Interdisciplinary Models: Difference between revisions

← Older edit

Interdisciplinary Models (view source)

Revision as of 16:25, 13 November 2022

41 bytes added , 16:25, 13 November 2022

no edit summary

Ml-evs

23

edits

Revision as of 16:04, 13 November 2022 (view source) Ml-evs (talk \| contribs) No edit summary Tag: Visual edit ← Older edit		Latest revision as of 16:25, 13 November 2022 (view source) Ml-evs (talk \| contribs) No edit summary Tag: Visual edit
Line 203:		Line 203:
	-Taking interdisciplinary in a very narrow sense, an example to bridge gaps between existing materials databases is OPTIMADE. We spent 5 years devising a common API format between 15 or so existing crystal structure databases, primarily of competitors, and with varying uptake. Should now be possible for anyone, to create a database and join the federation, e.g., a new project just tripeld the number of structures available. Each database has different slants even if the quanta of data shared is common. Unclear if this is socially sustainable. Hinged on developing a lightweight federation with "shoddy" infrastructure on top of a relatively API format.		-Taking interdisciplinary in a very narrow sense, an example to bridge gaps between existing materials databases is OPTIMADE. We spent 5 years devising a common API format between 15 or so existing crystal structure databases, primarily of competitors, and with varying uptake. Should now be possible for anyone, to create a database and join the federation, e.g., a new project just tripeld the number of structures available. Each database has different slants even if the quanta of data shared is common. Unclear if this is socially sustainable. Hinged on developing a lightweight federation with "shoddy" infrastructure on top of a relatively API format.

	- Specific example: as this data is machine-actionable, new software projects can now query all known materials, potentially significantly accelerating materials discovery. New databases can trivially join the federation, and previously "dead" data can be brought back to life by adding an API on top of it.		- Specific example: as this data is machine-actionable, new software projects can now query all known materials, potentially significantly accelerating materials discovery. New databases can trivially join the federation, and previously "dead" data can be brought back to life by adding an API on top of it. Automated labs can exchange information.