wiki-archive/twiki/data/TDWG_Systems/TDWGBiodiversityInformationProjectsRedevelopment.txt,v

head	1.5;
access;
symbols;
locks; strict;
comment	@# @;


1.5
date	2009.09.14.01.52.09;	author PiersHiggs;	state Exp;
branches;
next	1.4;

1.4
date	2009.08.22.14.33.52;	author PiersHiggs;	state Exp;
branches;
next	1.3;

1.3
date	2009.07.14.12.58.05;	author PiersHiggs;	state Exp;
branches;
next	1.2;

1.2
date	2009.07.14.00.43.18;	author LeeBelbin;	state Exp;
branches;
next	1.1;

1.1
date	2009.07.13.13.01.32;	author PiersHiggs;	state Exp;
branches;
next	;


desc
@none
@


1.5
log
@none
@
text
@%META:TOPICINFO{author="PiersHiggs" date="1252893128" format="1.1" version="1.5"}%
%META:TOPICPARENT{name="WebHome"}%
On Taxacom, there was a suggestion to redevelop the Biodiversity Information Projects of the World web site (http://www.tdwg.org/biodiv-projects/).  We were always meaning for it to be phase 1 of a longer project, but didn't get past phase 1.

---+++Taxacom suggestions

(these are edited for clarity)

*Cyndy Parr*
It would be cool if the TDWG project database also allowed for ratings and reviews.  Michael Galperin summarizes the state of molecular databases every year in the journal Nucleic Acids Research: http://nar.oxfordjournals.org/cgi/content/abstract/36/suppl_1/D2.  We discussed doing something like this at an ATOL meeting a couple of years ago. It would take a bit of cash to set up. At EOL we've got some annotations we've made to the TDWG project database that we'd be happy to share. Whoever take this and runs with it would be sure to have a highly cited review article every year.

*Michael Heads*
Yes, the TDWG site is very useful for sites' general specifications, but it doesn't point out the sites' weaknesses.  If you did write a review, everyone would use it but no-one would cite it - just like a taxonomic revision...!

*Jim Croft*
yes, that would be a great enhancement.  the TDWG twiki could be configured to enable ratings, reviews and comments, and perhaps even a hotness index based on the number of times a resource is looked up or a link is followed.

To be sustainable such 'metasites' (great term!) really need to be open to community contribution and comment.  with c,10^3 entries and potentially growing exponentially it is likely they will be difficult for a single individual or organization to maintain as a part time labour of love.  Really looking forward to seeing metasites with multiple user feedback and comment.  But with trepidation - what if the don't like what we have given them?  ;)

*Tony Rees*
First, apart from the initial population phase, I don't anticipate exponential growth in the number of relevant links relevant to a "metasite" such as mine, bearing in mind that there are only a few workers/groups pulling together global taxon lists or other general treatments at phylum or class level (the area I am concerned with in the main) and the composition of these key persons/groups changes only fairly slowly through time (once an echinodermologist, always an echinodermologist??)

Second, we already have at least four models happily co-existing: the community-editable one (TDWG Biodiversity Projects of the World); the volunteer/ labour of love type (such as my own and various others); the spinoff from a commercial operation (Thomson Reuters BiologyBrowser); and the funded public-good model (EOL and potentially others down the track -ALA springs to mind in Australia). If any one falls away, the others are a useful fallback I would say.

Many non metasites are also labours of love for their creators - we appreciate them while they are in existence, seek other routes (or replace them ourselves) when they stagnate or disappear - just a fact of life in web land :)

*Arthur Chapman*
This could be like a lot of hotel and software sites that have Reviews. The user doesn't always believe what is written, but goes through a number and makes up ones mind.

Remember the old http://webpagesthatsuck.com that we used when we first started the Internet.  That caused a lot of people to improve their web design, especially in the early days.  I see it is still operating - interesting!  Perhaps we need something similar for taxonomy-related databases.

More and more specialist web sites are being developed that are proving a real boon to researchers - things like Amphibiaweb, Fishbase, AlgaeBase, etc.  A lot more really specialised groups are getting together to design and maintain these databases - it is one of the things I am liking about our modern science - collaboration on projects like these rather than competition. Databases like the Hymenoptera Online Database, The World Spider Catalogue, Crustacea.net, etc. etc.  Such sites maintained by a collaboration of experts can only lead to much more rapid advances in our taxonomies as researchers spend less and less time on doing the chasing up.  As these begin to link to protologues (as I am sure many will over time), then the work of a taxonomist can be spent on taxonomy and not being a librarian.

So - how do we begin the process of developing a metaindex or a metasite to these - perhaps sites like EOL will, over time, develop pointers at the appropriate taxonomic levels to these sites.  Perhaps a new Google-meta-taxa-data-sites will develop - who knows.  We need to encourage these - and/or do we need to develop something new (a la TDWG) and how do we go about that?  Reviews could be added. Good metadata needs to be encouraged that MUST include information on data quality - what checks have been carried out and what is the over-all accuracy/uncertainty.  It is in the future, but I can see it coming and
I know GBIF, and the Atlas of Living Australia (ALA) are encouraging such metadata recording  - but they (and we) need to do more.

*Idea harvested from Twitter*
@@JoBrodie wanted a list of "citizen science" projects.  We could add in a tag that handles that in the BIPOTW site!

---+++Taxacom Summary

   * Add in user interaction - ratings, 'hotness'
   * Must be kept up to date - preferably through automated means
   * Publish an annual review (but where?)
   * Expand information to include site weaknesses, data cleaning processes, and overall accuracy/uncertainty (data quality)
   * Do we create something new or use something else?

---+++Plan

_As an aside: should we implement a metadata standard in this database?  e.g. ISO, etc?  I've used similar sort of fields in other metadata databases_

Of the summary points above, I've ordered easiest to implement are:

---++++Publish Annual Review

Just extract the data and publish it somewhere.... 

Costs/timeline: We could do a manual extract right now and work out some way of writing it up into a paper with some simple statistics?  Like numbers, records, trends of how many use technology A, B, C, etc.  Wouldn't be a bad idea to do this, I'm happy to start it off.

*Lee:* We could easily publish an annual update as a part of the (annual) Proceedings of TDWG. There has also been some discussion about a separate publication of 'brief papers' in Zookeys. This could happen this year in conjunction with TDWG 2009.

*Piers:* I've just checked and I can easily extract the data.  There's a few things we can do quite easily with this - I've got it out of the Typo3 implementation as an Excel sheet - it certainly needs some cleaning as I can see duplicate records in the database for some reason (ah, I see, there's a -1 indicating deleted records, god, that should be pruned).  I'll have to check what's going on there as well.  But I have to say, I am really excited at the idea of putting it into the Proceedings as a standard article each year.  It might be seen as adding value?  And I'm wishing we did something differently with the bloody technology fields... I think I'll set up a database that takes the output file from the Typo3 web site and automates some graph products, parsing, etc., so that makes this process easier to repeat.

*Piers update* Following the call for new projects by the end of August (which I'd like to re-do in this last week of August) I've also been working on the database in the background.  I've had a computer migration (major problems with my laptop) and so I'm behind where I would like to be.  But I'll get it done still.

*Piers update 2* (September) - have re-extracted the site and have begun working on a Microsoft Access database that can generate a range of "interesting" outputs in time for the conference.  I'm not sure we'll hit the deadline of the conference now but will still work towards it.

---++++Add fields to the site

We can add in additional fields to the database and not change the site in a major way.  This is only a small way towards the things outlined above.

   * Add "data cleaning processes" - text field that enables a verbose discussion
   * Add "positional accuracy" - numeric field in (units of measurement) for the average positional accuracy of the database
   * Add "taxonomic accuracy" - text field describing taxonomic accuracy (how to do that)
   * Add in some sort of General uncertainty field
   * Add in a "site weaknesses" field - doubt this will be filled out by many people!
   * Add a reference to a TDWG data standard (e.g. Do you use Darwin Core, etc)
   * Do we have a "checked by" field to see if the values are real?
   * Add in a tag that identifies that the project involves citizen science

Costs/timeline: probably do with the volunteers we have.  Time always an issue, though, especially for the sysadmin group members!

*Lee:* I agree that we should just add a few fields, unless there are some that are just not used. We do need to keep it simple. From comments, we need a numerical rating that we can average. I think we need to solicit and then filter it down to the 'really useful' fields. We also need to somehow encourage at least annual updates from the sites/projects. Sites that are regularly updated should rank higher?  

*Piers:* Yeah, updates should count too.  We could do something about that (as long as we are sensible and don't let people change a record to the same thing to increase their rank).  And simple is a very good idea.  Maybe we should solicit after we publish to get some more feedback on it?

---++++Add functions to the site

Add in new functions.  Specifically:

   * User rankings
   * Tracking statistics of click-throughs
   * Crawlers - this is a long way down the track I think, unless there is a google crawler API :)

When you get to this last one, should we bother?  Should we just let Google do the work for us somehow... @


1.4
log
@none
@
text
@d1 1
a1 1
%META:TOPICINFO{author="PiersHiggs" date="1250951632" format="1.1" reprev="1.4" version="1.4"}%
d66 2
d95 1
a95 1
When you get to this last one, should we bother?  Should we just let Google do the work for us somehow... 
@


1.3
log
@none
@
text
@d1 1
a1 1
%META:TOPICINFO{author="PiersHiggs" date="1247576285" format="1.1" reprev="1.3" version="1.3"}%
d37 3
d64 2
d77 1
@


1.2
log
@none
@
text
@d1 1
a1 1
%META:TOPICINFO{author="LeeBelbin" date="1247532198" format="1.1" version="1.2"}%
d55 2
d59 1
a59 1
Costs/timeline: We could do a manual extract right now and work out some way of writing it up into a paper with some simple statistics?  Like numbers, records, trends of how many use technology A, B, C, etc.  Wouldn't be a bad idea to do this, I'm happy to start it off.
a71 1
   * *Lee:* I agree that we should just add a few fields, unless there are some that are just not used. We do need to keep it simple. From comments, we need a numerical rating that we can average. I think we need to solicit and then filter it down to the 'really useful' fields. We also need to somehow encourage at least annual updates from the sites/projects. Sites that are regularly updated should rank higher?  
d73 15
a87 1
Costs/timeline: probably do with the volunteers we have.  Time always an issue, though, especially for the sysadmin group members!@


1.1
log
@none
@
text
@d1 1
a1 1
%META:TOPICINFO{author="PiersHiggs" date="1247490092" format="1.1" reprev="1.1" version="1.1"}%
d55 2
d70 1
d72 1
a72 1
Costs/timeline: probably do with the volunteers we have.  Time always an issue, though, especially for the sysadmin group members!
@