Background and Description

Currently GRSciColl is not structured to allow discovery of collections or individual specimens that could be critical for researchers. Support for describing a registered collection in GRSciColl is currently limited to single fields that capture broad statements of taxonomic coverage, geographic coverage and e.g. important collectors. With this enhancement, we intend to support richer, structured descriptions, such as the ability to upload an inventory of the species represented or e.g. a table representing the “species, sex, object count”. This is intended to both facilitate discovery of collections (“who holds preserved material of a specific species”) and to allow a more accurate description of a collections holding, whether digitized or not.

We anticipate supporting multiple descriptors for a collection, with a descriptor containing a title, textual explanation and a table of data edited inline or uploaded as e.g. a spreadsheet.

A simple example is illustrated.

We envisage the system would be flexible enough to accommodate differing levels of detail from simple lists, to detailed representations of the collection which we anticipate exist for some collections. By supporting multiple descriptors for a collection, the system would support different aspects of the collection to be documented at different detail levels. The contents will be indexed to help discoverability of collections. For example, we aim to index scientific names using the GBIF Backbone taxonomy in order to facilitate collection discovery by taxa.

It should be noted that this approach would mean that one may not be able to aggregate counts across descriptors as objects may be included in multiple places and thus double counted. However, the primary focus of this is to support the needs of taxonomists looking to discover collections of interest or where individual specimens may reside.

The Index Herbariorum has descriptor tables (example) which would be automatically incorporated during the synchronization process.

An API that exposes the descriptors as a Latimer Core document representing the collection will be available (likely in JSON format).

Notes based on feedback

• This item needs to be scoped before implementation. We will keep in mind the following comments:
  • We still don’t know how/if the search based on those descriptors could be integrated with the occurrence searches.
  • The breath of information that could possibly be shared this way might make the data confusing. We will need to define which descriptors/standards can supported.
  • During implementation, we will keep in mind that some institutions might need to upload multiple description tables for multiple collections.
  • The descriptors uploaded would have to follow standards (likely Darwin Core terms)
• We won’t be able to infer complex metrics across tables, but the information will be available to users.
• The collections without such collection descriptors will still be findable with the current search filters.