2020-05-21 Metadata Management Meeting notes

No product council meeting this week.

Of most concern for institutions maintaining MARC holdings, but also of interest to institutions that will be using MARC holdings to migrate holdings data at implementation.

Data export: Next week will be discussing export of MARC holdings. It would be great if there were some overlap on the group. On the fly records generated from native FOLIO holdings will be a simplified holdings record and will not be as comprehensive as a MARC holdings records. 

Item status: Emma Boettcher will be joining us in early June for an update.

FOLIO Forum: Good overview if Inventory and QuickMarc. Link to forum recording: 

MM Reporting working group: completed the External Holdings - HathiTrust prototype and are working on the report itself.

Homework for next week: see use cases for pre-order searching

Are there additional use cases for cross-app searching?

What fields would you expect to be able to search on?

How would you expect results to display? (What/how much information?)

Add use cases to the pre-order searching document, but specify that a different use case.

• Functionality available in Fameflower (today's demo)
  • Match based on incoming MARC and any field in Instance, Holding or Item
  • Create (only create, no update) MARC Bibs, Instances, Holdings, Items
  • Field mappings for Instances, Holdings, and Items, but some limitations (e.g. UUIDs)
  • Assembling matching, action, and field mapping profiles into job profiles
• Functionality available in Goldenrod (coming)
  • quickMARC-SRS-Inventory integration
  • Create and Update for MARC Bibs, Instances, Holdings, Items
  • Better field mappings for Instances, Holdings, and Items
  • MARC-MARC matching based on 001 (HRID), 999 ff (UUIDs)
  • MARC modifications for incoming files
• Key features still to be developed
  • MARC-MARC matching on fields besides 001 and 999 ff
  • MARC field protection during updates
  • Storage of MARC Holdings and MARC Authorities
  • MARC Holdings-Inventory Holdings default mapping
  • Importing Order and Invoice data via MARC Bibs
  • Updated import job logs and error/exception handling
  • Job preview functionality
  • EDIFACT invoice import
  • SRS-MARCcat Integration

The key to making data import work is having a job profile - defines what should happen to the records on import. job profiles are made up of match and action profiles.

• Profiles are managed in Settings > Data Import.
• Match profiles are used to target records during import.
• Action profiles define what should happen when match criteria are/are not met.
• Match happens first, then actions are taken.
• Actions are taken in the order they are in the action profile. 
• Can also have submatches. 
• There is a record which jobs use a profile in the view profile mode. 
• Actions will be create, modify, update.
• Currently, only create works and only for instance, holdings, and item records. Cannot create orders and invoices yet. 
• Update - will be able to handle protected fields. 
  • 001 and 999 with indicators ff re system protected fields. 
  • Will create a settings option to allow a library to indicate fields that should always be protected.
  • Will have options to provide exceptions to the global protection for specific imports. 
• Field Mapping
  • Default MARC to Instance mapping will be used for bibliographic data. 
  • Field Mapping profile is just for the fields that are not controlled by the MARC to instance mapping. 
  • Fields that cannot be mapped will be grayed out in the interface. 
  • Date picker:
    • could map from MARC if in YYYY-MM-DD.
    • Could map "Today's date"
    • Could assign a static data. 
    • Anything in "" is a static value that will be mapped literally - constant data.
    • Cataloging date is not a required field, so can be blank. 
  • Can reuse field mapping profiles with multiple Action profiles. 
  • If a list of reference data, then drop down menu will include that data. Cannot enter data as free text. 
  • There are still some bugs.
    • Statistical codes are currently problematic. 
    • Value tables - In drop down lists, system is expecting the UUID and not the name/label. 
      • Can get the UUIDs from the developer tool. 
      • Will definitely be fixed for Goldenrod. 
  • Can have fall back mappings, e.g., look for call number in 090. If not present, look in the 050. 
  • Item statuses don't have UUIDs; it is a hard-coded list. The wrong list is in Fameflower. 
• Data Import:
  • Secret button - uses old technology. New technology is pub-sub. 
  • Can view created jobs - new jobs will be at the top of the list. 
  • Others are "fake" sample data.
  • Run the job. 
  • Can view log as JSON and pull the IDs from the JSON then search for the record in Inventory to view the import. 
• Questions:
  • What about multiple matches?
    • Can the import match the first record matched? 
    • Still working through the data import response to multiple matches. 
  • What will replace encompass? 
    • Replace the entire record? (Needed for authority control and export - transform - load produres)
    • Protecting certain fields in the FOLIO records?
    • Replacing only a single field in the FOLIO record?
    • Update features will be available in GoldenRod. 

features:

https://docs.google.com/spreadsheets/d/1IoGBR4uWXCTDeag5tlMf0fRD0BFDb8r84IOgS0ZUUUE/edit?usp=sharing

ranking: https://docs.google.com/spreadsheets/d/1g8pCNmg2DrGJzUAtQHXPas8M9WIDXijnY4q666teeDc/edit#gid=0

Action: How can we effectively address features and prioritization? How can we look collectively at what has not yet been developed and what is most important?

Add ideas here

https://docs.google.com/document/d/1u1tvfVxn6i3fTImrWWCYHlh7RT5cf3_G7YgDuxKob4M/edit

VZG

Stanford

Index Data