Hi Howard (and everyone else :-)
The MAMS work I was involved in was in conducting federated searches
across repositories with access restrictions, using SAML assertions
as part of SRW queries. Federated searching is the approach that fits
best in the closed repository scenario as the federated search engine
can broker access between the user and the repository to return the
metadata set appropriate to the security context. In addition, the
negotiation enables the search engine to establish the likelihood of
access in the user's current role, either filtering or indicating in
the result set whether the resource would be available or not, and
what actions are needed to be taken to obtain access.
For open repositories, its probably more efficient to just harvest
them using OAI-PMH to make a metasearch engine along the Google
lines, as the metadata is always going to be the same for every
resource regardless of who's looking at it. Adding something like SRW
to an open repository is just adding another service to the
repository - if you can support it, great.
The MAMS AFS (Authenticated Federated Search) work should be
available from the project at some point - get in touch with James or
Erik for more details.
Btw, when I worked on the code, the MAMS AFS used a local registry
file in IESR format to profile its targets; this was being migrated
to a an actual registry service and combined with federation services
metadata of the SAML 2.0 variety.
- S
On 7 Oct 2005, at 14:22, Howard Noble wrote:
> Hello Scott,
>
> (I was going to reply to you off list but can't because I can't get a
> digital id (service seems to be down)...which makes this request
> kind of
> ironic!
>
> I have been reading up on what the kind people of this list have
> sent me and
> now need to know more about your work on authenticated queries (SAML
> assertions was it?) Can you send me an update on this i.e. is anyone
> implementing here in this area i.e. MAMS?
>
> I've seen this:
>
> http://www.imsglobal.org/query/imsQueryServices.html#1671727
>
> Thanks,
>
> Howard
>
> ----- Original Message -----
> From: "Andy Heath" <[log in to unmask]>
> To: <[log in to unmask]>
> Sent: Thursday, October 06, 2005 10:30 AM
> Subject: Re: metadata mapping
>
>
>
>> Hi Howard,
>>
>> I think its not quite that simple because its not just the
>> Meta-data its the aggregations themselves. I would argue
>> that ff only Meta-data approaches are used we miss taking advantage
>> of commonalities of functionality across different aggregation
>> formats.
>>
>> IEEE ltsc has started work on a solution to this - the
>> Resource Aggregation Model for Learning Education and Training.
>> The Public Information document on this is on the very edge
>> of being released, we're just tidying up the details of
>> formatting of contributors attributions. It will appear on
>>
>> http://www.ieeeltsc.org/wg11CMI/ramlet/
>>
>> It won't help a huge amount at this point with your problem
>> if you need the solution now or yesterday but it does point
>> out the functionality that we hope will solve this in the
>> near future.
>>
>> I'll try find out on this evening's conference call the
>> precise date the PI doc will appear and if you want me
>> to update you on that if you email me tomorrow to remind
>> me I hope I can reply with the date. There will be widespread
>> publicity anyway (Wilbert Kraan is writing the publicity
>> but I think he's travelling this week).
>>
>> Cheers
>>
>> andy
>>
>>
>>> Does anyone have a good reference/ link to a metadata mapping
>>> exercise
>>> i.e. considering object type (image, movie, data set etc.) which
>>> schema
>>> is recommended (i.e. MODS, DC, LOM, DDI, etc)
>>>
>>> We are building metadata repository services and need to ingest and
>>> export to multiple formats then search across them.
>>>
>>> Thanks,
>>>
>>> Howard Noble
>>> Educational Interoperability Specialist
>>>
>>> Oxford University, Learning Technologies Group (LTG)
>>> 13 Banbury Rd, Oxford. OX2 6NN. England
>>> Tel. +44 (0)1865 273281 internal 73281
>>> email: [log in to unmask]
>>> <mailto:[log in to unmask]>
>>>
>>
>>
>> --
>> andy
>> ___________________
>> Andy Heath
>> [log in to unmask]
>>
>>
>
|