|
This is version 17.
It is not the current version, and thus it cannot be edited.
[Back to current version]
[Restore this version]
Discussion session intended to address the integraiton of taxonomic tools with other components of the SEEK architecture.
Need to change EML to handle taxon concepts
- GCE is only institution with taxonomic coverage info in EML docs
- Morpho needs to handle taxonomic markup
nice use case for Kepler - interactive decision making, expand concepts, relationships
Matt: good to have loop version that will resolve a list of taxa
Bertram: need to talk to Ilkay about the wsdl harvester and how it can work with TOS
data integration
- map b/w 2 ecological datasets - must make value judgements on match b/w taxa within
- data integration in Kepler - provide the mapping so scientists can match datasets
- how does that relate to addRelationship
- must keep those 2 separate - use taxon output vs add content to SEEK cache
- do we want to store some of these decisions made during data integration?
- may want to share synthesis decisions by a working group of scientists so that others can use or discard those opinions
BEAM: what are the taxa that have changed a lot and would be useful for use case
- interesting ecological problem
- abundance/ whatever data with spatial, temporal range
- mammals is nice small group, name mapping issue is clear
- 2 issues
- name expansion
- misspelling problems
Use case
- NA mammal niche modeling (can get data with DiGIR query)
- pick top 5 taxa to model
- run mammal modeling with and without taxon service/ concept matching
- need a mammals expert to evaluate results
- western grasses concept data?
- big societal impact
- Vegbank should be getting data soon?
Engage taxonomists/data creators
- encourage addition of concepts/relationships
- catfish project?
- increase use case understanding
- Kepler functions here?
- collaborate with well organized group or NSF funded project that is working on it now?
General
- Get real concept data relevant to larger SEEK context
- BEAM group has identified NA mammals, let's get subset of that - 5 taxa
- Need links between APIs and Ecogrid
- Reindex existing datasets - ecological and specimen records, modify EML records
- Query with name for concepts then datasets fro concepts
Specific Assignments
- Mark Schildhauer: contact Reichman, Willig, and Stevens about mammal concept information
- Deana Pennington: better describe the requirements and intention of the mammal niche modeling case study, with assistance of Nico and Dave
- Populate concepts for those taxa
- and identify ecological datasets, markup in EML with Taxonomic Concepts
- Nico will help from Taxonomic perspective
- Dave V. will help from Ecogrid perspective
- Identify mechanism/plan for any given scientific study
- Formulate plan
- Walk the experts through the concept creation steps
- Create TCS doc for import into SEEK Taxon cache (Nico, Jessie)
- historical concepts/relationships
- integrate functionality of TOS with Ecogrid
- Incorporate TOS into current data queries
- term expansion in client for now
- smart searching - currently divided into 2 searches, need to integrate SMS and Taxonn searching, Shawn thinks - bring it in locally
- Aggregation node vs individual providers, regularly scheduled updates
- How do we annotate datasets - metadata generation
- Modify taxonomic part of EML
- Can use additionalMetadata with no changes to EML tools for experimentation
- Could use simple name/rank/publication composite key with/without LSID
- Using only LSID might be best long-term solution
- Maybe use controlled vocabulary (Smithsonian?)
- Modify Morpho tool which is pulling taxonomic names/info from datasets to use TOS for concept matching
- Ecological datasets vs specimen datasets
- How do we interface with ecological dataset creators that don't use Morpho?
- How do we use the rich TOS result to query on DiGIR resources with sparsely populated taxonomic information?
- Annotate Darwin Core records with proxy - LSIDs for specimen records and Concept LSID
- Map between TCS, DarwinCore, EML (Jessie)
|
