Date: August 10, 2009
Chris and Lam
Are scientific taxonomies qualitatively different from controlled vocabularies in other domains? The Getty AAT is similarly hierarchical. What are the use cases around taxonomies that will help us know what is different? According to Lam, one issue is that disagreements over taxonomies are prevalent. MVZ with the other Arctos institutions has decided to try and resolve these issues for their common deployment in Alaska. However, MVZ has been wanting a common thesaurus solution for a long time. Informatics Services had done some significant requirements gathering and design work on this, and we need to locate those materials. Lam: If we are going to restart any of the taxonomy work, we should talk to MVZ and other Arctos participants. MVZ was disappointed when nothing came out of the earlier work. In SMaSCH, specimens are linked directly to a taxonomic entry by taxon ID. A record of history (changes) is kept. However, Arctos and other systems model this relationship differently -- adding a many-to-many table between object and the taxonomy tables and allowing for formulas and relationships to help make the state of the identification more accurate and informative. For example, in SMaSCH, when a plant specimen is a hybrid, a new entry in the taxonomy tables is created for 'SpeciesA x SpeciesB'. In Arctos, the identification table employs a formula to identify the hybrid between two acknowledged species (for example).
Committees have to be in place before an accession can be entered. Committees can be one or multiple people, though for historical reasons, SMaSCH only records the committee chair for groups. Committees have different functions: collecting, annotations, identification, publication. Arctos uses the concept of agents, and an agent can be a group of individuals. SAGE (Botanical Garden) also uses the entity agent, rather than committee. It's not clear whether there is something qualitatively different here.
Here's some additional information from Lam:
Committees in SMaSCH are comparable to Agents in SAGE and Arctos. There are committee_types for SMaSCH and agent_types for SAGE and Arctos. In all three databases, there is also a Person table to keep track of committee members in SMASCH and agent_type = "person" for SAGE and Arctos. In SMaSCH, when a Committee is made up of a number of people, each person should have an entry in the person table. I meant to say that because of data entry limitations due to project funding, mostly only the committee chair has been entered in the person table.