[SEMANTICS] Re: semantics
Franck Le Petit
Franck.LePetit at obspm.fr
Mon Jun 19 08:53:51 PDT 2006
Dear members of the theory group.
Now that the deadline of the 16th of june is behind us, we would
like to suggest a modification of the Twiki page concerning Semantics
to continue the structuration of the informations.
We would like to add to the main items further informations which
should be required when we will ask to the semantic group to add UCDs
as :
- Is the item a required information or an optinal one for
registration ?
- Are multiple choice possible or not ?
- Do the values associated to the UCDs have to be choosen from a
given list or not and if yes, from which list ?
You will find below what we suggest for the presentation (a mixture
of our previous email with some new inputs as well as Bernard
Debray's last comments). The list of words on the present twiki page
could be found at the end.
Apart these modifications we see several difficulties up to now :
1 - for the algorithms, the list may become very very long and so
unusable for an efficient search in the registries. A solution has to
be find...
2 - Bernard Debray mentionned a very nice problem adding to the
twiki page "Stellar synthesis Populations" in "physical processes".
This is neither a "physical process", nor an "algorithm", nor a
"subject". But that is the way everybody call such codes.
It may be possible to cover such codes with "subject =
Galaxies:Content" (if we want to use the thesaurus.). But who would
think to do a search this way ?
I think I mentionned this difficulty with "PDR codes" which, as
stellar synthesis populations codes cover many physical processes.
Should not we add a new item "Category of code" to cover large
domains of codes with a finite list of possibilities as :
- Large structures formation
- Hydrodynamic
- MHD
- Stellar synthesis populations
- PDR
- Numerical Relativity
- Radiative transfer
- ...
In some cases, the value associated could be the same as in
"physical processes" but not always as for "stellar population
synthesis". It would help a lot the search in registries.
3 - The item "Result parameters" seems unuseful for the
registration of codes. It may be useful for the registration of
simulation results but it is completly linked to the datamodel
describing the simulation results. So I suggest to forget it for the
moment.
4 - I added a new item "Description" where it would be possible to
add an ASCII text describing the code. This exist yet in the
registration of any VO service in the present registries. I do not
know if the registries will allow to search by words in such
description. Does anybody know if we plan to have in the VO a tool
as "spotlight" in the Mac OS X which is able to find any document
when we give it a few words contained in this document ?
Best regards
Franck Le Petit & Fabrice Roy
-------------------------------------------------------------------
Observatoire of Paris
LUTH - Laboratory Universe and Theories
--------------------------------------------------------------------
========================================================================
================================
Suggestion for the new twiki page
========================================================================
================================
For each main items :
- Description of the content of the UCD
- Information on the content :
- Is it a required information or an optional one
- Are multiple choices possible
- Do the informations have to be choosed in a list and if yes,
from which list
- Note : a place where to discuss on the twiki page each main item.
Note : need to specify the differences between what is required to
register codes and results from simulation.
Remark : Required informations to register a code or a simulation
result are not exactly the same. In the latter case, parameters of
the simulation have to be specified which is not true for the
registration of the code.
========================================================================
=================================
Main Items
========================================================================
=================================
Items required to register a code :
1 - Name of the code
2 - Name of the developper / team / contact
3 - Version of the code
4 - Description of the code
5 - Physical processes
6 - Subject
7 - Algorithm
8 - Time evolution
9 - Type of results
10 - Results format
11 - Associated tools
12 - Web address
Items required to register simulation results :
1 - Name of the dataset
2 - Name of the developpers / team / contact
3 - Version of the dataset
4 - Description
5 - Physical processes
6 - Subject
7 - Algorithm
8 - Time evolution
9 - Type of results
10 - Results formats
11 - Associated tools
12 - Web address
13 - Provenance
14 - Results parameters
From 1 to 12 it is similar to the registration of codes.
13 - Provenance : should give access to the code used to produce the
dataset (optional in case of simulation results coming from a code
not described in the VO)
14 - Result parameters : should give the parameters corresponding to
the production of the simulation results. It is difficult to discuss
this item now since it is directly linked to datamodels that we do
not have now.
1 --- Name of the code
------------------------------------------------------------------------
---------------------------
Description : Give the name of the code
Required : YES
Multiple choice : NO
List : NO
Examples : Zeus, Ramses, Cloudy , ....
Note : A UCD seems to exist : meta.id
2 --- Name of the developpers / team / contact
-------------------------------------------------------------------
Description : Give the name of the developpers
Required : Yes
Multiple choice : yes
List : no
Examples : Albert E.
Note : 1 - Implementation done. UCD is meta.curation
3 --- Version of the code
------------------------------------------------------------------------
----------------------
Type:
Required : yes
Multiple choice : no
List : no
Examples :
Note :
- Hervé suggested to use common standards for versionning
- A UCD seems to exist for that : meta.version
4 --- Description of the code
------------------------------------------------------------------------
---------------
Description : ASCII text describing the code / the simulation results
Required : yes
Multiple choice : no
List : no list
Note : Should be covered by the UCD meta.note (?)
5 --- Physical Process
------------------------------------------------------------------------
-------------------------
Description : General keywords to describe the physical
processes that the code deal with
Required : yes
Multiple choice : yes
List : MNRAS keywords, section "Physical data and processes" :
http://www.blackwellpublishing.com/pdf/mnraskey.pdf
Examples : MHD, radiative transfer, molecular processes,
gravitation, ...
Note : To avoid problems of synonyms during the search a list may be
required. In the list on the former twiki page, nearly all words can
be found in the thesaurus but sometimes the meaning is not as precise
as we would wish. Does it seem sufficient ?
6 --- Subject
------------------------------------------------------------------------
-------------------------------------
Description : General keywords describing the kind of objects
the code can deal with
Required : yes
Multiple choice : yes
List : MNRAS keywords : http://www.blackwellpublishing.com/pdf/
mnraskey.pdf
Examples : Star, Galaxy, Large scale structures, Molecular clouds, ...
Note : 1 - Advange to use a list is that the search should be
easier since it will avoid synonyms
2 - All words can be found in the thesaurus.
7 ----
Algorithm --------------------------------------------------------------
------------------------------------------
Description : Most important algorithms in the code
Required : optional
Multiple choice : yes
List : Yes but has to be defined
Examples : nbody, tree, sph, Ali, ...
Note : 1 - A list is necessary for an efficient search or we will
have troubles with specific numerical methods and synonyms
2 - On the other hand, new algorithms can be found and a fixed
list is a problem.
Example :In radiative transfer it is necessary to specify
algorithm. Up to now the main ones were : LVG, Monte-Carlo and Ali.
But recently Moshe Elitzur for Kentucky university has just found a
new revolutionnary method called CEP. How would it be possible to add
this to a fixed list ?
3 - I choosed to group algorithm and time
evolution algorithm. As numericians we clearly do the difference
between the two, but this separation may not be required in the
registries. Since multiple-choice are possible for this item, the
time evolution algorithm should be
8 ---- Time evolution
------------------------------------------------------------------------
-------------------------
Description : Flag saying if the code is stationnar or time dependant
Required : yes
Multiple choice : yes
List : 2 possibilities : yes or no
9 ---- Type of result
------------------------------------------------------------------------
---------------------------
Description : Give the type of results :
Required : yes
Multiple choice : yes
List : yes
- table
- VO-table
- spectre
- VO-spectre
- snapshot
- animation
- fits
- catalog
- statistics
Note : 1 - The list has to be completed.
2 - Should we add to the list : "rawdata" for
binary results from a simulation ?
10 --- Results formats
------------------------------------------------------------------------
-------------------------
Description : Give the format of the results
Required : yes
Multiple choice : yes
List : yes
- ASCII / binary
- HDF5
- giff, jpeg, ...
- mpeg4, quicktime, ...
Note : This information could be usefull to find a code producing
outputs in a specific format to be plug in a workflow.
11 --- Associated tools
------------------------------------------------------------------------
----------------------
Description : give the name in the VO registries of associated tools
to the code
Required : no
Multiple choice : yes
List : no
Note : 1 - This would be usefull if a user looking for a code could
find with the description of the code, associated VO-service for this
code. For example the name/acess to a theoretical database of results
produced by this code, or the name of another code used to analyse
the results of the first code.
12 --- Web Adress
------------------------------------------------------------------------
----------------------------
Description : give the web address where to access the code /
simulation result
Required : yes
Multiple choice : no
List : no
Note : A UCD exists for that : meta.ref.url
===========================================================
Simulation results
===========================================================
13 - Provenance
Description : Give the name of the code used to produce the
theoretical data
Required : Optional
Multiple : No (yes for results from the combination of 2 codes ? )
List : no
Note : 1 - Optional since the code used to produce the data may not
be registered
14 - Results parameters
Description : Give the parameters used to produce the data
Required : yes
Mutliple : yes
List : no
Note: 1 - Maybe it is too early to know how to deal with this item
since it is directly linked to datamodel(s)
========================================================================
======
B **** Thesaurus Keywords
========================================================================
======
Do the keywords of the thesaurus cover all the physical processes :
Thesaurus
radiative
transfer Yes
gravitational
dynamics Gravitation
GR What does GR means ? (General Relat. ?)
relativity
hydrodynamics
Yes
fluid
dynamics W
hat is the difference with hydro ?
plasma
physics
Plasma
magneto-
hydrodynamics Yes
stellar
evolution
No --- see subject below
stellar population
synthesis No --- see subject below
photoionization
atomic processes / molecular processes
photodissociation
molecular processes
physical process (secondary?)
hydrostatic equilibrium (atmosphere?) No (but may
be covered by an hydrodynamics physical process and a stationnary
algorithm)
==============================================================
subject
===============================================================
Subject
Thesaurus
star
star:individual:
stars
stars:
stellar
clusters
galaxies: star clusters
galaxy
galaxies :
volume of space
accretion disc(k)
jet
ISM: jets and outflows
interstellar
medium ISM
molecular
cloud ISM :
clouds
atmosphere (of planet, exo-planet, star...)
stars:atmosphere, nothing for planet atmosphere
stellar evolution stars: evolution
stellar population
synthesis stars: evolution /
galaxy: evolution (?)
===============================================================
Algorithm
===============================================================
* nbody (What does it mean ? particle-particle ? )
* mesh
* collisionless (not an algorithm ? - just a physical
approach ? )
* eulerian grid
* lagrangian grid
* tree
* adaptive refinement mesh
* adaptive refinement tree
* sph
* tree-sph
* pppm
* pm
* pp (What is "nbody" compare to pp ? )
* fokker-planck codes
* vlasov solvers
* orbit solver
* pic (particle in cell)
* tree-pm
* scf (self consistent field)
* ffm (fast multipole method)
* schwarzschild
* ppm (piecewise parabolic method)
* plm (piecewise linear method = godunov)
* Isochrone synthesis
* Fuel consuption theorem
* Spectral method
* Radiative transfer code :
* ETL
* LVG
* Monte-Carlo
* ALI
* CEP
o leap frog
o individual timesteps
o static
o adaptative timesteps
o runge-kutta
o symplectic integrator
Gear
Predictor/Corrector
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/theory/attachments/20060619/ec3cb866/attachment-0001.html>
More information about the theory
mailing list