Criteria Matrix for Identifying and Prioritising Discovery datasets.

There’s an accompanying table for Adrian’s Timeline and Workplan post which I have attached here.

Criteria	Required	Desirable	Not required	Rationale
Has the institution adopted an explicit open licensing statement plus clear and reasonable T&C		X		For the purposes of the exemplar it may not be necessary to mandate open licensing as a necessary condition for inclusion in the aggregation. The exemplar itself is intended as a mean to demonstrate the value of opening data, and so could drive participation. Requiring fully open data may mean that important collections are left out and the user experience is diminished. However, if it is possible to develop an API on top of fully open metadata, this would be highly desirable, would provide a more compelling business case, and so support the longer-term sustainability of the API.
Have the adopted documented information standards	X			In addition to adopting common standards, the quality and comprehensiveness of the metadata records must be assessed. A baseline for quality assessment will also be developed.
Are they deploying persistent identifiers for unique identification of ‘things’?		X		This is highly desirable as a means to promote awareness and best practice, but we recognise that it may not be feasible to require this. Quality metadata sets with persistent IDs should be prioritised.
Are they establishing data relationships by reusing authoritative controlled vocabularies?		X		Highly desirable, and potentially should be required – could contribute directly to a more positive user experience if aligned well with the suggested use case requirements
Do they provide a clear mechanism for accessing APIs	X			Fundamental. If there is no API, then the ability of the exemplar to demonstrate the Discovery ecosystem ‘at work’ is not possible.
Are the APIs well documented?		X		Highly desirable, but Mimas developer could likely invest the effort required if the documentation is not there.
Do they adopt widely understood data formats?		X		Can be assessed on a case by case basis.
Is the API regularly updated and maintained?			X	Not necessary for this project, but must be promoted as a fundamental part of any sustainability plan. If this exemplar were to become a service, then this is critical.
Are service dependencies fully understood and addressed?			X	See above
Are they using their own APIs?			X	Hallmark of best practice in sustainability. But see above.
Is data being collected to measure use?			X	Not necessary for this project, but we might want to consider asking participating institutions to track API usage on their end