SWSA Use Cases
The Architecture Committee has organized its use case collection
around
a number of different potential environments for semantic web services.
In
particular, there is interest in all of the following communities that
we
believe will some day be reliant on this technology:
- B2B - Business to Business and in particular Enterprise
Integration Systems. This area represents some of the most mature
service-oriented technology, but is largely based on custom solutions
to large-scale e-commerce problems. It is one of the main targets of
the current generation of Web Service development. The key motivation
here for SWS is in the area of semantic mediation between different
organizational vocabularies, as service models and protocols have been
coordinated among a few dominant organizations, although there are
several competing standards.
- Grid - Scientific Grid Computing is an emerging technology
for widely distributed large-scale heterogeneous computing. It is an
area
where there are independently developing standards efforts, such as the
Open Grid
Services Architecture proposal.
- Ubiquitous - Ubiquitous Computing efforts are largely
focused on the emerging area of wireless computing, and the use of
portable devices to tap into a variety of spatially proximate computing
services in a very context-specific fashion. Due to the extremely
dynamic variation in available services this implies, semantic service
descriptions will be an important element of the success of these
efforts.
- Web - Commercial and Informational Web Services were the
original motivation for research on SWS. Key motivators here are the
ability to dynamically discover and combine web services under the
control of software agents, for such uses as comparative shopping,
coordination of specialized business services (e.g. travel plans,
meeting arrangements...), etc.
- Agent - Semantic Agent-oriented organizational computing
addresses the needs of large heterogeneous organizations to do
distributed planning and control or coordination. Logistics, for
example, is an organizational function that can benefit from SWS
techniques.
By developing cases focused in each of these areas, our we seek to
elucidate technical and architectural differences and similarities that
must be addressed in our abstract architecture. The aspects of semantic
web service functions that will be addressed, and, where appropriate,
highlighted by each community, are:
a) Service invocation planning (message
formulation) and response interpretation based on process descriptions
b) Choreography (protocol) interpretation and execution
c) Semantic translation/mediation (e.g., of message
content, process descriptions or advertisments)
d) Candidate service identification (matchmaking) and
selection
e) Automated Process composition
f) Process mediation and delegation
g) Service process status tracking
h) Ontology management and access
i) Security (including identification,
authentication, delegation and policy-based authorization)
j) Reputation services
k) Service failure handling and compensation
l) Negotiation and contracting
m) Server executable process management (service factories,
instantiation, migration)
These areas will be addressed through the development of specific use
cases within the community-based scenarios developed. Emphasis
throughout should be on requirements for specific agents (clients,
servers, middle agents) and associated (abstract) protocols.
We are using the shared
SWSI
Use Case Template as a guide.
Draft Usage Scenarios (group leader)
B2B (Christoph Bussler)
Semantic Web enabled
Business Protocol Standards (RosettaNet and Enterprise Integration).
(Zaremba/Bussler)
Web Services (Massimo Paolucci)
Amazon B2C Use Case based on the W3C Web Services Architecture Model
(Paolucci)
Conversational
Protocol of a book purchase service (Boualem
Benatallah, Fabio Casati, Farouk Toumani)
Quality of Service (Amit Sheth)
Dynamic QoS based Supply Chain
GRID (Carole Goble)
Scientific Computing uses of semantic services
descriptions and data models.
Ubiquitous Computing (Tim Finin)
Dynamic Service Discovery for UC, including security
policies.