[seek-taxon] ConceptMapper for SEEK Taxon R.I.P.

[Chandramouli, Aravind]

We are working on an application that will take TOS as input, display the relationships between species in the form of a tree, allow users to edit the data and finally, save the output again as TOS.  

An example of this tree viewer is at http://www.ittc.ku.edu/~aravindc/test.html

Let me know if there is a better starting point than this in terms of the UI or places in the application development process that we can colloborate on so that we do not duplicate what has been done before.

Aravind.

________________________________

From: seek-taxon-bounces at ecoinformatics.org on behalf of Beach, James H
Sent: Tue 10/31/2006 11:21 AM
To: SEEK Taxon
Subject: [seek-taxon] ConceptMapper for SEEK Taxon R.I.P.



I regret to report that ConceptMapper software is unusable for any SEEK Taxon purposes related to TCS or the TOS.  We suspected this earlier but Aimee has pretty much put the nail in the coffin, here on Halloween. The good news is that it is usable for Bob's concept mapping work as evidenced by his recent success with (What was it?) 100,000? mapped plant concepts.
 
We are back to square one about a UI application to enable end users or people with a TCS-level of awareness of concepts, to enter data into TOS or to author TCS docs.  Our bat taxonomist, Heather York, will be using a spreadsheet.  Despite having data for bats for MSW 93 and 05 in the TOS, we have no application that supports that mapping between them, we'll need to find another way.  Aimee and I will be thinking about it.
 
Jim B.
_____________________________
James H. Beach
Biodiversity Research Center
University of Kansas
1345 Jayhawk Boulevard
Lawrence, KS 66045, USA
T 785 864-4645, F 785 864-5335

 

________________________________

From: Stewart, Aimee Marian 
Sent: Tuesday, October 31, 2006 9:54 AM
To: Beach, James H
Subject: ConceptMapper for Bat project


Hi Jim,

Last week I started a How-To document for the bat project using the most recent version of ConceptMapper (18 Oct 2006).  I ran into some severe problems with the application's interpretation of a simple (7 concept, 11 relationships) test dataset.  There are some other issues with ConceptMapper that make it difficult to use though they do not necessarily corrupt the data.  These may be issues that Laura addressed, but have not yet been incorporated into the application.

In short, Taxonomic Concepts in ConceptMapper do not match any version of TCS.  The differences are not merely a simplification of TCS structure, they change or ignore some basic simple versions of objects.  I don't think that this application in its present form will work for creating and/or relating concept data.  Here are some of the problems that led to that conclusion:

Import Data:

   1. Data is imported inconsistently
         1. AccordingTo's on Concepts are imported as Publications, called References in some places, AccordingTo in other places
         2. Those Publication/AccordingTo/Reference items are imported differently depending on items that occur near them
         3. Publications are not imported
         4. Of the 11 External Relationships (RelationshipAssertions), only 2 were imported
         5. One of the Internal relationships was imported as an External relationship
         6. AccordingTos, Specimens, Institutions, Collections, and SpecimenItems all have a required field that, when imported, is overwritten by the application
   2. Taxon Names - Only the simple string is imported, all other data is ignored/lost (Publication, CanonicalName, CanonicalAuthorship, Rank, Typification, etc)


Export Data:

   1. Obviously, data ignored or interpreted incorrectly during import is lost or exported incorrectly
   2. Export file is not well-formed (parseable) for included dataset
   3. Export file for small test dataset (7 concepts) is well-formed, but data is incorrect
   4. Required fields for some objects are generated by the application - replacing those fields that were originally in the imported data


Add Data:

   1. Can only add a single Specimen - error on additional specimens
   2. AccordingTos, Specimens, Institutions, Collections, and SpecimenItems all have a required field that cannot be added and does not appear in the application.  The required fields are generated by the application - inappropriate for a simple string
   3. External relationships (RelationshipAssertions):
         1. Zero of the hierarchical realtionships can be added
         2. Only 5 of the 18 non-hierarchical relationship types can be added
   4. Internal relationships (within a concept):
         1. Only 1 of the 6 hierarchical internal relationships can be added
         2. Zero of the non-hierarchical relationships can be added
   5. "Synonyms" can be added to a concept, but they are just a single string - this is not a part of TCS

Aimee