August 13, 2009
Chris and Susan
Regarding (biodiversity) taxonomy as a kind of controlled vocabulary, Susan described the tension involved:
Rather than looking at SMaSCH, we should probably be looking more closely at how Specify and Arctos have decided to manage this tension. We do need to check in with John Wieczorek (MVZ) too since he is a local expert in this field. Susan also shared some documents as background -- one from TDWG's Taxon Concept Schema (TCS) project and several from a thesaurus project that IST conducted with the BNHM consortium (though especially with MVZ).
Most collection management databases' handling of taxonomy is intentionally simplified and incomplete; their structures are inadequate for actually managing taxonomy. The issue is how much taxonomic information it is necessary to maintain for collection management and other collection needs. There is a tradeoff between having this data and the need to maintain it.
Also, can the taxonomy be used throughout the collection management system in the ways expected by natural scientists. So for example, the taxa identification of a specimen is usually expected to become the name of the object in the collection management system.
We talked about some potential use cases (e.g., referencing an externally hosted taxonomic service vs. storing and managing the information locally; local terms; classification changes and history; managing identifications separately from taxonomy; taxonomic ordering; multiple taxonomies). Susan emphasized that the requirements are important. Use cases can also be gathered from some of the thesaurus work that Debbie and Susan did with BNHM/MVZ.
Chris will list some more specific use cases for taxonomy on the use case page for this project, and assign them to Lam and Susan, and talk with Patrick S about the approach.
Toward the end of the meeting, Susan talked about a meeting she had earlier in the day with Jason Miller at the College for Environmental Design. Their collection management system (in XDB) is seriously limiting them right now, forcing them to enter information about their visual resource collection using the old slide library metaphor (even though they do not do many of the library functions any longer). They are considering moving to FileMaker. We need to develop some use cases for visual resource collections (e.g., works vs. images; OAI-PMH harvesting for CDL).