Monthly Authorities Processing of Bibliographic Records

Summary: At the start of each month, the Metadata Management Unit sends new bibliographic records from the previous month to Backstage for authorities processing and loads the updated bibliographic records as well as any associated authority records back into FOLIO.

Gathering bibliographic records

1. Use the LDP app to collect the UUIDs of records that have a cataloged date from the previous month. Because of the 1,000 result limit in LDP, records need to be gathered in batches of approximately one week's worth of cataloged dates.
   1. Settings are:
      1. Schema: public
      2. Table: inventory_instances
         
      3. Filter: cataloged_date <= [end date in the form of YYYY-MM-DD]
      4. (add filter) Filter: cataloged_date >= [start date in the form of YYYY-MM-DD]
      5. Show Columns: id
      6. Limit results: 1000
   2. For each search, download the resulting CSV file. Open the CSV and copy its contents to a single CSV for the month.
2. After collecting the entire month's UUIDs, dedupe the CSV. There will likely be between 3,000 and 4,000 UUIDs.

Creating a MARC file for Backstage

1. Load the monthly CSV created in the previous steps into Data Export in FOLIO.
2. Select "Default Instances Export Job Profile," then select "Instances" in the drop down box. Click "Run."
3. After a few minutes, the MARC file will appear in the export log. The name of this file will be the same as the name of the CSV. Check that the number of records retrieved is the same as the number of UUIDs in the CSV.
4. Click on the name of the MARC file to download the MARC records.

Checking the MARC file

1. Open the MARC file in MarcEdit.
2. Deduplicate the file based on the 001 field.
   1. If there are duplicates, delete both records from the file and send the HRIDs to the Head of Cataloging. Duplicates usually result from a Data Import job that includes two MARC records that point to the same instance HRID. The only way to solve the problem is to abandon the existing instance and bring in the MARC record again on a new instance (thus, the new record will be included in the following month's authority processing). To confirm that this is the problem, look at the instance to find the cataloged date and cataloger name, then browse through Data Import to find the import job and confirm that there were two identical MARC records in that job.
3. Save changes and compile back into .mrc.

Sending the file to Backstage

1. In a browser, go to https://ac.bslw.com/mars/. (Login info can be obtained from the Authorities Librarian if needed.)
2. Select the "Jobs" tab and click "Add a new job."
3. In the drop down box, select "MAX."
4. Click "Browse," select the MARC file, and upload it.
5. When the file is uploaded, click "Send."

Collecting the processed bibliographic records, authority records, and reports

1. When the file is done processing, the Authorities Librarian will receive an automated email from MSUL's Backstage representative stating that the files are ready to be downloaded.
2. Navigate to the "Jobs" page at https://ac.bslw.com/mars/ and click "Retrieve files" under the relevant job. Download both \_MARS_Files.zip and \_MARS_Reports.zip (these include all the other files and reports listed on the page).
3. Once files have been downloaded, click "Mark files as downloaded" at the bottom of the page. This moves the job from "New" to "Previously delivered."

Opening and preparing the Backstage files

1. Unzip the Backstage files.
   1. \_MARS_Files.zip contains all MARC records needed for the next step as well as the Excel versions of the reports. \_MARS_Reports.zip contains the same reports in HTML and the R01-Statistical Summary report that is only delivered in HTML.
      
2. The following files are not needed and can be deleted:
   1. All files ending in ".formerly.csv"
   2. All files that are 0 bytes in size

Loading records back into FOLIO

1. Load the authority records according to the procedures described in Monthly Authority File Updates
2. Load the bib records file (BIB.mrc) using Data Import with the AUTH3 loader. Generally, Data Import can only handle approximately 5,000 records per load so it may be necessary to split BIB.mrc into multiple files using MarcEdit.
3. When the record load is complete, check for errors by clicking on the file name in the Data Import log. Investigate any records that failed to load and keep track of any records that are not able to be edited or overlaid.
4. Record the number of bibliographic and authority records loaded in the "Authority Statistics" spreadsheet ("Bibs" tab) in Google Drive. (Contact the Authorities Librarian if access to this spreadsheet is needed.)
5. Begin work on Backstage reports in the following order:
   
   1. Priority 1: Names, Subjects, Titles, Series Delete reports.
   2. Priority 2: C1XX Authority Change report.
   3. Priority 3: Near Match report. (Work on Personal names first, and other sections as there is time.)
   4. Priority 4: All the other reports in the following order:
      1. Leading Article Deleted
      2. Filing Indicator Changed
      3. Suspicious Filing Indicators
      4. Possible Leading Articles
      5. Heading Usage Not Authorized
      6. Split Headings (LC Subjects) – also see Bad Subject Flips - List of Terms
      7. Tags Flipped
      8. Subdivisions Flipped (LC Subjects)
      9. Subdivisions Flipped (Local Subjects)
      10. Unrecognized $z (Geographic subdivisions)
      11. Unrecognized Relator Terms (scan for spelling errors)
      12. Unrecognized $c in personal names (scan for spelling errors)