From - Mon Jan 27 09:07:34 2003 X-Mozilla-Status: 0001 X-Mozilla-Status2: 00800000 Message-ID: <3E352105.2090105@cria.org.br> Date: Mon, 27 Jan 2003 09:07:33 -0300 From: Ricardo Scachetti Pereira Organization: Centro de =?ISO-8859-1?Q?Refer=EAncia_em_Informa=E7=E3o?= =?ISO-8859-1?Q?_Ambiental_=28CRIA=29?= User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0 X-Accept-Language: en-us, en MIME-Version: 1.0 To: "Stewart, Aimee Marian" CC: "Beach, James H" , "Vorontsov, Gregory Yakovlevich" , "Downie, Scott Russell" , reed.beaman@yale.edu Subject: Re: Value added products to collection as webservices References: Content-Type: multipart/alternative; boundary="------------010809040406030205070706" --------------010809040406030205070706 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit It is good that you liked, Aimee. See answers below: Stewart, Aimee Marian wrote: > Hi Ricardo, > > Great suggestions! I think that this is an excellent way to increase > our value to data providers. I will think about some more services > that can be implemented. > > By the way, 2 services are available > at http:\\beta.lifemapper.org\Services\QueryByLocation > and http:\\beta.lifemapper.org\Services\QueryTaxa . Number 6 on your > spreadsheet could be easily be put together from methods in these > services. You are right! There is also another implementation of this service that creates many models, jackknifing the input datapoints (submitted paper by Town and I), that might be interesting, too. Reed is involved on this one, too. > > Number 7 would be easy once we gather layers for areas outside the > US. I will get to that once we have successfully moved to production. Yeah! The data repositories that those kind of services rely on can be build in phases, as more data is added on. For the parts of the world that we don't have data, we output a diagnostic message: no data available. Those datasets can also be used to normalize those same data fields. > > We can create the initial services, then create some applications > using the services to implement commiting the new data back to > provider databases (interactively, one record at a time or groups at a > time for normalization, in Specify perhaps). To get some interest > quickly, I think we need to allow providers a way to improve their own > data without on-site programmers. Right. > > Report cards could include geostatistical analysis of entire > collections, genus, species, subspecies. I've been wanting to explore > that ESRI extension! Yeah! I've been thinking about measuring correlation between a given set of points (per species, genus, museum, collector, etc) and features related to infrastructure (roads, rivers), and also with county centroids. That would give a basis for comparing sampling between datasets. Also, we could measure correlation of datapoints and ecological dimensions, but that is for other purposes. In this line, I was also thinking about some kind of density measures (spatial and temporal, such as, points per square mile, or per year), and also measures of clustering. > > Perhaps a short questionaire for existing contributors could determine > whether there is any interest in these services, or if they have some > ideas we haven't yet come up with. The initial Fishnet (Ed) reaction > was very positive! Did you show that to Ed? Cool!!! > > Aimee > > > > -----Original Message----- > From: Ricardo Scachetti Pereira [mailto:ricardo@cria.org.br] > Sent: Friday, January 24, 2003 11:47 AM > To: Beach, James H; voron999@UKANS.EDU; Stewart, Aimee Marian; Downie, > Scott Russell; reed.beaman@yale.edu > Subject: Value added products to collection as webservices > > Hey, guys. > How have you been? > > Let me bring something up, so we can discuss it and think > about it a little, if you want. > I've been thinking about value added products to collections > and came up with some ideas. > Among the categories of value added products, I could list those: > 1) Error checking and correction tools, such as: normalization > of field values, checking for invalid characters, checking > compliance of date or coordinate formats, checking mispellings and > typos, cross-reference of textual location descriptions (country, > state, county, etc) with georeferences, etc; Reed's georeferencing > tool could be put in this category, if you consider that putting a > georeference in a record that does not contain one means > correcting it, i.e., filling in the blanks on georeference field. > 2) Visualization tools: Showing data on maps (like GIS tools > and WMS services); > 3) Statistics and other analysis tools: Measuring clustering > of points, correlation between data points and infrastructure, > sampling densities, etc. > As far as I know, the tools in category #3 don't lend > themselves well to a webservice implementation. They could be > implemented as services, but they are not very important to stand > up for themselves. > Tools in #2 are currently being well covered by Aimee's WMS. > So I tried to think about possible #1 tools that would become > good webservices. The result of that is summarized in the Excel > spreadsheet attached. > I'm still not pretty sure what would be the best architecture > to implement them. > A possibility would be implement each of them as one object, > or methods within an object. > Each object would have a SOAP and/or XML/HTTP wrapper so they > become webservices. > This way, local server components can use the objects > directly. Remote components and other users can call the services > programmatically using the webservices. Some can be implemented in > Kansas, other at CRIA, or we could have copies in both places. > Pretty flexible. > Last but not least, there are the report cards to the > collections, that could use some of those services and some other > statistics from Lifemapper. > Questions: > 1) What do you think of the general idea and the proposed > architecture? > 2) What do you think of the services listed? Can you add some > more to the list? > 3) Do you guys have more ideas for the report cards? What > should be in there? > 4) Market analysis: would any one out there (collections > community) in need to use such services? > > That is about it. > Cheers, > > Ricardo > > > > Attached is a set of 10 possible value added products that > >-- >Ricardo Scachetti Pereira >Gerente de Pesquisa >Centro de Referência em Informação Ambiental - CRIA >http://www.cria.org.br/ > > > > -- > This message has been scanned for viruses and > dangerous content and is believed to be clean. > > > -- > This message has been scanned for viruses and > dangerous content and is believed to be clean. -- Ricardo Scachetti Pereira Gerente de Pesquisa Centro de Referência em Informação Ambiental - CRIA http://www.cria.org.br/ --------------010809040406030205070706 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit     It is good that you liked, Aimee.
    See answers below:

Stewart, Aimee Marian wrote:
Hi Ricardo,
 
Great suggestions!  I think that this is an excellent way to increase our value to data providers.   I will think about some more services that can be implemented.  
 
By the way, 2 services are available at http:\\beta.lifemapper.org\Services\QueryByLocation and  http:\\beta.lifemapper.org\Services\QueryTaxa .  Number 6 on your spreadsheet could be easily be put together from methods in these services. 
You are right!
There is also another implementation of this service that creates many models, jackknifing the input datapoints (submitted paper by Town and I), that might be interesting, too. Reed is involved on this one, too.
 
Number 7 would be easy once we gather layers for areas outside the US.  I will get to that once we have successfully moved to production.
Yeah! The data repositories that those kind of services rely on can be build in phases, as more data is added on. For the parts of the world that we don't have data, we output a diagnostic message: no data available. Those datasets can also be used to normalize those same data fields.
 
We can create the initial services, then create some applications using the services to implement commiting the new data back to provider databases (interactively, one record at a time or groups at a time for normalization, in Specify perhaps).  To get some interest quickly, I think we need to allow providers a way to improve their own data without on-site programmers. 
Right.
 
Report cards could include geostatistical analysis of entire collections, genus, species, subspecies.  I've been wanting to explore that ESRI extension!
Yeah! I've been thinking about measuring correlation between a given set of points (per species, genus, museum, collector, etc) and features related to infrastructure (roads, rivers), and also with county centroids. That would give a basis for comparing sampling between datasets. Also, we could measure correlation of datapoints and ecological dimensions, but that is for other purposes.
In this line, I was also thinking about some kind of density measures (spatial and temporal, such as, points per square mile, or per year), and also measures of clustering.
 
Perhaps a short questionaire for existing contributors could determine whether there is any interest in these services, or if they have some ideas we haven't yet come up with.  The initial Fishnet (Ed) reaction was very positive!
Did you show that to Ed? Cool!!!
 
Aimee
 
 

 -----Original Message-----
From: Ricardo Scachetti Pereira [mailto:ricardo@cria.org.br]
Sent: Friday, January 24, 2003 11:47 AM
To: Beach, James H; voron999@UKANS.EDU; Stewart, Aimee Marian; Downie, Scott Russell; reed.beaman@yale.edu
Subject: Value added products to collection as webservices

    Hey, guys.
    How have you been?

    Let me bring something up, so we can discuss it and think about it a little, if you want.
    I've been thinking about value added products to collections and came up with some ideas.
    Among the categories of value added products, I could list those:
    1) Error checking and correction tools, such as: normalization of field values, checking for invalid characters, checking compliance of date or coordinate formats, checking mispellings and typos, cross-reference of textual location descriptions (country, state, county, etc) with georeferences, etc; Reed's georeferencing tool could be put in this category, if you consider that putting a georeference in a record that does not contain one means correcting it, i.e., filling in the blanks on georeference field.
    2) Visualization tools: Showing data on maps (like GIS tools and WMS services);
    3) Statistics and other analysis tools: Measuring clustering of points, correlation between data points and infrastructure, sampling densities, etc.
    As far as I know, the tools in category #3 don't lend themselves well to a webservice implementation. They could be implemented as services, but they are not very important to stand up for themselves.
    Tools in #2 are currently being well covered by Aimee's WMS.
    So I tried to think about possible #1 tools that would become good webservices. The result of that is summarized in the Excel spreadsheet attached.
    I'm still not pretty sure what would be the best architecture to implement them.
    A possibility would be implement each of them as one object, or methods within an object.
    Each object would have a SOAP and/or XML/HTTP wrapper so they become webservices.
    This way, local server components can use the objects directly. Remote components and other users can call the services programmatically using the webservices. Some can be implemented in Kansas, other at CRIA, or we could have copies in both places. Pretty flexible.
    Last but not least, there are the report cards to the collections, that could use some of those services and some other statistics from Lifemapper.
    Questions:
    1) What do you think of the general idea and the proposed architecture?
    2) What do you think of the services listed? Can you add some more to the list?
    3) Do you guys have more ideas for the report cards? What should be in there?
    4) Market analysis: would any one out there (collections community) in need to use such services?

    That is about it.
    Cheers,

Ricardo



Attached is a set of 10 possible value added products that
-- 
Ricardo Scachetti Pereira
Gerente de Pesquisa
Centro de Referência em Informação Ambiental - CRIA
http://www.cria.org.br/

--
This message has been scanned for viruses and
dangerous content and is believed to be clean.

--
This message has been scanned for viruses and
dangerous content and is believed to be clean. 

-- 
Ricardo Scachetti Pereira
Gerente de Pesquisa
Centro de Referência em Informação Ambiental - CRIA
http://www.cria.org.br/

--------------010809040406030205070706--