Applied Linked Data 2016-07-14
Agenda:
- Linked Data Fragments Update
- Fedora 4 Backend Experimentation
- No way to query Fedora 4 at all. The only option is to hit the root and then cascade down the objects. So... not possible to use it as a backend for linked data fragments.
- Fedora 4 Backend Experimentation
- Harvesting Discussion
- There was an attempt to prototype using ResourceSync and SiteMap at LDCX a few years in the past as an alternative to OAI.
- Discussion from LCDX^4:
https://lib.stanford.edu/libdevconx4/harvesting-oai-pmh-or-not - And the initial code from mjgiarlo & jronallo:
https://github.com/mjgiarlo/sitemapmicrodata
- Discussion from LCDX^4:
- There is a Fedora 4 blaclight implementation by University of Alberta. No one has tried it yet.
- NYU considering an option of writting an OAI provider that references their Solr directly under the hood. Essentially use the Solr API's to do most of the heavy lifting.
- Some code done using Java and XSLT in the past to create such an OAI endpoint exists?
- Steven plans to experiment using an external triplestore to that syncs to Fedora 4 to make triples available for access.
- But the group doesn't think this is the best approach. A document approach like OAI may be better and it is quite well standardized.
- End result is that much will depend on what DPLA wants. As such, Steven will reach out to DPLA to schedule a custom Applied Linked Data call on the topic of harvesting.
- Corey will also ask about the state of blacklight-oai (lacks tests, no sets, resumption token bug, etc)
- There was an attempt to prototype using ResourceSync and SiteMap at LDCX a few years in the past as an alternative to OAI.
- Hydra Connect 2016 Workship
- https://github.com/projecthydra-labs/hydra-connect/issues/63
- Suggestion to add QA to this talk.
- Others may be able to help with it but unsure at the moment for feasability.
- Anna has volunteered to try out any tutorials beforehand.
- Suggestion to perhaps reach out to the OpaqueNamespace people for help.
-
Suggestions for HC2016 program
- Anna may add a group unconference talk about how to choose and how to knit together different linked data components. (Recipes for implementations).
- For example, when do you need QA? Linked Data Fragments? Etc.
- Also perhaps should include the Descriptive Metadata Group in some of this.
- Essentially a standardized way to do Authority Control?
- Anna may add a group unconference talk about how to choose and how to knit together different linked data components. (Recipes for implementations).
- https://github.com/projecthydra-labs/hydra-connect/issues/63
- Call for other topics
- No other topics.
Notes:
- Next meeting: To be Scheduled with DPLA
Next Meeting: To be Scheduled with DPLA