Brand Niemann, Chairman of the Federal
Semantic Interoperability Community of Practice (SICoP),
and Chris Harding, Forum Director for SOA and Semantic Interoperability,
The Open Group gave a brief
introduction,
setting out the viewpoints of the sponsoring organizations and the objectives
of the meeting. Brand spoke on behalf of
Denise Warzel for the FMMC, as well as for the SICoP.
Ron Schuldt, Senior Staff Systems Architect,
Lockheed Martin Enterprise Information Systems gave the opening keynote presentation:
Convergence of the Semantic
Naming and Identification Technologies -
What are the Choices - What are the Issues?.
He explained the problem of semantic interoperability in
relation to structured and unstructured data, and outlined
the roles of ISO 11179 and the Semantic Web in addressing
the problem in these respective spheres. He explained
the Universal Data Element Framework (UDEF), and how it
can play a key role.
Conor Shankey, CEO of Visual Knowledge, presented on
Semantic
Wikis for Information Management. He described how
semantic agents can help with metadata management
by dealing with the impacts of changes in the
underlying ontologies. He went on to explain
the concept of the semantic wiki, and to show how
semantic wikis can enable the addition of metadata
(which in turn enables inferencing) to
information created collaboratively through wikis,
and can also facilitate collaborative development of
ontologies.They are a key element of tooling for implementation
of the Federal Enterprise Architecture Data Reference Model and
of knowledge reference models in general.
Garry Duvall of OASD presented on RFID/Wireless Coordination
and Collaboration. RFID is an important labelling technology,
enabling item label information to be accessed by reader devices
using wireless communication.
It will dramatically increase the volume of information
available to enterprises, and will also dramatically increase
the amount of item coding that enterprises do.
It is possible to include metadata in RFID identifiers.
There are issues relating to security and privacy,
to the assignment of universally meaningful
identifiers (distributors, for example, sometimes
assign identifiers that omit manufacturer information),
and to how RFID and coding should be managed, but
the biggest challenge is in understanding the impact
on business practices.
Denise Warzel, Assistant Director caDSR,
National Cancer Institute, Center for Bioinformatics described the Common
Framework for Creating, Managing and Deploying Semantically Interoperable Systems
that was developed by the NCI and is in current use by them.
It is a complete metadata management infrastructure,
that combines use of ISO 11179 and ontologies,
and uses model-driven architecture techniques to
derive information management APIs.
It is an excellent use case for semantic information management.
Neil Lovering, of Cisco, described the
Impact of IPv6 on Semantic Interoperability.
IPv6 is the next-generation Internet protocol defined
by the IETF to replace the current version (IPv4),
whose address space will be exhausted some time between
2008 and 2015. IPv6 has an enormous address space, that
will enable components, as well as systems,
to be addressed individually. A part of each address
is assignable within the host system, and can be
used as an identifier. One possibility would be to
make it the same as the RFID identifier for the
component in question.
Brand Niemann posed the key
questions as he sees them, and Chris Harding then added a description of his
key questions.
The speakers and the audience then combined in a discussion moderated by John Yanosy,
chair of the SII WG of the Network Centric Operations
Industry Consortium (NCOIC), to address the question of whether
whether convergence is possible or practical.
Following the discussion, there were summaries of the key points by
Brand, Chris, and John.
The discussion was wide-ranging. It started with
semantics for machine-machine communication and for
person-machine communication, and the crossover of
human-understandable and machine-processable semantics.
It stressed the importance of semantic context: both business context
and IT architecture context. This led to the
need to identify business objects and semantic concepts, and
on appropriate methods and standards for such identification.
ISO 11179 and the Semantic Web will both have roles
to play in enterprise IT architecture. Methods and practices
by which they can play those roles effectively and
in combination must be found.
Identification is crucial because it relates the concept to the referrent object.
Some identifiers can also include semantic information.
Two specific forms of identifier were discussed:
RFID identifiers for business objects, and UDEF identifiers
for semantic concepts. It could be possible to embed UDEF
identifiers as semantic information within RFID identifiers,
which might in turn be embedded within IPv6 host addresses.
No conclusion was reached, however, on the desirability
of doing this.