[Ala-portal] Names Generator Issues

Mon Mar 17 07:12:54 CET 2014

Hi Allan,

The ala-name-generator is useful if you want to use the Australian National Species list as the main source for your namematching index.  We would not suggest using this to supplement the name matching index with additional species.

In order to support custom species lists we are planning an enhancement to generate the namematching index from a DarwinCore Archive. We would envision that all the species would be provided as a single DWCA with the attached meta.xml.   We think that this could be achieved in the 1-2 week window that you mentioned.  We could provide a DWCA which contains Catalogue of Life  as a basis for you to start with. You can then add additional names to the DWCA as you please. Do you think that this would suit you needs?

Regards
Natasha

From: Allan Koch <allan.kv at gmail.com<mailto:allan.kv at gmail.com>>
Date: Saturday, 15 March 2014 4:19 AM
To: "Martin, Dave (CES, Black Mountain)" <David.Martin at csiro.au<mailto:David.Martin at csiro.au>>
Cc: "Ala-portal at lists.gbif.org<mailto:Ala-portal at lists.gbif.org>" <Ala-portal at lists.gbif.org<mailto:Ala-portal at lists.gbif.org>>
Subject: Re: [Ala-portal] Names Generator Issues

I thank you for the quick answer David and Tim .

We have studied the process to create the namematching index based on the National List of Australia and I see that reproducing the same process to create a new National List is quite complex.

But, for now, we just have the demand to include some names that aren´t included in current namematching index.

If we understood, we need at first, run the Names Generator with the input of a set of CSVs from APNI, APC e AFD. Based on the output of the Names Generator we run the Name Mathcing for creating the LUCENE index, right?

If it´s right, it would be great If we could execute this same standard process, but with de input CSVs modified, with our set of names included in these CSVs (in the same format).

In the future (after this 3 months) we can study the possibility to generate our complete National List .

But for now, we need to include a set of names in the namematching index.
It could be possible to be realized in a short time, in one or two weeks?

Best regards,

Allan Koch Veiga

Núcleo de Pesquisa em Biodiversidade e Computação - BioComp
Laboratório de Automação Agrícola - LAA
Depto. de Engenharia de Computação e Sistemas Digitais - PCS
Engenharia Elétrica - Escola Politécnica da USP
Celular: +55 11 8401-2277
Email: allan.kv at usp.br<mailto:allan.kv at usp.br>

"Stay hungry, stay foolish." Stewart Brand

2014-03-12 23:15 GMT-03:00 <David.Martin at csiro.au<mailto:David.Martin at csiro.au>>:
Thanks Allan, Paulo, Tim.

We appreciate your efforts in setting this software upm locally, and thanks for emailing the list.

1) Versioning

While we are on the track of making this software re-usable by other projects/organisations, it is still very early days. Versioning and packaging are things that we need to tackle properly in the 3 month evaulation period [2] and we are working with GBIF on the best approach here (see Tim's email regarding Ansible). To date, the ALA environment itself is the only place these components are used in production and we manage these closely ourselves. We havent had a need to tightly version components, but as other projects become reliant we need to do this properly. At this point it time, I'd recommend ignoring developments on branches within SVN.

2) ala-name-generator

We didn't anticipate that other projects would be using the ala-name-generator code at this stage (or at all), and instead would rely on the Catalogue of Life names lucene index we've produced [1]. The ala-name-generator code as it currently is isnt suitable for use outside the Australian context. It is dealing with some of the quirks of Australian species lists and merging some elements from different sources. We should have marked wikis to that effect.

That said, we appreciate the need for other projects to use their own taxonomic checklists. This was something I'd hope we tackle in the 3 month evaluation period [2]. There's a few of potential approaches here we are exploring and we'll email this list soon with some progress on this front. I suggest in the meantime, projects make use of the existing index [1].

Thanks again,

Dave Martin
ALA

[1] http://biocache.ala.org.au/archives/col_namematching.tgz
[2] See GBIF's email sent 21st Feb 2014 - "Biodiversity data portals: Using the ALA tooling"

From: "Tim Robertson [GBIF]" <trobertson at gbif.org<mailto:trobertson at gbif.org>>
Date: Thursday, 13 March 2014 2:16 am
To: Paulo André <pfilipak at gmail.com<mailto:pfilipak at gmail.com>>
Cc: "Ala-portal at lists.gbif.org<mailto:Ala-portal at lists.gbif.org>" <Ala-portal at lists.gbif.org<mailto:Ala-portal at lists.gbif.org>>
Subject: Re: [Ala-portal] Names Generator Issues

Hi Paulo

Those are all good comments - I'll make sure the ALA dev team are following those issues.
As this goes forward, it is clear that code releases are going to be needed, so we get immutable binaries in nexus and tagged SVN branches.  I'll try and raise this with Dave Martin.

I'll try and follow the resolutions for the issues you log, build an artifact and verify the same results.
I'm not so much into scala, but IIRC I saw that issue with another artifact.  The solution was to run this before running the command line:
  jar -xf ala-names-generator-1.0-SNAPSHOT-assembly.jar lib

I found this in the way they run the biocache command line tools in:
  https://ala-portal.googlecode.com/svn/trunk/biocache-install/ubuntu/install.sh

It may not be the solution, but worth trying.

I hope this helps,
Tim

On Mar 12, 2014, at 3:57 PM, Paulo André wrote:

Tim

I had have several issues on https://code.google.com/p/ala-portal/source/browse/#svn%2Ftrunk%2Fala-names-generator<https://code.google.com/p/ala-portal/source/browse/#svn/trunk/ala-names-generator>

I wrote on Jira: http://dev.gbif.org/issues/browse/ALA

[]'s
Paulo Andre Filipak

2014-03-12 11:51 GMT-03:00 Tim Robertson [GBIF] <trobertson at gbif.org<mailto:trobertson at gbif.org>>:
Hi Allan,

I am sure the ALA folks will comment when they wake up.  But...

It doesn't appear to be published as an artifact in the ALA maven repository:
  http://maven.ala.org.au/repository/au/org/ala/

You could build from source from: https://code.google.com/p/ala-portal/source/browse/#svn%2Ftrunk%2Fala-names-generator<https://code.google.com/p/ala-portal/source/browse/#svn/trunk/ala-names-generator>

I presume using something along the lines of "mvn clean assembly:assembly"

I hope this helps provide some options,
Tim

On Mar 12, 2014, at 3:32 PM, Allan Koch wrote:

Does anyone knows where I can download this jar: ala-names-generator-1.0-SNAPSHOT-assembly.jar?

I´m trying to generate a new Taxon Name List based on NSL for the Biocache processing.
This instructions has been followed:

http://code.google.com/p/ala-portal/wiki/UpgradeALANames

According the instructions, I need to run this command:

java -Xmx1G -Xms1G -cp .:ala-names-generator-1.0-SNAPSHOT-assembly.jar au.org.ala.names.NamesGenerator --all

But, I can´t find this JAR.

We are trying to build the Scala Project, but we are having some troubles.
Would help me, for while, if I could run a ready JAR.

Best regards,

Allan Koch Veiga

Research Center on Biodiversity and Computing - BioComp
University of São Paulo

Laboratório de Automação Agrícola - LAA
Depto. de Engenharia de Computação e Sistemas Digitais - PCS
Engenharia Elétrica - Escola Politécnica da USP
Celular: +55 11 98401-2277<tel:%2B55%2011%2098401-2277>
Email: allan.kv at usp.br<mailto:allan.kv at usp.br>

"Stay hungry, stay foolish." Stewart Brand
_______________________________________________
Ala-portal mailing list
Ala-portal at lists.gbif.org<mailto:Ala-portal at lists.gbif.org>
http://lists.gbif.org/mailman/listinfo/ala-portal

----------------------------------------------------------------------------------------
Tim Robertson - GBIF Head of Informatics - trobertson at gbif.org<mailto:trobertson at gbif.org>
Global Biodiversity Information Facility http://www.gbif.org/
GBIF Secretariat, Universitetsparken 15, DK-2100 Copenhagen Ø, Denmark
Tel: +45 3532 1487<tel:%2B45%203532%201487>  Mob: +45 2826 1487<tel:%2B45%202826%201487>  Fax: +45 2875 1480<tel:%2B45%202875%201480>
----------------------------------------------------------------------------------------

_______________________________________________
Ala-portal mailing list
Ala-portal at lists.gbif.org<mailto:Ala-portal at lists.gbif.org>
http://lists.gbif.org/mailman/listinfo/ala-portal

----------------------------------------------------------------------------------------

Tim Robertson - GBIF Head of Informatics - trobertson at gbif.org<mailto:trobertson at gbif.org>

Global Biodiversity Information Facility http://www.gbif.org/

GBIF Secretariat, Universitetsparken 15, DK-2100 Copenhagen Ø, Denmark

Tel: +45 3532 1487<tel:%2B45%203532%201487>  Mob: +45 2826 1487<tel:%2B45%202826%201487>  Fax: +45 2875 1480<tel:%2B45%202875%201480>

----------------------------------------------------------------------------------------

_______________________________________________
Ala-portal mailing list
Ala-portal at lists.gbif.org<mailto:Ala-portal at lists.gbif.org>
http://lists.gbif.org/mailman/listinfo/ala-portal

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.gbif.org/pipermail/ala-portal/attachments/20140317/43840ba7/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: meta.xml
Type: application/xml
Size: 2554 bytes
Desc: meta.xml
Url : http://lists.gbif.org/pipermail/ala-portal/attachments/20140317/43840ba7/attachment-0001.rdf 