dataset_description.jsonld

[CPernet]

Given BEP18 (provenance) and the genetic descriptor, @jbpoline pointed to me we could turn dataset_description.json into a dataset_description.jsonld (ie add that) to download automatically @satra @cmaumet

[satra]

@CPernet - we can turn any of the json docs in bids to jsonld. we just have to agree with the way the context file is encoded. we can keep the json as is, for backwards compatibility, and simply add an "@context" key i believe. we would have to go through the exercise of crafting this context file though, and there will be some places where we will have to introduce some new constructs.

@jbpoline is using DATS in CONP, we in DANDI are using a mixed variant from schema.org + DATS + Datacite, and if we can agree on some alignment here with BIDS, that would be great. within the next few weeks we will have datasets and assets in DANDI speak a jsonld compliant schema. so perhaps that would also help.

perhaps we can arrange a 2-3 hour discuss/hack session around this.

[effigies]

Tagging in @nellh. OpenNeuro does automatically promote some dataset-level data into a JSON-LD snippet in the dataset landing page to make them intelligible to external services.

Has a call/hack already been arranged, or were you waiting for more interested parties? I'm happy to join in, but I'm not sure how useful I would be...

[CPernet]

@effigies as you know @jbpoline mentioned it - and given @satra @cmaumet interest via provenance BEP that would make sense - if you already have a hack to turn json to jsonld I'm sure that's of interest! nothing done on that front (or any on my side anyway)

[cmaumet]

Very happy to join the effort as well! Looping in @remiadon who is currently working on the Provenance BEP with us!

[effigies]

Should probably get around to scheduling this. Would the second or third week in November work for people? If so, I can send around a poll with times.

[effigies]

Ah, thumbs-ups don't notify me... Here's a poll, though it might be a tight squeeze now. https://doodle.com/poll/8uc8mcg8pzsm22ar

[effigies]

Nov 16 @ 8-10am EST is the only consensus point for all current respondents. Are we interested in a short call to get on the same page or should we block off the whole two hours to try to get something done live?

This would obviously exclude any potential Pacific participants, so if there are others interested in getting involved, I can extend the poll range.

[cmaumet]

Thanks @effigies! I'd be in favour of a short call to get us on the same page and identify todos we can work asynchroneously on. Happy to hear what others think!

[effigies]

For me, the number of things I should be doing asynchronously is a towering stack, so if there is something concrete to hack on, I feel like I'm more likely to be able to contribute. But also happy to do a short call to get things moving and then see where we get to.

@yarikoptic @jbpoline @remiadon Are any of you available for a portion of Nov 16 @ 8-10am EST?

[yarikoptic]

I can be.

[effigies]

I hope everybody's schedules haven't filled up. Sending out an invitation now...

[effigies]

If anybody has not received an invitation who would like to, please let me know. You can email this username @ gmail.