Seek Taxon Notes Nov 2

Taxon Group Breakout November 2, 2004

note taker: dave thau

in attendence:

jim beach, nico franz, ed wiley, susan gauch, bob peet, dave vieglais, dave thau, aimee stewart, robert gales, xianhua lui, trevor patterson, jioanna t., gwen davis

jim - we have to clarify mammals of the world - what is it that people want to do with it

jesse - we need data which extends over time so that we get different data marked up with concepts so we can demonstrate what happens when you don't go to concepts

nico - optimally it would have information which would help drive policy information

ed wiley - you'd want well worked out species level taxononomy, and easy access to synonyms and alternate names

jim - that scoping leads toward a work plan that is narrow and deep, unless there are pre-canned data like german moss or pyle's fish data

ed - look at the energized communities - the vertebrates, the mammal group is a good group

jim - birds?

ed - maybe, as town, the herps might be good too - fishes has eschmeyer's synonymy list and they're fairly energized

nico - what about the catfish - 200 taxonomists who are supposed to do the catfish for good

ed - yeah, the catfish is a great idea - you have eschmeyer as a baseline and lots of people, larry page is the lead on it

susan - what about data, do they have the data?

ed - there's occurance data in fishnet, there's also minnows of the world

bob - are any of these using concepts, or just names and synonyms?

jim - there are 2700 species of catfish and 1500 left to be described (reading off the larry page catfish of the world site)

jim - there are four planetary biodiversity inventories: solanum, plant bugs, catfish, slime molds

ed - ecological niche modelling of aquatic species has to happen with terrestrial surrogates, so maybe it's not good to start with something that uses surrogate data

dave v - that's why mammals were chosen - there's lots of environmental data

ed - and lots of literature

jim - well catfish group has a huge base of people working with them, compared to the mammals of the world - who's working on it

bob - catfish are too obscure, but we could invite the catfish community to use our tools

susan - we should use mammals as the near term goal if that's the project's goal

nico - we could invite all the pbi groups

jim - but for mammals we'd need to get the concepts ourself

jessie - if we use a checklist or another database we just have to figure out how we want to treat that data - the datasource might be a mishmash which would be hard to sort out without help

jim - the mammals of the world people may all be dead, who are we going to call for mammal data?

ed - so the concept of a binominal is those things included in it

jessie - that's one person's definition - different people define concepts differently, some people use the taxa, some people use character circumscription, there are other ways

ed - do you want to give equal currency to all of these, or just those that makes sense? For binominals the only thing that makes sense is pointing to the specimens

jessie - you're the first taxonomist we've met who's said that - that's the prometheus model

nico - we're not trying to limit practice in any way - if the transfer schema enforced one view it would fail to transfer data in other forms

jim - the catfish folks have a full bibliography

susan - do we have ecologists who are dealing with catfish

ed - there's lots of fishery information - lots of people eat catfish

susan - cod would be interesting, they're all vanishing

ed - all deep sea marine information if georeferenced from ship information

jim - first, we need to find a source of data, then we need to have a nice demo using those data

jim - in knb there are 67 data sets on mammals

jessie - the mammal data can come from itis, natureserve, mammals of the world

..... general hububary and impossible to separate out conversations... I count 3 going... one about using mammal data, one about what's the issue, and one about too far for me to hear ....

jessie - we should ask aimee and robert to start loading in that data

susan - and that serves what seek needs

jim - who's going to do the mappings between concepts

susan - we can put together the visualization tools

jessie - and some might refer to others, and we can fake some stuff

susan - yes, and then we're serving the seek needs, and most of the techniques we're using will apply

jessie - we have mammals of the world and mammals of texas, we'll need somebody to make the relationships

nico - but the catfish people have all these people

jessie - no point talking about it amongst ourselves, we'd have to convince the whole seek group

bob - it might be practical to use the trees of the united states, we have 800 taxa, the usda list, 1/2 in the flora of north america, one other checklist

dave v - that would be interesting, there's lots of ecological data available for trees

ed - you can use american fishery society checklist - it's updated every 10 years, you can map between the checklist, some parts change dramatically, it's complete for 4 or 5 iterations

susan - so the seek demo is ecological while the one we want to do is taxonomic, so they're quite different use cases

ed - one interesting thing about the american fishery society is that all fish ecologists have to use it, so it's a good link to the data

nico - that's what we want for the mammal case but don't have

susan - that seems perfect

ed - the checklish uses common names, there's only one common name per binomial

jim - i'm for the mammal one to demonstrate the computational aspects, then we have the catfish folks to help us building tools for concept management and authoring

susan - we need to map to the users of the data

jim - the third thing is using collections as noisy playgrounds to stimulate the assembly of concept maps by the collections community

susan - the fourth is the taxonomist using the data sets

nico - the fourth is bob's hickories and my weavils

jim - that's in with the collections category

jim - it might be valuable to talk about our various things, are we ready to launch into these new things, is populating the database still #1 or are we still defining services

bob - i'm not sure.. we should walk through the use cases

aimee - they're all addressed by the service that we have

aimee - we need to update the object model with the new relationships, that shouldn't take too long as long as we have time to work on it, we need more data to test the model

nico - if we want to populate the database with collections data... can we do that now? should we use the tcs to transfer collection data

jessie - depends on if you treat the data as specimen-focused or concept-focused. If you have the specimens as specimens of concepts, but we shouldn't treat the specimens as concept definitions

nico - can you say that bob peat accepts or rejects another concept - a concept status assignment

jessie - you might be able to have a relationship with only a from and no to

bob - so do we agree that getting data in there is our #1 priorty?

jessie - well there's this linnaen core thing - on sunday there was a terrible meeting - a massive wiki has developed since new zealand, they're not considering other people's point of view

dave t - where do data entry tools fit in with getting data into the database

jim - good point, what tools do we have?

jim - what's our second priority?

dave t - there's whatever seek integration tasks we have

susan - there's the whole data set registration - even if you know it's concept 42 how do you tie that to the EML? how do you find it?

jim - who's responsible for that software - is it morpho?

susan - it's extending EML

jim - I'm concerned with finding a user base. Worst case scenario is that we have some cool demos. If we don't have tools which people are using, what will happen with our work?

nico - either we start going to meetings at museums or we latch onto projects like the pbi projects, or we integrate with programs already being used with a large user base

jim - so what else, data entry, integration

susan - we're working on populating the database from a TCS and jessie's going to work on putting the mammal data into TCS

dave v - there are tools to map sql databases to xml documents and advertise those conversions as an easy way to contribute data

dave t - we need feedback mechanisms to encourage people to contribute concepts susan - we need to create ways for people to contribute concepts in the first place

dave v - now that we have the exchange schema this shouldn't be as hard as it has been - there are tools

xianhua - those are tool based, I'm not sure they'll work

nico - what about other software programs that have been established - lucid, macclade, mesquite.. things which have long records, we should try to integrate with them

susan - I don't know if we should be creating a tool for people to author concepts - people should use whatever tool they're familiar with, we should just be a place where people can share the concepts they've developed

bob - people should be able to use our service to author relationships between concepts

nico - we also need to publish more

.....

conclusions

we are going to do mammals of north america by taking various data sets and turning them into concepts for demo purposes

then we're also going to look for groups which we might be able to get more rich data, and more importantly help

• trees of north america
• fish in general
• catfish via the pbi group
• any other pbi

there's also the tack of using collections to drive concept creation

a major priority is getting information into the database

another major priority is integrating with the rest of seek