Copy Link
Add to Bookmark
Report
NL-KR Digest Volume 02 No. 02
NL-KR Digest (1/12/87 12:54:11) Volume 2 Number 2
Today's Topics:
Seminar - Time Modeling with Intervals (SU)
Upenn Colloquium
From CSLI Calendar, January 8, No.12
----------------------------------------------------------------------
Date: 07 Jan 87 1642 PST
From: Vladimir Lifschitz <VAL@SAIL.STANFORD.EDU>
Subject: Seminar - Time Modeling with Intervals (SU)
Commonsense and Nonmonotonic Reasoning Seminar
NEW RESULTS ON TIME MODELLING WITH INTERVALS
Peter Ladkin
Kestrel Institute
(ladkin@kestrel.arpa)
Thursday, January 15, 4pm
Bldg. 160, Room 161K (NEW PLACE!)
James Allen introduced a calculus for reasoning about time using
intervals, instead of points. In this talk, we shall indicate two
new results for time modelling using intervals, and indicate why
they help overcome some of the objections to using an interval system
for reasoning about time. Much of this work is joint with Roger
Maddux. Briefly, we have shown that there is only one countable
representation of the calculus, up to isomorphism, and that the
system of time units introduced in [Ladkin AAAI-86] is isomorphic
to this countable representation.
------------------------------
Posted-Date: Thu, 8 Jan 87 08:44 EST
From: Tim Finin <Tim@cis.upenn.edu>
Subject: Upenn Colloquium
Date: Thu, 8 Jan 87 08:44 EST
CIS Colloquium - University of Pennsylvania
3pm Monday, January 19, 1987 - 216 Moore
UNDERSTANDING DIGITAL SYSTEM SPECIFICATIONS WRITTEN IN NATURAL LANGUAGE
JOHN J. GRANACKI, JR. - UNIVERSITY OF SOUTHERN CALIFORNIA
This talk describes PHRAN-SPAN, a natural language interface to ADAM, the USC
Advanced Design AutoMation System. The interface allows the designer to
specify the behavior of asynchronous digital systems in restricted English
text. This interface will facilitate the construction of a formal model of the
digital systems behavior.
An extension of the USC DDS (Design Data Structure, a knowledge structure for
digital design information) is described with emphasis on timing and control
flow. This extension is used to model various concepts found in system
specifications, such as unidirectional value transfers and temporal
constraints. These models then provide a basis for templates against which
input specifications are matched. This allows PHRAN-SPAN to identify missing
or inconsistent information and to notify the user of the problem.
The interface processes the specification text and produces an internal
representation of the specification. This declarative knowledge representation
scheme is loosely based on Schank's Conceptual Dependency Theory. The use of a
declarative representation simplifies adding vocabulary to the knowledge base
and facilitates mapping of the information into other forms.
A system prototype that parses single sentences will be described and examples
of the types of sentences understood by the system will be presented.
------------------------------
Date: Wed 7 Jan 87 17:51:22-PST
From: Emma Pease <Emma@CSLI.STANFORD.EDU>
Subject: From CSLI Calendar, January 8, No.12
Tel: (415) 723-3561
[Excerpted from CSLI Calendar]
An Application of Default Logic to Speech Act Theory
C. Raymond Perrault
Artificial Intelligence Center and
Center for the Study of Language and Information
SRI International
January 15
One of the central issues to be addressed in basing a theory of speech
acts on independently motivated accounts of propositional attitudes
(belief, knowledge, intentions, ... ) and action is the specification
of the effects of communicative acts. The very fact that speech acts
are largely conventional means that specifying, for example, the
effects of the utterance of a declarative sentence, or the performance
of an assertion, requires taking into consideration many possible
exceptions to the conventional use of the utterances (e.g., the
speaker may be lying, the hearer may not believe him, etc.). Previous
approaches to the problem have attempted to deal with these exceptions
by stipulating the consequences of the utterance as the strongest
condition which is true in all possible conditions of utterance. We
will argue against this approach and present an alternative solution
within the framework of an extension of Reiter's nonmonotonic default
logic. Default rules are used to embody a simple theories of belief
adoption, of action observation, and of the relation between the form
of a sentence and the attitudes it is used to convey. This allows
quite a simple picture of the relation between certain illocutionary
and perlocutionary acts. The emphasis will be on uses of declarative
sentences.
--------------
"Events and LF"
by Stephen Neale
discussion led by Peter Ludlow
January 15
In this article (forthcoming in Linguistics and Philosophy),
Neale argues against a criticism of Situation Semantics due to
Higginbotham (1983). Higginbotham's argument is a response to a
treatment of perceptual reports due to Barwise (1981). Roughly put,
Barwise's suggestion was that perceptual reports such as `John saw
Bill run' express relations between agents and situations.
Higginbotham's counter was that in fact per- ceptual reports contain
implict quantification over events, and that the logical form of the
above sentence is something akin to the following:
(Ex, x an event: Run(x, Bill))[John saw x]
Neale offers four reasons for rejecting Higginbotham's position.
1) Higginbotham's analysis does not predict all the available
inferences in perceptual reports.
2) Higginbotham's analysis leads him to make several errant
predictions about causative constructions.
3) Higginbotham's analysis does not comport well with his chosen
syntactic framework (GB).
4) Higginbotham purports to reject model theory, but must in fact
exploit its resources to capture inferences.
I will BRIEFLY clarify and defend Higginbotham's views and then
open the floor for discussion.
------------------------------
End of NL-KR Digest
*******************
