Computer based expert system Polychaeta

[Robin Wilson]

The recent little flurry of comment on computer based identification 
systems for polychaetes is based on a very vague initial query.  It is 
not clear to me if the original questioner wants to develop new software 
or if he/she wants to use existing software to ease polychaete 
identifications.  Whichever is the case, a good starting point is the 
DELTA (= DEscriptive Language for TAxonomy) home page at 
http://muse.bio.cornell.edu/delta/  Anyone contemplating development of 
new software would do well to investigate what the DELTA team has 
accomplished over many years.  At the very least, the DELTA standard data 
format  should be supported by any taxonomic database/intelligent 
system.  There are other taxonomic database and identification systems 
out there (try net searches on ALICE and PANDORA in the U.K, for 
example), but I think most support the DELTA format for data files.  

On the polychaete front, and in response to what Geoff Read wrote:
...
> based on the Delta method of descriptions for example. No polychaete
> datasets are publicly available or imminent yet though others will know
> more about that than I. ...

Well, I have some polychaete DELTA datasets in development.  They are not 
publicly available yet, but I like to think they are imminent!  At 
present I have about 12 polychaete families in various stages of 
completion, and more underway.  My project intends to complete 
these DELTA databases for soft-bottom benthic polychaetes from 
SE-Australia, with 2 goals:
1. to document taxa accurately from various environmental surveys
2. to facilitate accurate identifications in the future
The initial target audience is taxonomists and ecologists working with 
polychaetes in southern Australia, but maybe others will find it useful 
also, notably taxonomists doing global revisions who are looking for 
summary information about the southern Australian fauna for inclusion in 
revisions (I hope).  I have broached this subject on the list before, but 
I will say again that I hope the polychaete community out there will be 
able to make some critical and constructive comments as I am able to make 
databases available (family by family).  And I hasten to add to those 
who have already expressed an interest, please be patient, I haven't 
forgotten!

I also wanted to comment briefly on the very pertinent words of 
wisdom from Geoff and Helmut, to the effect that what is needed first are
good taxonomic revisions rather than new computer toys.  Although this is 
partly true, the two goals are complementary, they are NOT in conflict.  
I use DELTA because it vastly simplifies taxonomic work, it forces me to 
be more rigorous (not skipping characters, won't let me get away with 
taxa that cannot be separated!), and it makes the results of that work 
much more useful (I hope!).  The huge benefits to users should be the 
ability to have multiple starting points in an electronic key when making 
identifications, and having all important information handy (like 
character state glossaries etc).  And it is easy to make a paper (or 
html) version for those who want it.  I know I sound like an evangelist, 
but when you see a good thing ...

You can expect to hear from me again on this (in about a month), when I 
have finished incorporating images into the first databases and have 
talked to my museum's computer people about making files available 
for downloading.

bye

Robin

_____________________________________________________________________
Robin Wilson				rwilson at mov.vic.gov.au
Museum of Victoria			
71 Victoria Crescent			telephone 61-3 9284 0216
Abbotsford  Victoria			fax       61-3 9416 0475
Australia  3067
_____________________________________________________________________