You are here: Home > Events > Using the IESR: what's in it for you? > Summary of afternoon discussion session

Summary of afternoon discussion session

Using the IESR: what's in it for you?

A summary of the afternoon discussion session at the Using the IESR: what's in it for you? workshop held on 12 January 2005.

Points raised by attendees included the following.

There should be a number of IESRs, all interoperating. This would involve the IESR becoming more invisible and distributed, rather like the DNS. It was not known at this stage whether this would then involve dynamic searching or harvesting as a method for exchanging data. This distributed model would need to develop one step at a time in step with emerging peer to peer technologies.
UDDI would be attractive to investigate as it is designed for distributed services, but UDDI is not without its downsides, including low usage currently.
The JISC Information Environment community has more familiarity with Z39.50 and OAI-PMH so should look to these standards to distribute the IESR.
Investigating and implementing a distributed IESR will not be trivial and needs to be done with JISC backing.

Licensing for subject terms is a problem.
Obtaining a DDC licence for use by the IESR is proving problematic. If the IESR moves from a project basis into a service then the IESR team should like to provide access to DDC to users. It was thought that the IESR needs DDC to be useful.
What is missing is a terminology service. This is in the semantic web area and it is a very intensive process to map subject schemes to each other.
The list of subject schemes (controlled vocabularies) recorded by the IESR would be linked to a future terminology service. It was pointed out that this linkage might be provided by the portals that make use of the IESR.
The work done by IESR in adding DDC subject terms to collections would be useful to the HILT project as it will provide a correlation between DDC and terms in other subject schemes.
Subject terms are the biggest issue with regard to data quality in the IESR.
IESR collection records have a field to say which controlled vocabulary is in use by the collection. A point was raised whether there could be something in the metadata record to record which field this is. There is overlap here with the IEMSR project and there should be a link from the IESR to the IEMSR to say which application profile is in use by the service. This is something that will be tackled by the IEMSR and will be a use case for IESR.

There should be links from the service record to the collection record in the IESR. Currently the only link is the other way round. This was a known issue and links are to be added immediately by the IESR team.

Service providers will need to specify times when their repositories may be harvested. This might be recorded in the access conditions to the service and would need to be machine readable in order for e.g. portals to do access negotiation invisibly.
A harvestable service would also need to specify what might be harvested.
IESR are aware that they need to address the recording of access management in more detail in a future phase.
An availability service would be useful that regularly checks that described services are available. This would be specific to the service type and the IESR should contain enough information to check whether a service is up or down.

Commercial portal software providers are allowed to include IESR records in their databases for academic customers. The Creative Commons Licence on the records is such that records may be re-used for non-commercial use.
The database and information retrieval software used by IESR is Open Source (Cheshire) and IESR uses a simple configuration file. The application is not yet in a state to be distributed as it is still under development.