Re: JISC-website and PROD PIMS SIMAL DOAP
Sam Easterby-Smith
I sent an email to Ross and Simone earlier in the week trying to pick
up on the matter of getting PIMs to export some useful data for us
all. Ross very rightly pointed out that this should be done in-the-
open on these lists.
Ross recently said:
> As for waiting for PIMs. I disagree. Lets just get on with it, if we
> need to change the name of a field or two, or add a new field it
> will hardly break the bank it's just be a textual search and replace
> on the code.
Actually I agree too on this one - the point is more that we need
access to the pims developers so that we can actually get them to
implement something useful for us all at their end.
So let's really try to kick this off as you say without waiting. My
pass at the things we might want to know about jisc projects which are
not in the regular doap spec but _might_ be in PIMS are as follows
(along with data-type suggestions):
programme-manager (full name/foaf)
institution (name/url/unique number)
funding-period (start and end dates)
jisc-page (url)
programme (name/url)
committee (name/url)
programme-manager (name/foaf)
What do people think?
Cheers,
Sam
Ross Gardler
> Hello Simals,
>
> I sent an email to Ross and Simone earlier in the week trying to pick
> up on the matter of getting PIMs to export some useful data for us
> all. Ross very rightly pointed out that this should be done in-the-
> open on these lists.
>
> Ross recently said:
>> As for waiting for PIMs. I disagree. Lets just get on with it, if we
>> need to change the name of a field or two, or add a new field it
>> will hardly break the bank it's just be a textual search and replace
>> on the code.
>
> Actually I agree too on this one - the point is more that we need
> access to the pims developers so that we can actually get them to
> implement something useful for us all at their end.
>
> So let's really try to kick this off as you say without waiting. My
> pass at the things we might want to know about jisc projects which are
> not in the regular doap spec but _might_ be in PIMS are as follows
> (along with data-type suggestions):
We should not extend DOAP unless there is a particular need to do so.
Wherever possible we should use the appropriate RDF schema
> programme-manager (full name/foaf)
Is this a fact about the project? I would have thought this is a fact
about the programme and therefore should be recorded against the
programme resource. I agree that we should use FOAF to do this - no need
to extend DOAP.
> institution (name/url/unique number)
FOAF can be used for representing organisations as well (I think).
Therefore this is a <doap:maintainer> entry.
> funding-period (start and end dates)
This does belong in the DOAP, does anyone know of a suitable schema for
this?
> jisc-page (url)
This is already available under DOAP
<doap:homepage rdf:resource="http://www.jisc.ac.uk/foo/bar"/>
Simal also supports dc:title so you can have:
<doap:homepage rdf:resource="http://www.foo.org" dc:title="User site"/>
<doap:homepage rdf:resource="http://dev.foo.org" dc:title="Dev site"/>
<doap:homepage rdf:resource="http://www.jisc.ac.uk/foo/bar"
dc:title="JISC page"/>
> programme (name/url)
> committee (name/url)
Simal already supports these as using category:
<category rdf:resource="http://www.jisc.ac.uk/committee/foo" />
Again, Simal uses dc:title to give it a meaningful name.
> What do people think?
I think we only need to find a solution for funding-period.
Ross
Simone Spencer
I'm sorry that I have not been very communicative or proactive recently.
I will post something meaningful soon!
Simone
Assistant Programme Manager
JISC Executive, London
Email. s.sp...@jisc.ac.uk
Tel. 020 3006 6031
Mob. 07929 378139
Skype. spencer.simone
Anything in this message which does not clearly relate to the official
work of the sender's organisation shall be understood as neither given
nor endorsed by that organisation.
----------------------------------------------------------------------
Ross Gardler
On 19 Nov 2008, at 17:01, Ross Gardler wrote:
>
> Sam Easterby-Smith wrote:
>>
...
>> So let's really try to kick this off as you say without waiting. My
>> pass at the things we might want to know about jisc projects which
>> are
>> not in the regular doap spec but _might_ be in PIMS are as follows
>> (along with data-type suggestions):
>
> We should not extend DOAP unless there is a particular need to do so.
> Wherever possible we should use the appropriate RDF schema
>
>> programme-manager (full name/foaf)
>
> Is this a fact about the project? I would have thought this is a fact
> about the programme and therefore should be recorded against the
> programme resource. I agree that we should use FOAF to do this - no
> need
> to extend DOAP.
See expansion on this below
>
>
>> institution (name/url/unique number)
>
> FOAF can be used for representing organisations as well (I think).
> Therefore this is a <doap:maintainer> entry.
I've rethought this. a funding body is not a maintainer since a
maintainer maintains the outputs of the project. A funder would rarely
do that directly.
See below for a suggestion.
>
>
>> funding-period (start and end dates)
>
> This does belong in the DOAP, does anyone know of a suitable schema
> for
> this?
>
FOAF has the ability to define organisations, agents and groups. It
also provides an experimental fundedBy attribute.
So, we can capture everything we want about a funding body with
something like (I've not checked this is 100% valid):
<foaf:Organization rdf:about="http://www.jisc.ac.uk">
<foaf:Name>Joint Information Systems Committee</foaf:Name>
<foaf:member>
<foad:Person rdf:resource="..."/>
<foad:Person rdf:resource="..."/>
</foaf:member>
</foaf:Organizaton>
Then use foaf:fundedBy to link to the Project:
<doap:Project ...>
<foaf:fundedBy rdf:resource="http://www.jisc.ac.uk"/>
</doap:Project>
Now to capture the period of this funding we need a start and end
date. I've not looked for a suitable schema for this yet, but lets
suppose we choose to use resumeRDF (this doesn't seem right though):
<doap:Project ...>
<foaf:fundedBy rdf:resource="http://www.jisc.ac.uk">
<cv:startDate>2007-10-01</cv:startDate>
<cv:endDate>2009-04-30</cv:endDate>
</foaf:fundedBy>
</foaf:fndedBy>
</doap:Project>
As well as the question of which schema to use for start and end dates
we need to be able to capture the roles that people play within the
organisaton (programme manager, programme director etc.).
Does this seem to make sense?
Ross
Ross Gardler
On 19 Nov 2008, at 23:04, Ross Gardler wrote:
>
> As well as the question of which schema to use for start and end dates
> we need to be able to capture the roles that people play within the
> organisaton (programme manager, programme director etc.).
roleVocab looks perfect: http://wiki.foaf-project.org/RoleVocab, I'm
not sure of the level of acceptance/use this schema has.
Ross
Sam Easterby-Smith
>> We should not extend DOAP unless there is a particular need to do so.
>> Wherever possible we should use the appropriate RDF schema
Agreed. Introducing other RDF schema to the mix is exactly what i had
in mind "extension" being an ambiguous term! If we can use existing
ones rather than re-inventing all the better.
>>> programme-manager (full name/foaf)
>>
>> Is this a fact about the project? I would have thought this is a fact
>> about the programme and therefore should be recorded against the
>> programme resource. I agree that we should use FOAF to do this - no
>> need
>> to extend DOAP.
I've found the relationships between "programme managers" and
"programmes" is not necessarily one-to-one and is often subject to
change with the direction of the wind... particularly when looking
back over historic projects. Which is why I've modelled PMs in
relation to projects not programmes.
HOWEVER there is something to be said for doing it the "logical" way
round as this would then give you the member of staff who is
"currently" responsible for the project rather than whoever was
responsible at the time it was done...
@simone your thoughts on this would be helpful!
Mashing a few of these things together... Firstly we can say that the
project is funded by a given programme rather than JISC as a whole...
And secondly we can add the PM relationship at the same time by add
members to the foaf:fundedBy at the project level too - thusly:
> <doap:Project ...>
> <foaf:fundedBy rdf:resource="http://www.jisc.ac.uk/programme-name">
> <cv:startDate>2007-10-01</cv:startDate>
> <cv:endDate>2009-04-30</cv:endDate>
> <foaf:member>
> <foad:Person rdf:resource="http://www.jisc.ac.uk/staff/davidkernohan
> "/>
> </foaf:member>
> </foaf:fundedBy>
> </doap:Project>
(...obviously made up urls here)
Ross - is that kind of thing allowed?
> Now to capture the period of this funding we need a start and end
> date. I've not looked for a suitable schema for this yet, but lets
> suppose we choose to use resumeRDF (this doesn't seem right though):
An interesting suggestion! And yes it seems a bit odd.... but
strangely enough it works... and is making me think re-casting all
this RDF as e-Portfolios for individuals (though that's another story
which we can talk about later)
Oh and I've invited some folks from coventry (Jim Hensman and Ajdin
Brandic) who are working on visualisations of inter-project
dependancies based on all this same data. Welcome to the list!
More soon.
Sam
Simone Spencer
to project rather than programme (or at least strand - not reflected in
current PIMS structure). You would probably want to show who they all
were with the presentation of the programme information though. It's not
quite fair to say the programme manager changes with the wind, there is
always a reason, but it does happen fairly frequently and is not always
helpful to the projects!
Funding is also not as simple as you might think - there is not always a
one-to-one relationship to a programme.
Simone
Assistant Programme Manager
JISC Executive, London
Email. s.sp...@jisc.ac.uk
Tel. 020 3006 6031
Mob. 07929 378139
Skype. spencer.simone
-----Original Message-----
From: simal...@googlegroups.com [mailto:simal...@googlegroups.com]
On Behalf Of Sam Easterby-Smith
Sent: 21 November 2008 15:49
To: simal...@googlegroups.com
Subject: [simal-users] Re: Extending the schema (Re: [simal-users] Re:
JISC-website and PROD PIMS SIMAL DOAP)
Ross Gardler
> Hi folks,
>
>>>> programme-manager (full name/foaf)
>>> Is this a fact about the project? I would have thought this is a fact
>>> about the programme and therefore should be recorded against the
>>> programme resource. I agree that we should use FOAF to do this - no
>>> need
>>> to extend DOAP.
>
> I've found the relationships between "programme managers" and
> "programmes" is not necessarily one-to-one and is often subject to
> change with the direction of the wind... particularly when looking
> back over historic projects. Which is why I've modelled PMs in
> relation to projects not programmes.
>
> HOWEVER there is something to be said for doing it the "logical" way
> round as this would then give you the member of staff who is
> "currently" responsible for the project rather than whoever was
> responsible at the time it was done...
We should be capturing who was responsible and who is responsible now.
FOAF with some kind of time stamping on it is fine for that purpose.
> Mashing a few of these things together... Firstly we can say that the
> project is funded by a given programme rather than JISC as a whole...
> And secondly we can add the PM relationship at the same time by add
> members to the foaf:fundedBy at the project level too - thusly:
Yep
>> <doap:Project ...>
>> <foaf:fundedBy rdf:resource="http://www.jisc.ac.uk/programme-name">
>> <cv:startDate>2007-10-01</cv:startDate>
>> <cv:endDate>2009-04-30</cv:endDate>
>> <foaf:member>
>> <foad:Person rdf:resource="http://www.jisc.ac.uk/staff/davidkernohan
>> "/>
>> </foaf:member>
>> </foaf:fundedBy>
>> </doap:Project>
>
>
> (...obviously made up urls here)
> Ross - is that kind of thing allowed?
I believe your intent is correct, although the use of foaf:fundedBy and
foaf:member is incorrect - but that's minor, we need to work on a real
case and build the data model for it.
Be careful with URLs vs URIs too, you've used URLs in the above which is
wrong. The problem raises its head because I incorrectly used URLs for
the category definitions in the 0.1 version of Simal and as a result
<foaf:fundedBy rdf:resources="http://www.jisc.ac.uk/programme-name">
now points to a category not a funding organisation.
This needs to be corrected in Simal
(http://code.google.com/p/simal/issues/detail?id=217)
We need to make sure we don't repeat the error in other places.
>> Now to capture the period of this funding we need a start and end
>> date. I've not looked for a suitable schema for this yet, but lets
>> suppose we choose to use resumeRDF (this doesn't seem right though):
>
> An interesting suggestion! And yes it seems a bit odd.... but
> strangely enough it works... and is making me think re-casting all
> this RDF as e-Portfolios for individuals (though that's another story
> which we can talk about later)
I'd be interested to here about this at an appropriate time, it's partly
why I liked the resumeRDF way of doing it. That works well for capturing
when a person is acting as programme manage:
<foaf:member>
<cv:startDate>2007-10-01</cv:startDate>
<cv:endDate>2009-04-30</cv:endDate>
...
</foaf:member>
( I have not checked to see if FOAF has a facility for this)
But it seems wrong when saying a project is fundedby there must be a
more appropriate schema.
> Oh and I've invited some folks from coventry (Jim Hensman and Ajdin
> Brandic) who are working on visualisations of inter-project
> dependancies based on all this same data. Welcome to the list!
Indeed. Welcome. And thanks for turning up, we need all the help we can
get - more brains solve more problems in better ways.
Ross
Ross Gardler
> I had also come to the conclusion that Programme Manager is best linked
> to project rather than programme (or at least strand - not reflected in
> current PIMS structure). You would probably want to show who they all
> were with the presentation of the programme information though.
relationship between a project and multiple programme managers over a
period of time. From there we can work out how to capture this
information in RDF.
Can you point us to a project which has such a complex setup. Generally
if we can capture the complex stuff we can capture the simple stuff.
> Funding is also not as simple as you might think - there is not always a
> one-to-one relationship to a programme.
Ross