Nomina IX in the Astor Crown Plaza Hotel in New Orleans, on the afternoon of October 21 and on October 22nd - at the end of the 2011 "TDWG meeting". The goal of the meeting was to identify priorities.
Attendees
Frank Bisby (Species2000)
Stan Blum (GN US)
Nico Cellinese (ADBC)
Jerry Cooper (Landare New Zealand)
Jim Croft (ALA)
Gregoriev, Teodor (PenSoft)
Stinger Guala (ITIS)
Yde de Jong (PESI)
Paul Kirk (Index Fungorum)
Jerry Lu (iPlant)
James Macklin (gov.ca)
Narim Matasci (iPlant)
Chuck Miller (MOBOT, TROPICOS)
Dick Moe (Index Nominum Algarum)
Paul Morris (FileredPush)
Dmitry Mozzherin (GN US)
Nicky Nicholson (Kew)
David Patterson (GN US)
Lyubo Penev (Pensoft)
Rich Pyle (GN US)
Tony Rees (IRMNG)
Kevin Richards (LandCare New Zealand)
Greg Whitbread (ALA)
Rob Whitton (GN US)
FRIDAY
Presentations
Rich Pyle: Background to Global Names Architecture
David Patterson: Overview of Global Names (US), the MBL role, and our purpose
Rich Pyle: ZooBank
Stan Blum: Architectural issues
Q&A on GNA and GN US
Discussion
GN must achieve a 'Marquee success'
Success is 'acceptance', our goal is to see GN being used in preference to other options
We cannot achieve across a broad spectrum, so we must define the target audiences
A study is needed to assess practices of relevant community
SATURDAY BREAKOUT GROUPS
Addressed 6 issues. Three (Audiences and criteria for success, Delightful services, and Taxonomic priorities) sought to limit GN options to achievable goals. Two themes, 'Nomenclatural registration' and 'Content replication and annotation', target areas of high current relevance, and the last (Core Services and the GNUB data model) sought to discriminate foundation elements of GNUB.
Audiences and their criteria for success
There are many classes of users:
with infrastructural needs
with software needs
data needs (such as EOL)
needing value-adding services - I have a name, give me back ... (all names, the ID according to, the basionym, the paper in which it was described, the preferred name according to X) Examples, see 'Registration'
Identification of the most appropriate /urgent / valuable tasks will require careful market analysis and the development of prioritization criteria. By what means?
Nomenclatural registration - joining the dots
Prokaryotes nomenclature is well organized
Fungal registration is voluntary, more than 80% successful, and will emerge as a code (ICNAFP) requirement. As there are three likely initial registration locations (Index Fungorum, Mycobank, and a Chinese Registration Facility under development), this is a good use case for synchronization (replication)
Plants will probably follow suit within 6 years
Zoology is expected to have multiple registration points, all informing ZooBank
Index Fungorum and ZooBank will work towards a common data model and common database (exploring merger of data in March 2012, Nomina X)
Delightful services
Think why; thinking who before thinking how
Discussion of web browser plugins that offer name-linking services led to the opinion that we should see ourselves as being middle-ware; with emphasis on API services. This would allow third parties to build services for end users. This perspective can lead to the conclusion that name-linking software plug-ins for web browsers is not in scope. Agreed or not agreed?
GN space as a black box that provides "I give you a name "services much along the form of: "I give you a name, and you give me back your 'menu item 5' (where Menu item 5 might be a name resolution service that returns the preferred lexical variant according to a particular source; as opposed to menu item 8 which might be a taxonomic resolution service that returns the preferred lexical variant of the correct name for the taxon according to a particular authority; other services might return all variant names, the ID according to, the basionym, the paper in which it was described, etc.)
Use case - GN working with iPlant to improve their services to their clients, and through the interaction to generalize the service.
Taxonomically intelligent alert services, see Annotation
Track taxonomic name usage over time (see Ryan Schenk's synynymy.no.de application)
Expansion into NLP, co-occurrence of items in BHL, such as known pathogens with plant names, co-occurrence of taxa in the same geographical region
Servicing specimen-related projects (BiSciCol, iDigBio)
See 'Audiences and Criteria for Success'.
Content Replication and Annotation
Replication covers situations where an entry in one part of a (distributed) system MUST be entered into remote systems. Fungal registration is a good example as there will be multiple registration locations (Index Fungorum, Mycobank, and a Chinese Registration Facility under development); IPNI may have a similar need for synchrony among three data centers. Zoology is expected to have many registration points and will require an effective replication infrastructure.
Annotation addresses situations in which changes in one system are communicated to remote systems but leave the implementation as an option. A good example is the discovery of inconsistencies within or between databases.
Both require the adoption of a global system of IDs that can be used as the reference point for annotations. An ID minting service for names will have to be provided (by GNI). This needs to include a same-as service to allow persistence of existing ID systems
The annotation infrastructure can be adopted to provide taxonomically Intelligent alerts that broadcast the occurrence of a change in taxa of interest (single taxa, clades, or members of lists)
Core services and GNUB data model
There are many users, many use cases, many options. We should think of the GN space a little like a DNS system (= GNS), hierarchically organized, perhaps with a single master copy (of names to index all other components).
GNS (DNS-like) services to satisfy the need for mapping different databases
Reconciliation services are much sought-after. Consider a use of the IUCN ratings to annotate other taxa. Such annotations would require a reconciliation step. In this discussion, it was noted that in acting like a GNS, GN can be used to correct a user's misunderstanding that IUCN or NCBI contains ALL species. That could be addressed with a service that 'fills in' missing taxa
Replication (see above)
Taxonomic priorities
It is impossible to provide rich services across all taxa. While many areas of biodiversity management require life-wide services, BHL being a good example, it will be necessary to start with selected taxonomic areas for some services - such as names resolution services that incorporate all types of synonyms. Such a service requires that the names architecture accesses and integrates a considerable body of taxon-specific knowledge. As we work out the 'why', we may wish to chose from the following for which useful data are available in an accessible form include:
Fungi offer a potential use case for synchronizing activities across multiple registration locations (Index Fungorum, Mycobank, and a Chinese Registration Facility under development)
Solanaceae (Sandy Knapp / Rich Pyle)
Fish generally (for animals) and a more targeted group would be Catfish (ties in with an NSF PBI project and to iDigBio)
Trichomycetes is a well informed exemplar for fungi (Index Fungorum, Species Fungorum)
Conifers (Conifers of the world, http://dps.plants.ox.ac.uk/bol/conifers Aljos (A.Farjon@kew.org.) is a CoL GSD and subject of an ATOL project
Solanaceae (Sandy Knapp / Rich Pyle)
Orthoptera Species Files system (David Eades)
Action items
Collaborate with the iPlant names resolution services.
Seek specialist input to formally explore the IPR issues to inform data-use agreements with content provides as appropriate
Add taxonomic context to names in GNI
Seek specialist input (? Hillmar Lapp, NESCENT or group at University Illinois, Champaign) to acquire a better understanding data cultures within taxonomy, to ensure that emerging GN structure will satisfy its clientele
Reshape description of GN to distinguish and emphasize 'write relationships', 'read relationships' and 'response relationships'
Service specimen-related projects (BiSciCol, iDigBio)
Set up a meeting about names and the semantic web (RDF / ontologies)
GN to attend i4Life/CoL meeting in Paris to identify areas for co-operation.
