Stakeholders
Data publisher: to describe they own catalogues/repositories
Data consumer: to find catalogues/repositories containing relevant data
Requirement / Use Case
The publisher wants to describe their catalogue at a high level, e.g.: experimental vs computational data, for pure substances vs mixtures, etc. High level properties on the way the data was obtained and the material class type.
This enables the consumer to filter at the level of data sources (catalogues) rather than at the dataset level, speeding up/narrowing the search.
Additional context
For example, this would consolidate the functionalities of services such as PSDI cross-data search, where multiple data sources are simultaneously searched.
Relevant background work by Dr. Aileen Day (metadata lead at PSDI): Comparison of subject classification systems for resources in Physical Sciences Data Infrastructure (PSDI) [DOI]. This report points to and compares a number of subject classifications.
Some ideas from the DOME 4.0 project could be re-used too, see Deliverable D3.2 - “Ecosystem information model ontology” [link]. E.g., here concepts such as standard_identifier (InChi, etc) and action (e.g., VIEW, TRANSFORM, etc) were used to characterize resources and their APIs, so to enable filtering at the catalogue rather than dataset level.
Question: how to tag mixed catalogues (e.g., both experimental and computational), use the union of properties that appear or only the majority property to describe the catalogue?
Stakeholders
Data publisher: to describe they own catalogues/repositories
Data consumer: to find catalogues/repositories containing relevant data
Requirement / Use Case
The publisher wants to describe their catalogue at a high level, e.g.: experimental vs computational data, for pure substances vs mixtures, etc. High level properties on the way the data was obtained and the material class type.
This enables the consumer to filter at the level of data sources (catalogues) rather than at the dataset level, speeding up/narrowing the search.
Additional context
For example, this would consolidate the functionalities of services such as PSDI cross-data search, where multiple data sources are simultaneously searched.
Relevant background work by Dr. Aileen Day (metadata lead at PSDI): Comparison of subject classification systems for resources in Physical Sciences Data Infrastructure (PSDI) [DOI]. This report points to and compares a number of subject classifications.
Some ideas from the DOME 4.0 project could be re-used too, see Deliverable D3.2 - “Ecosystem information model ontology” [link]. E.g., here concepts such as standard_identifier (InChi, etc) and action (e.g., VIEW, TRANSFORM, etc) were used to characterize resources and their APIs, so to enable filtering at the catalogue rather than dataset level.
Question: how to tag mixed catalogues (e.g., both experimental and computational), use the union of properties that appear or only the majority property to describe the catalogue?