- From: tim finin <finin@cs.umbc.edu>
- Date: Mon, 04 Jun 2001 16:40:29 -0400
- To: Sandro Hawke <sandro@w3.org>
- CC: www-rdf-logic@w3.org, daml-umbc@cs.umbc.edu
Sandro Hawke wrote:
> Seth Russell wrote:
> >From: "Sandro Hawke" <sandro@w3.org>
> >> The point is that some RDF vocabular terms need to be defined as
> >> "operational" or "performative" for particular agents.
> >I think all we need is just the one term, 'do'. So the triple would be
> >{Agent do Action}. But the action is never performed unless it is issued
> >through a privledged channel ... we wouldn't want an agent to to just read
> >some RDF feed somewhere and act upon it. ...
We should consider an approach in which agents use explicit
communicative acts or speech acts. Since we are still embedded in a
paradigm in which everything is communicated in a fairly high-level
language, this seems reasonable. If you want to get an agent to do X,
you REQUEST or COMMAND it to do X. If you want to let the world know
that you've done X, you INFORM the world thusly.
A mature ACL like KQML or the FIPA ACL is a good place to start, since
a lot of the issues have been worked out. However, this kind of ACL
was designed for a communication paradigm in which messages are
intentionally sent from one agent to another agent and typically both
parties know their conversational partner. The web paradigm is
different -- speech acts can be published for anyone to read. Phil
Cohen has been thinking about some of the changes that will have to be
made to adapt ACLs for this kind of use:
Kumar, Huber, McGee, Cohen, & Levesque: "Semantics of Agent
Communication Languages for Group Interaction," Proc. AAAI 2000,
pp. 42-47, Austin TX, August, 2000.
http://www.cse.ogi.edu/CHCC/Papers/sanjeevPaper/aaai2000.pdf
We've defined an ontology which provides a DAML encoding for the FIPA
ACL (http://www.cs.umbc.edu/~yzou1/daml/acl.daml) as well as one
enabling DAML to be used as a FIPA compliant content language
(http://www.cs.umbc.edu/~yzou1/daml/acldaml.daml. Currently we are
using DAML as a content language in a demo involving KQML speaking
agents, so it's still in the traditional peer-to-peer mode, but we
anticipate experimenting with other ways for agents to communicate via
DAML encoded communicative acts.
We've also been working on using DAML to support various kinds of distributed
trust scenarios in which agents distribute authority to perform certain
actions by a DELEGATE communicative act combined with descriptions
of permissions, obligations and policy constraints.
--
Tim Finin, Prof Computer Science & Electrical Eng, Director Inst. for Global
Electronic Commerce, U Maryland Baltimore County, 1000 Hilltop, Baltimore MD
21250. mailto:finin@umbc.edu 410-455-3522 fax:-3969 http://umbc.edu/~finin/
Received on Monday, 4 June 2001 16:40:51 UTC