50 min
| Presentation on Linked Data App | Doug Loynes | Presentation: https://folio-project.slack.com/files/U04TN5JGMCM/F084D41MN6M/linked_data_module_product_council_pre-final_dec_12_2024.pptx Scoped to monographs as the logical place to start, but not limit of what we want to do Fully integrated with FOLIO There are workforms in the LD editor for cataloging, and everything is in sync with FOLIO inventory, this was big effort for Ramsons Looking at the Roadmap, we’ve accomplished all that listed under LD on the Roadmap for Sunflower Still working on authority look-up, validation, polishing, accessibility Looking further ahead, would like to move beyond MARC-centric workflow to allow for direct import/export of RDF BIBFRAME Promise of where we’re going is to allow an entity management knowledge graph and be able to manage the entities directly, but this is post-sunflower
Review of back-end and front-end modules have been submitted to TC for the TCR - all under the OLF license and to be added to Github for Sunflower Linked data storage is populated through MARC in SRS. The MARC records are transformed and processed into linked data records in the library’s data graph. The only editing that will be initially available is on the bib records and not authority records. We can populate the data graph with authority resources. When something is linked data, it can only be edited as linked data and not as a MARC record Interactions with existing FOLIO modules Data import works the same as previous, import is an a MARC record Inventory: now there will be an option to edit the MARC record in the Linked Data Record Must be a monograph Must either have MARC record with in SRS or be brand-new in the linked data editor Cataloger must linked data editor permissions. Once a resource is edited as linked data, editing further in MARC will not be possible
Data Export: this will also be MARC
In Sunflower, linked data editor will be available for testing in the Bugfest environment. Previous testing done with Pilot libraries Documentation will be added to docs.folio.org
Demo: When not a monograph, the linked data editor option will not appear If you choose to not edit in linked data, you can cancel out In editor, work components are on the left, and instance components on the right - this is based on the monograph workform for BIBFRAME Can see when multiple instances are associated with a single work (e.g., editions of the same title) and select the appropriate instance. Integrated the MARC authority API into the linked data editor, so catalogers can query authorities to find authorized headings If MARC comes in without a match on a heading, it will create an authority on the fly, and catalogers can use the lookup features if needed to bring in an authorized heading Changes to the work need to be saved prior to editing the instance Demo ran into some difficulties with fetching the data, but there is an option to see the MARC-derived record. If you make changes in the linked data interface, you can see the changes reflected in the MARC-derived record Can export out combination of source=MARC and source=linked data records, because a MARC-derived record is made available in SRS
Q&A For discovery, Locate will work off the records in Inventory. Works will be represented in discovery because of inventory records UUIDs are synced up between FOLIO and the linked data module. Inventory assigns all UUIDs. APIs: currently driven by Kafka messages, so there are no APIs currently; however this could be coming
How is data stored: As Builde - rebranded resource that was known as BIBFRAME Lite Data graph storage uses Builde scheme that allows it to mapped to multiple vocabularies and profiles, e.g., LRM Openly licensed and shared through GitHub repo Looking at extending Builde to item descriptions
Looks like a lot is hardcoded right now, e.g., Kafka messaging, so how configurable is it if using other LD editors? Also would like to be able to edit more than just monographs Use REST API and limited with those constraints, so we will need to extend the endpoints or create new endpoints For Kafka, you should be able to connect to Kafka, depending on existing flows Each editor will need to work on the backend to front end components Not yet ready to make API more flexible to work with other formats, but know this is needed. Iterative approach to begin with only one profile
Under new Instance creation, how much is inherited and how much do you have to redo? 1. create instance from scratch with blank form 2. edit selected instance 3. Duplicate new instance on based on existing instance
Will the syncing for Locate be available for other Discovery layers? It will be based off of Inventory, as resources cataloged through the LD editor will be available through inventory, but there is not an API to connect to the LD editor directly If you are pulling anything from mod-record-storage, it will not have the record in the LD Still working on mapping between MARC and LD What’s in the LD editor today will not be available through the OAI-PMH harvester, but that is desirable
to add: video of editing/creating instances and documentation diagram from WOLFCon Meeting chat: 09:17:56 From Kristin Martin to Everyone:
Can you edit resources where source=FOLIO?
09:19:55 From PK Jacob to Everyone:
Replying to "Can you edit resourc..." 09:20:47 From PK Jacob to Everyone:
Replying to "Can you edit resourc..."
“Edit in linked data editor” option is available when source=MARC or source=LINKED_DATA
09:21:24 From Charlotte Whitt to Everyone:
Will the end-user documentation be incl. in http://docs.folio.org ?
09:22:32 From Christie Thomas (she/her) to Everyone:
Does there need to be an instance in Inventory to start a new linked data description?
09:23:42 From Charlotte Whitt to Everyone:
Replying to "Can you edit resourc..."
Will it be possible later to edit records with source=FOLIO ?
09:24:18 From Charlotte Whitt to Everyone:
Replying to "Can you edit resourc..."
Or other source IDs for that matter; e.g. LIBRIS?
09:34:23 From Stewart, Caitlin to Everyone:
Live theater! We've all been there.
09:34:28 From Gloria Gonzalez to Everyone:
Reacted to "Live theater! We've ..." with 🤣
09:34:31 From Lisa McColl to Everyone:
Reacted to "Live theater! We've ..." with 👍
09:34:37 From Charlotte Whitt to Everyone:
Reacted to "Live theater! We've ..." with 💯
09:35:26 From Stewart, Caitlin to Everyone:
Want to describe expected behavior and send us a video of what it looks like when it's cooperating later?
09:35:46 From Laura Daniels to Everyone:
Reacted to "Live theater! We've ..." with 💯
09:36:01 From Charlotte Whitt to Everyone:
Doug it is looking really pretty. I like the way you display the titles connected to the given Work
09:36:02 From Alexis Manheim to Everyone:
Reacted to "Live theater! We've ..." with 💯
09:36:14 From Laura Daniels to Everyone:
Reacted to "Doug it is looking r..." with 👍
09:37:47 From Charlotte Whitt to Everyone:
Replying to "Does there need to b..."
No. The record can start out as original catalogued Linked Data record
09:37:51 From Gloria Gonzalez to Everyone:
Reacted to "Doug it is looking r..." with ❤️
09:38:02 From Gloria Gonzalez to Everyone:
Replying to "Want to describe exp..." 09:39:19 From PK Jacob to Everyone:
Looks like someone restarted the Kubernetes env. Environment is available now.
09:39:28 From Anuar Nurmakanov to Everyone:
Reacted to "Looks like someone r..." with 👍
09:40:14 From Gloria Gonzalez to Everyone:
Reacted to "Looks like someone r..." with 👍
09:40:53 From Charlotte Whitt to Everyone:
Will the syncing you have implemented for the EBSCO Discover, be something VuFind implementations can take advances off too?
09:42:34 From Gloria Gonzalez to Everyone:
Here’s the new Builde website: https://bibfra.me/
09:42:49 From Gloria Gonzalez to Everyone:
And here’s the Github repository: https://github.com/ebsco/builde
09:43:12 From Lisa McColl to Everyone:
Replying to "And here’s the Githu..." 09:46:05 From Gloria Gonzalez to Everyone:
Reacted to "No. The record can s..." with ❤️
09:55:52 From Kirstin Kemner-Heek to Everyone:
I need to leave. Thanks a lot for this.
09:57:04 From Andreas Mace (NLS) to Everyone:
Read/write with the api (or several api endpoint) is a key point for all integrations. Both other editors, as well as frontend/discovery.
09:57:36 From Gloria Gonzalez to Everyone:
Reacted to "Read/write with the ..." with ❤️
09:58:35 From PK Jacob to Everyone:
Reacted to "Read/write with the ..." with 👍
09:59:27 From Anuar Nurmakanov to Everyone:
Reacted to "Read/write with the ..." with 👍
10:00:15 From Charlotte Whitt to Everyone:
It is exciting that FOLIO finally will be compliant with Linked Data editing; and that in future FOLIO can connect with external editors
10:00:20 From Gloria Gonzalez to Everyone:
Reacted to "It is exciting that ..." with ❤️
|