[Prev][Next][Index][Thread]
CSLI Calendar, 7 January 1998, vol. 13:15
C S L I C A L E N D A R O F P U B L I C E V E N T S
______________________________________________________________________
7 January 1998 Stanford Vol. 13, No. 15
______________________________________________________________________
A weekly publication of the
Center for the Study of Language and Information (CSLI)
Stanford University, Ventura Hall, Stanford, CA 94305-4115
____________
ACTIVITIES DURING 7 JANUARY TO 16 JANUARY 1998
THURSDAY, 8 JANUARY
4:15pm Seminar on Computational Learning and Adaptation (SCLA)
Gates 100
Supervised Classification with Temporal Data
Stefanos Manganaris
IBM Almaden
Abstract below
FRIDAY, 9 JANUARY
12 Noon Logic Lunch
Room 380:383N
The Logic of Partial Information and the Dynamics of
Logic Systems
Areski Nait Abdallah
U. of Toronto, visiting Stanford
Abstract below
12:30pm Seminar on People, Computers, and Design
Gates B01 (HP classroom)
Integrating User Interface Agents with Conventional
Applications
Henry Lieberman
MIT Media Lab
Abstract below
MONDAY, 12 JANUARY
4:30pm Stanford Digital Libraries Seminar
Gates B08
Pam Samuelson
Berkeley
THURSDAY, 15 JANUARY
12 noon CSLI CogLunch
Cordura Hall, Room 100
Planning, Learning, and Acting by Autonomous Agents
Nils Nilsson
Stanford, CS
4:15pm Seminar on Computational Learning and Adaptation (SCLA)
Gates 100
Folke Axel Rauscher
DaimlerBenz, Ulm, Germany.
FRIDAY, 16 JANUARY
12:30pm Seminar on People, Computers, and Design
Gates B01 (HP classroom)
WBI: Intermediaries for Manipulating Web Content
Rob Barrett
IBM Almaden Research
Abstract below
3:30pm Linguistics Department Colloquium
Margaret Jacks Hall 460:146
The ach/ich Alternation and the Representation of German
Palatals
Rob Robinson
Stanford University
____________
SEMINAR ON COMPUTATIONAL LEARNING AND ADAPTATION (SCLA)
on Thursday, 8 January 1998, 4:15pm to 5:30pm
Gates 100
[http://www-csli.stanford.edu/cll/scla.html]
Supervised Classification with Temporal Data
Stefanos Manganaris
IBM Data Mining Solutions
[and Vanderbilt University]
IBM, Almaden, CA
mailto:stefanos@almaden.ibm.com
This talk examines a special case of the supervised classification
problem, where predictor measurements have a natural ordering; in
particular, an ordering due to a dependence with a special variable
such as time. An observation consists of a finite series of
measurements rather than an unordered set. The talk's thesis is that
the information encapsulated in the ordering of predictors can be used
to improve both the accuracy and understandability of classification
models.
The ordering of predictors enables the characterization of
observations over intervals of time, in terms of trends, effectively
reducing the dimensionality of the input space. Questions relating to
the representation, extraction, and evaluation of such features are
addressed. A method is proposed to compress, segment, and smooth 2D
scatter-plots, using MDL-estimation. The concept of line episode is
introduced to capture local trend characteristics. Line episodes can
be extracted and used effectively by any classifier using
trend-episode analysis. Trends can be described in various scales,
with varying levels of detail. The role of scale in the induction of
classification models is explored, and a local scaling algorithm for
line episodes is developed. Trend-episode analysis is extended to
enable any classifier to identify appropriate features at appropriate
scales and induce classification models for instances with multiscale
representations. The proposed framework is evaluated on synthetic
data, under a range of controlled experimental conditions, and on a
NASA telemetry monitoring application. The results show that
multiscale trend-episode analysis can increase both the separability
of classes (leading to more accurate models) and the understandability
of models.
____________
LOGIC LUNCH
on Friday, 9 January 1998, 12 noon
Math Corner 380:383N
[http://math.stanford.edu/upcoming/upcoming.html]
The Logic of Partial Information
and the Dynamics of Logic Systems
Areski Nait Abdallah
U. of Toronto, visiting Stanford
In real-life situations, as well in most scientific disciplines,
reasoning with partial information is the rule rather than the
exception. Classical mathematical logic, however, assumes that all the
information is on the table before any reasoning is performed.
We present a logic for reasoning with partial and tentative
information. We show the need to go beyond the view of logic as a
geometry of static truths, and see logic, both at the proof-theoretic
and at the model-theoretic level, as a dynamics of processes. We see
this dynamics of processes bear with classical logic, the same
relation as the one existing between classical mechanics and Euclidean
geometry. This approach, inspired from theoretical physics, leads to
a geometrization of practical reasoning with partial information.
The applications of the logic of partial information include the
electron interference problem and Heisenberg principle in quantum
mechanics, some classical philosophical logic puzzles, natural
language understanding, formal commonsense reasoning, reasoning about
distributed systems, and reasoning about actions.
The essential research in commonsense reasoning has been developed in
isolation from the disciplines of theoretical computer science and
classical logic. Our work breaks the isolation and establishes deep
links.
Several fragments of this logic have been implemented.
____________
SEMINAR ON PEOPLE, COMPUTER, AND DESIGN
on Friday, 9 January 1998, 12:30-2:00pm
Gates B01 (HP Classroom)
[http://www-pcd.stanford.edu/seminar/]
(SITN Channel E2)
Integrating User Interface Agents with Conventional Applications
Henry Lieberman
Media Laboratory, Massachusetts Institute of Technology
mailto:lieber@media.mit.edu
In most experiments with user interface agents to date, it has been
necessary either to implement both the agent and the application from
scratch, or to modify the code of an existing application to enable
the necessary communication. Instead, we would like to be able to
"attach" an agent to an existing application, while requiring only a
minimum of advance planning on the part of the application developer.
Commercial applications are increasingly supporting the use of APIs
and scripting languages as mean of achieving external control of
applications. Are these mechanisms sufficient for software agents to
achieve communication with applications?
I report some preliminary experiments in developing agent software
that works with existing, unmodified commercial applications and
agents that work across multiple applications. I describe a
programming by example agent, ScriptAgent, that uses a scripting
language, Applescript, to record example procedures that are
generalized by the agent.
Another approach is examinability, where the application grants to the
agent the right to examine internal data structures. We present
another kind of learning agent, Tatlin, that compares successive
application states to infer interface operations. Finally, we discuss
broader systems issues such as parallelism, interface sharing between
agent and application, and access to objects. Picture
Biography: Henry Lieberman has been a Research Scientist at the MIT
Media Laboratory's Software Agents group since 1987. His interests are
in the intersection of artificial intelligence and the human
interface. His current projects involve intelligent agents for the
Web that learn by "watching what you do". He has also built an
interactive graphic editor that learns from examples, and from
annotation on images and video. He worked with graphic designer Muriel
Cooper in developing systems that supported intelligent visual design.
Other projects involve reversible debugging and visualization for
programming environments, and new graphic metaphors for information
visualization and navigation.
From 1972-87, he was a researcher at the MIT Artificial Intelligence
Laboratory. He started with Seymour Papert in the group that
originally developed the educational language Logo, and wrote the
first bitmap and color graphics systems for Logo. He also worked with
Carl Hewitt on actors, an early object-oriented, parallel language,
and introduced the notion of prototype object systems and the first
real-time garbage collection algorithm. He holds a
doctoral-equivalent degree from the University of Paris VI and was a
Visiting Professor there in 1989-90.
Picture
____________
SEMINAR ON PEOPLE, COMPUTER, AND DESIGN
on Friday, 16 January 1998, 12:30-2:00pm
Gates B01 (HP Classroom)
[http://www-pcd.stanford.edu/seminar/]
(SITN Channel E2)
WBI: Intermediaries for Manipulating Web Content
Robert C. Barrett
IBM Almaden Research Center
mailto:barrett@almaden.ibm.com
The current architecture of the web gives the server sole
responsibility for determining the content that results from a
browser's request. Because of the well-defined information stream
between browser and server, this architecture is easily extended to
allow multiple entities to cooperate in producing the final delivered
content. Applications can then be built as intermediary components
which live along the information stream. WBI defines intermediaries
which can observe, produce, and edit web content anywhere between
browser and server. WBI provides a place for applications to
manipulate web content. These applications can be used for
personalization, protocol extensions, collaboration, advising, and
dynamic content generation. The same concepts can be applied to
non-web information streams, such as mouse/keyboard input, information
push, and e-mail.
Biography: Dr. Barrett is a Research Staff Member of the User
Ergonomics Research department in the Computer Science function at the
Almaden Research Center. He received B.S. degrees in physics and
electrical engineering and a M.S. degree in physics from Washington
University (St. Louis) in 1987, and M.S. and Ph.D. degrees in applied
physics from Stanford University in 1989 and 1991, respectively. He
subsequently joined IBM at the Almaden Research Center, where he has
worked on magnetic storage, scanned probe storage, computer pointing
devices, information retrieval technologies, and web-based
intermediaries.
____________
END MATERIAL
The CSLI Calendar appears weekly on Wednesdays throughout the academic
year. Announcements, abstracts, and other information to appear in
the Calendar should be submitted to the editor, who reserves the right
to decide what does or does not go in the calendar
[mailto:incalendar@csli.stanford.edu].
Requests to be added to the mailing list should be sent to
[mailto:majordomo@csli.stanford.edu]. With the lines in the body of
the text of either
subscribe csli-calendar
for the long form or
subscribe csli-short-calendar
for the short form. Problems with subscribing or unsubscribing should
be sent to owner-csli-calendar@csli.stanford.edu.
The full current issue is at
[http://www-csli.stanford.edu/csli/Archive/calendar/1997-98/current.html]
and the archives at
[http://www-csli.stanford.edu/csli/Archive/calendar/].
People on most of the CSLI computers can type 'help csli-calendar' to
see the current issue. They can also type 'help quip' to see the
Linguistics Department newsletter, the Sesquipedalian,
[http://www-linguistics.stanford.edu/Linguistics/].
The CSLI Calendar is also posted each week to
[news://nntp-csli.stanford.edu/csli.bboard].
Information about CSLI's research program is available at
[http://www-csli.stanford.edu/csli/].
For maps to the Stanford University campus see
[http://www.stanford.edu/home/visitors/maps.html].
____________