in-line with the finding aid, as so <https://archives.lib.umn.edu/repositories/7/archival_objects/1086792> Examples - Example of SERP display order on current Umedia <https://umedia.lib.umn.edu/search?utf8=%E2%9C%93&sort=&q=y.usa.40> - Example of finding aid order <https://archives.lib.umn.edu/repositories/7/archival_objects/1086792> desired — Reply to this email directly, view it on GitHub <#126>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AARCFFVLHAV3CF6P7Q2PD4D3VK75RAVCNFSM6AAAAACH5ST7HCVHI2DSMVQWIX3LMV43ASLTON2WKOZTGQ3TANBYG44DSMQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- Jason Roy Director | Cataloging, Metadata, & Digitization Services 160B Wilson Library | 309 - 19th Avenue South | Minneapolis, MN 55455 University of Minnesota Libraries | lib.umn.edu | 612-625-0028 ORCID: 0000-0002-3644-1970 <https://orcid.org/0000-0002-3644-1970> he, him

[theberg75]

Agree with Jason that sorting by DLS identifier might be the easiest way to address this; however DLS will need to do some metadata clean-up to ensure that all DLS identifiers associated with an archival collection (i.e., Box, Folder situation) are formatted with the same padded zeroes. Historically, we have not always used 3 digits/placeholders so the formatting of identifiers may vary within a collection. If we make a decision about this soon, I can have my student volunteer work on this clean-up.

Another consideration, though, if we sort by DLS identifier: For items whose identifier is a barcode (e.g., the identifier starts with 3 or X), the barcode is just a random number. The sorting may thus appear random to users, so we may need to write an exception for theses cases to rely on standard alphabetical sorting.

I would also like SOME collections (not all) that have barcodes as the DLS identifier - mainly serials and newspapers - to sort chronologically by the Date field. OR, I would like a button at the top of the search results page for these particular collections that says "Sort by date" for users to choose that sort capability. This would incorporate user studies that confirm newspaper searching is done first by date, and then by search term within a date range.

[theberg75]

Thinking out loud: To help with the box, folder situation and sorting, maybe we sort by DLS identifier ONLY if the Item Type = Mixed Material (which is what we apply to all of our archival collections). (DLS can review those 2 fields and make sure the metadata is clean before implementing if we decide to go this route).

This doesn't address the point about newspaper sorting though.

[sourslaw]

update based on 2025-10-16 meeting:

• DLS will clean up dls_identifier field
• IT can then parse to solr field to sort on (cdmdexer transform)

[mberkowski]

Parse box and folder from dls_identifier and store in new solr fields archive_box, archive_folder