[Date Prev][Date Next][Date Index]
CSLI Calendar, Wednesday, 3 May 2000, vol. 15:29
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
______________________________________________________________________
3 May 2000 Stanford Vol. 15, No. 29
______________________________________________________________________
A weekly publication of the
Center for the Study of Language and Information (CSLI)
Stanford University, Ventura Hall, Stanford, CA 94305-4115
____________
ACTIVITIES FROM 3 MAY TO 12 MAY 2000
WEDNESDAY, 3 MAY 2000
12:45pm CS548: Distributed Systems Research Seminar
McCullough 150
Taming the Internet Service Construction Beast:
Persistent, Cluster-based Distributed Data Structures
Steven D. Gribble
University of California at Berkeley
http://www.stanford.edu/class/cs548/
Abstract below
4:15pm Broad Area Colloquium in AI,
Geometry, Graphics, Robotics, and Vision
TCseq201, Lecture Hall B
Humanoid Robots and Social Interaction
Rodney Brooks
Director, AI lab, MIT
http://robotics.stanford.edu/ba-colloquium/
Abstract below
4:15pm EE380: Computer Systems Laboratory Colloquium
Gates B03 (NEC Auditorium)
Wiring Tonga: From the Ground Up and the Sky Down
Dewayne Hendricks
Dandin Group
http://www.stanford.edu/class/ee380/contents.html
8:00pm Kant Lecture 2
Bldg. 300:300
Rational Beliefs, Rational Desires:
Rational Desires
Peter Railton
University of Michigan
http://www-philosophy.stanford.edu/ce.html
THURSDAY, 4 MAY 2000
12 noon CSLI CogLunch
Ventura Hall 17 (note different room)
Josh Tenenbaum
Stanford University
http://www-csli.stanford.edu/events/Coglunch/
Abstract below
3:00pm PhD Oral Examination
Packard Bldg., Room 101
Supporting Ubiquitous Computing with
Stateless Consoles and Computation Caches
Brian Schmidt
Computer Systems Laboratory
http://suif.stanford.edu/~bks/
Abstract below
3:15pm Stanford Learning Lab Speaker Series
Press Warehouse, Press Staff Training Room
Evolution of an Online Education Community of Practice
Mark Schlager, Judith Fusco, Patricia Schank
SRI's Center for Technology in Learning
http://sll.stanford.edu/index.shtml
Abstract below
4:00pm Xerox PARC Forum
George Pake Auditorium at Xerox PARC
Rethinking (& Redoing) R&D
Arno Penzias
Senior Technology Advisor
Lucent Technologies and Bell Labs Innovation
http://www.parc.xerox.com/ops/projects/forum/
Abstract below
4:30pm ACM Allen Newell Award Lecture
Terman Auditorium
Computers and Trust
Nancy Leveson
MIT
http://calendus.stanford.edu/CS/read/event_8964_CS_read.html
Abstract below
8:00pm Kant Lecture 3
Bldg. 300:300
Rational Beliefs, Rational Desires:
Rational Beings, Being Rational
Peter Railton
University of Michigan
http://www-philosophy.stanford.edu/ce.html
FRIDAY, 5 MAY 2000
12 noon Logic Lunch
Bldg. 380:383N (math corner)
Abstract versus Concrete Models of
Computation on Partial Metric Algebras
Jeffrey Zucker
McMaster University
http://www-logic.stanford.edu/seminars.html
Abstract below
12:30pm CS547: Human-Computer Interaction Seminar
Gates B03
The Scent of a Site:
A System for Analyzing and Predicting Information
Scent, Usage, and Usability of a Web Site
Ed Chi
Xerox PARC
http://www-pcd.stanford.edu/cs547/
Abstract below
3:15pm Infolab Seminar
201 tcSEQ (across from Gates)
Ecologies on the Web
Lada Adamic
XEROX PARC
http://www-db.stanford.edu/dbseminar/dbseminar.html
3:15pm Kant discussion seminar
Bldg. 90:92Q
Rational Beliefs, Rational Desires
Peter Railton
University of Michigan
http://www-philosophy.stanford.edu/ce.html
3:30pm Semantics and Pragmatics Workshop
Margaret Jacks Hall 460:126
Discourse Structure
Wallace Chafe
UC Santa Barbara
http://www-linguistics.stanford.edu/semgroup/
Abstract below
3:30pm Art and Art History Symposium
Annenberg Auditorium, Cummings Art Bldg.
Photography and Memory
till 6:30pm
two-day symposium
http://www.stanford.edu/dept/ccva/program.html
SATURDAY, 6 MAY 2000
9:00am Art and Art History Symposium
Annenberg Auditorium, Cummings Art Bldg.
Photography and Memory
till 6:00pm
two-day symposium
http://www.stanford.edu/dept/ccva/program.html
TUESDAY, 9 MAY 2000
4:00pm Biology Talk
Geology Corner, Room 105
The Evolution of the Brain
John Allman
Cal Tech
http://calendus.stanford.edu/biology/
4:15pm Statistics Seminar
Sequoia Hall 200
Estimating Probability of Causation
Judea Pearl
UCLA
http://www-stat.stanford.edu/seminars/
Abstract below
WEDNESDAY, 10 MAY 2000
12 noon Psychology Developmental Brownbags
Jordan Hall 420:286
On Artistry in the Conduct of Social Science Research
Elliot Eisner
Stanford University
http://matia.stanford.edu/html/talks.html#dev_brownbag
12:45pm CS548: Distributed Systems Research Seminar
McCullough 150
Proactive DNS Caching: Addressing a Performance Bottleneck
Edith Cohen
AT&T Labs-Research
http://www.stanford.edu/class/cs548/
Abstract below
4:15pm Broad Area Colloquium in AI,
Geometry, Graphics, Robotics, and Vision
TCseq201, Lecture Hall B
Searching for the Actual Cause
Judea Pearl
UCLA
http://robotics.stanford.edu/ba-colloquium/
Abstract below
THURSDAY, 11 MAY 2000
12 noon CSLI CogLunch
Cordura Hall, Room 100
Intelligent Communication - AI and NL
John McCarthy
Stanford
http://www-csli.stanford.edu/events/Coglunch/
4:00pm Xerox PARC Forum
George Pake Auditorium at Xerox PARC
Launching Products in the Internet Age
Catherine Kitcho
Consultant and Author, "High Tech Product Launch"
http://www.parc.xerox.com/ops/projects/forum/
7:30pm Stanford Phonology Workshop
Margaret Jacks Hall 460:126
Title to be announced
Luc Baronian
Stanford University
http://www-linguistics.stanford.edu/pinterest/
FRIDAY, 12 MAY 2000
12:30pm CS547: Human-Computer Interaction Seminar
Gates B03
Facial Affect in Human and Machine
Diane Schiano
Formerly Interval Research
http://www-pcd.stanford.edu/cs547/
3:15pm Philosophy Department Colloquium
Bldg. 90:92Q (Philosophy)
Title to be announced
Harry Frankfurt
Princeton University
Undergraduate Philosophy Conference Distinguished Lecturer
http://www-philosophy.stanford.edu/ce.html
3:15pm Infolab Seminar
201 tcSEQ (across from Gates)
Preliminary Topic: Agents, e-commerce and XML
Charles Petri
Stanford University
http://www-db.stanford.edu/dbseminar/dbseminar.html
____________
CS548: DISTRIBUTED SYSTEMS RESEARCH SEMINAR
on Wednesday, 3 May 2000, 12:45pm
McCullough 150
http://www.stanford.edu/class/cs548/
Taming the Internet Service Construction Beast:
Persistent, Cluster-based Distributed Data Structures
Steven D. Gribble
University of California at Berkeley
mailto:gribble@cs.berkeley.edu
This talk presents the design, implementation, and evaluation of
highly available, durable, scalable distributed data structures
(DDS's) on clusters of workstations, intended primarily to vastly
simplify the construction of scalable Internet services. A DDS is a
self-contained, self-managing, consistent and available repository
that exports a data-structure interface. A DDS platform decouples data
persistence and consistency requirements from the rest of
cluster-based Internet service logic, greatly simplifying service
design and implementation. The main hypothesis of this work is that by
providing service authors a small but carefully chosen selection of
DDS's (such as a log, a hash table, and a tree), these authors will
have enough flexibility to implement a wide variety of interesting
services, but will also be shielded from many of the complexities of
scalable, available, consistent state management on clusters. The
DDS's are built on an asynchronous I/O layer that uses state machines
to achieve high concurrency and data throughput, and design of the
DDS's exploits properties of clusters (such as ample bandwidth, low
latency, and a very small probability of a network partition) in areas
such as the design of consistency protocols and recovery techniques.
Example services built on the DDS platform, such as the instant
messaging gateway and translation proxy "Sanctio", will be discussed
in addition to the core platform.
About the speaker: Steven D. Gribble is a graduating Ph.D. student
from the University of California at Berkeley, working primarily with
Dr. Eric A. Brewer as part of the Ninja research group. Steve's
research interests include the design, implementation, and evaluation
of highly available, scalable Internet service platforms on cluster of
workstations, and in the use of such platforms to aid the deployment
and adoption of "Post-PC" devices and their infrastructure. Steve
received his B.Sc. in Computer Science and Physics from the University
of British Columbia in 1995, and his M.Sc. in Computer Science from
Berkeley in 1997. He was also a co-founder of ProxiNet, Inc. (recently
acquired by Puma Technology), which is commercializing thin client
technology created at UC Berkeley.
____________
BROAD AREA COLLOQUIUM FOR
AI-GEOMETRY-GRAPHICS-ROBOTICS-VISION
on Wednesday, 3 May 2000, 4:15pm
TCseq201 (across from Gates)
http://robotics.stanford.edu/ba-colloquium/
Humanoid Robots and Social Interaction
Rodney Brooks
Director, AI lab
MIT
Robots with humanoid form are a tool for investigating and validating
cognitive theories. We are building a number of humanoid robots (Cog,
Kismet, Coco, plus partial systems K2, M4, Lazlo, etc.) and use four
guiding principles: embodiment, multi-modal integration, development,
and social interaction. With these principles but without any
kinematic or dynamic models, and without any central system with
symbolic representations we are able to get our robots to carry out
complex manipulation tasks, and interact with people using the same
social cues that people have evolved over millions of years. Our
systems are modeled on how humans and animals achieve the same sorts
of performance.
About the Speaker: Rodney Brooks is the Fujitsu Professor of Computer
Science and Engineering the Director of the MIT Artificial
Intelligence Lab, and is the Chairman and CTO of iRobot
Corporation. He received a PhD in Computer Science from Stanford in
1981, was on the research staff at CMU and MIT, then on the CS faculty
at Stanford, and joined the MIT faculty in 1984. He developed many of
the key ideas of behavior-based robots which are finding applications
from consumer products to oil production to planetary exploration. His
current research focuses on humanoid robots, models of development of
cognition, and social interaction between people and robots.
____________
CSLI COGLUNCH
on Thursday, 4 May 2000, 12:00pm
Ventura 17 (note room change)
http://www-csli.stanford.edu/events/Coglunch/
Is probabilistic inference a natural human ability?
Josh Tenenbaum
Stanford University
Numerous studies from the judgment and decision-making literature have
demonstrated that people are prone to systematic errors on
probabilistic reasoning tasks. On their face, these findings suggest
that probabilistic inference is not a natural human ability. Rather,
people employ shortcut strategies, such as Tversky and Kahneman's
"representativeness heuristic", that follow a logic distinct from
probability theory. This conclusion contrasts sharply with recent
findings in other areas of human learning and inference, where a
number of researchers have constructed successful Bayesian models of
tasks ranging from concept learning, syllogistic reasoning, and
hypothesis testing to causal inference, word learning and predicting
the future. Are we therefore forced to conclude that people have a
natural ability for probabilistic inference in every situation
*except* when they are explicitly asked to manipulate probabilities?
In this talk, I will first review some of my work on Bayesian models
of concept learning and then try to reconcile this work with the
conclusions of Tversky and Kahneman. I will argue that the
representativeness heuristic, rather than being seen as an alternative
to probability theory, is itself best described as a Bayesian
computation, and one that is intimately related to the Bayesian
computations underlying our abilities to learn concepts, words, and
causal relations from examples. Time permitting, I will also discuss
a parallel with Popper's arguments against probability theory as a
viable logic of confirmation in scientific inquiry.
____________
PHD ORAL EXAMINATION
on Thursday, 4 May 2000, 3:00pm
Packard Building (EE department), Room 101
Supporting Ubiquitous Computing with
Stateless Consoles and Computation Caches
Brian Schmidt
Computer Systems Laboratory
Stanford University
http://suif.stanford.edu/~bks/
Rapid technological advances are leading us to an era in which we will
have ubiquitous access to virtually unlimited computing resources.
Harnessing the power of these resources requires our computing
environments to become hassle-free and much easier to manage, as users
cannot be concerned with machine specifics and management details.
This talk describes two core concepts to help enable this transition.
First, we propose decoupling displays from their servers, i.e. using
network-attached frame buffers to access active computations from any
location. These simple consoles are cheap, stateless, low-level
access devices that require no administration, support transparent
user mobility, and isolate users from desktop failures. We
demonstrate that an architecture based on this concept is feasible
with modern technology, providing an interactive experience that is
indistinguishable from the traditional desktop, even for highly
interactive video games.
The next step is to manage the computing resources at the back end,
and we place several requirements on this remote computational
service. Users must have privacy, security, and isolation from each
other, and they should be free to roam throughout the system. The
system must be scalable, reliable, maintainable, available,
extensible, and efficient. To meet these goals, we restructured the
operating system using the concept of "compute capsules." Capsules
provide a private, portable, persistent, customizable computing
environment with active processes. They are self-contained, running
computations, and the CPU/OS merely serves as a cache to hold them.
We show how capsules help meet the demands of the future computing
infrastructure by supporting persistent computations on a globally
distributed collection of anonymous processors.
____________
STANFORD LEARNING LAB SPEAKER SERIES
on Thursday, 4 May 2000, 3:15pm - 4:30pm
Press Warehouse, Press Staff Training Room
http://sll.stanford.edu/index.shtml
Evolution of an Online Education Community of Practice
Mark Schlager, Judith Fusco, Patricia Schank
SRI's Center for Technology in Learning
Traditionally, the K-12 education sector has lagged far behind the
corporate sector in providing its professional staff with the
communication channels, information resources, tools, and workplace
conditions needed to manage their own professional development. Most
teachers remain isolated from innovative tools and expertise, as well
as from their peers (who may not be in their own school). They are
left without support after a professional development project or
workshop ends. SRI's Center for Technology in Learning (CTL) has
developed a new community-based model of online teacher professional
development that we believe can help overcome isolation, access, and
support problems that teachers face. We have built an online
environment called TAPPED IN ( http://www.tappedin.org/ ) that serves
as a testbed for our research to into the efficacy and sustainability
of our model ( http://www.tappedin.org/info/papers/ ). TAPPED IN is
not a conventional Web portal. Rather, it is a virtual workplace where
education professionals can collaboratively serve as human resources
for one another. We are now preparing to scale up our research and
development efforts to support large segments of the K-12 educator
population. This talk will describe some of the successes we have had,
some surprises, and the challenges that lie ahead as we scale up the
TAPPED IN community, technology, and services.
Biographies:
Dr. Mark Schlager is a Senior Cognitive Scientist and Associate
Director of Learning Communities at SRI's Center for Technology in
Learning. Mark has worked in the area of online collaboration R&D in
industry, government, and education for more than 15 years. His
current research focuses on the development of community-based
pedagogy and online technologies for teacher professional development.
Mark earned his Ph.D. in Cognitive Psychology at the University of
Colorado, Boulder, and a B.A. in Psychology at Temple University.
Dr. Judith Fusco is a Research Scientist at SRI's Center for
Technology in Learning. Judith's research focuses on the social and
technical infrastructures needed to cultivate and sustain online
communities. As TAPPED IN's Director of Online Community Research,
Judith has grown and managed the TAPPED IN online community, now
approaching 8,000 education professionals, with the help of a
dedicated crew of teacher affiliates, summer interns, and community
volunteers. Before joining SRI, Judith ran Convomania, an online
community of sick and disabled children hosted by Apple Computer.
Judith earned her Ph.D. in Cognitive Psychology from the University of
California, Riverside, and a B.A. in Psychology from Washburn
University.
Dr. Patricia Schank is a Senior Research Scientist at SRI's Center for
Technology in Learning. Patricia's research focuses on the design,
development, and evaluation of computer-mediated learning
environments. Recent projects include the development of collaboration
environments to support learning and professional development,
interactive shared representations to support inquiry and
argumentation in chemistry, and Web-based resource banks of evaluation
products and performance assessments. Patricia received her Ph.D. from
the University of California at Berkeley in Education in Mathematics,
Science, and Technology (EMST). She has an M.S. in Computer Science
from U.C. Berkeley, as well as a B.S. in Mathematics and Computer
Science from the University of Nebraska, Lincoln.
____________
XEROX PARC FORUM
on Thursday, 4 May 2000, 4:00pm - 5:00pm
George Pake Auditorium, Xerox
http://www.parc.xerox.com/ops/projects/forum/
Rethinking (& Redoing) R&D
Arno Penzias
Senior Technology Advisor
Lucent Technologies and Bell Labs Innovation
As Bell Labs' Vice President of Research, I led that organization from
just before the 1982 breakup of the Bell System, to its transition
into Lucent Technologies in 1995. For much of that time, my colleagues
and I prided ourselves on keeping our internal environment as
unchanged as possible. Then, when I finally saw this approach as no
longer tenable, I began a long (and still evolving) process of change,
making Bell Labs Research far more relevant to marketplace needs,
while maintaining its long-standing traditions of creativity and
excellence. This talk recounts the story of what we did, how we did
it, and what has happened as a result.
Biography: Arno Penzias is a Venture Partner at New Enterprise
Associates, as well as a Senior Technology Advisor of Lucent
Technologies, Bell Labs Innovations. In this role, he prowls Silicon
Valley and similar places, seeking out promising technology futures
and catalyzing their applications.
Prior to assuming his current position in May of 1998, he held a
series of leadership positions in Bell Labs' Research organization.
As Vice President of Research -- from the Bell Systems' breakup in
1982, to the creation of Lucent Technologies -- he boosted Bell Labs'
impact upon the global marketplace, while maintaining its
long-standing reputation for scientific excellence.
Dr. Penzias began his scientific career in 1961 when he joined Bell
Laboratories as a Member of Technical Staff. He conducted research in
radio communication and took part in the pioneering Echo and Telstar
communications satellite experiments.
As a scientist, Dr. Penzias is best known for his contributions to
astrophysics which earned him the Nobel Prize for Physics in 1978.
A sought-after speaker on emerging trends, he has written a number of
articles on information technology, especially its impact on business
and society. His highly acclaimed book, Ideas and Information, was
published by W.W. Norton. A second book, Digital Harmony: Business,
Technology and Life After Paperwork, published by Harper Collins,
charts the course of the Information Revolution and its demands for
higher levels of system integration.
Dr. Penzias received a Bachelor of Science degree from the City
College of New York. After serving the U.S. Army Signal Corps, he
attended Columbia University where he received his master's and
doctorate degrees. He has also received a number of honorary degrees,
as well as other awards for his contributions to science, R&D
management and public service. His affiliations include advisory
relationships and board memberships with several small- and
medium-sized technology-based companies.
____________
ACM ALLEN NEWELL AWARD LECTURE
on Thursday, 4 May 2000, 4:30pm - 6:30pm
Terman Auditorium
sponsored by the Knowledge Systems Lab
Computers and Trust
Nancy Leveson
MIT
"We seem not to trust one another as much as would be desirable. In
lieu of trusting each other, are we putting too much trust in our
technology?" - T.B. Sheridan.
Computers are being introduced into the control of virtually every
dangerous system, including nuclear weapons, transportation systems
(aircraft, automobiles, trains), medical devices, and chemical and
nuclear power plants. Few engineering techniques exist to provide
assurance that safety is not being degraded by the substitution of
digital systems for the electro-mechanical designs that have been
perfected through decades and sometimes centuries of experience. At
the same time, nothing is absolutely safe, and computers provide
important advantages over the human operators, social systems, and
engineered devices they are replacing.
This talk will attempt to examine whether concern is justified. Are
we putting too much trust in computers? Will introducing computers to
assist or replace human operators eliminate or reduce the problem of
human error? Are there limits to the reasonable uses of computer
technology? If so, what do we need to do to stretch those limits?
____________
LOGIC LUNCH
on Friday, 5 May 2000, 12 noon
Math Corner 380:383N
http://www-logic.stanford.edu/seminars.html
Abstract versus Concrete Models of
Computation on Partial Metric Algebras
Jeffrey Zucker
McMaster University
A model of computation is abstract if, when applied to any algebra,
the resulting programs for computable functions and sets on that
algebra are invariant under isomorphisms, and hence do not depend on a
representation for the algebra. Otherwise it is concrete. Intuitively,
concrete models depend on the implementation of the algebra.
The difference is particularly striking in the case of topological
partial algebras, and notably in algebras over the reals. We
investigate the relationship between abstract and concrete models of
partial metric algebras. In the course of this investigation,
interesting aspects of continuity, extensionality and non-determinism
are uncovered.
This is joint work with J.V. Tucker (Swansea, Wales).
____________
HUMAN-COMPUTER INTERACTION SEMINAR
on Friday, 5 May 2000, 12:30-2:00pm
Gates B03 (NEC Classroom)
http://www-pcd.stanford.edu/cs547/
The Scent of a Site:
A System for Analyzing and Predicting Information Scent,
Usage, and Usability of a Web Site
Ed Chi
Xerox PARC.
mailto:echi@parc.xerox.com
Designers and researchers of users' interactions with the World Wide
Web need tools that permit the rapid exploration of hypotheses about
complex interactions of user goals, user behaviors, and Web site
designs. We present an architecture and system for the analysis and
prediction of user behavior and Web site usability. The system
integrates research on human information foraging theory, a reference
model of information visualization and Web data-mining techniques. The
system also incorporates new methods of Web site visualization (Dome
Tree, Usage Based Layouts), two new algorithms from a new predictive
modeling technique for Web site use (WUFIS and IUNIS), and new Web
usability metrics. The research was done with Peter Pirolli and James
Pitkow.
Biography: Ed H. Chi is a research scientist at Xerox Palo Alto
Research Center's User Interface Research Group. He has been working
on Information Visualization and User Interfaces since 1993. His area
of research and expertise is software systems for computer-human
interaction and 2D or 3D user interfaces. His most current project is
the study of information ecology---understanding how users navigate
and understand information environments, including the Web. His
Ph.D. project was "Spreadsheet for Visualization" -- a data
exploratory tool using a 'spreadsheet metaphor' that allows each cell
to hold an entire data set with a full-fledged visualization. In the
past, he has also worked on computational molecular biology, and
recommendation systems. Ed received his Ph.D. 1996-1999,
M.S. 1994-1996, and B. Comp Sci. 1992-1994.
____________
STANFORD SEMANTICS AND PRAGMATICS WORKSHOP
on Friday, 5 May 2000, 3:30pm
Margaret Jacks Hall 460:126
http://www-linguistics.stanford.edu/semgroup/
Discourse Appreciation
Wallace Chafe
UC Santa Barbara
The implication of my title is that a piece of discourse resembles a
work of art, in the sense that it offers multiple levels of
understanding that one can learn to appreciate, more or less as one
might learn to appreciate a painting or a piece of music. After some
introductory remarks on the value of examining language in actual use,
I will illustrate various aspects of discourse structure with an
excerpt from a three-person conversation, emphasizing how such
material can provide evidence for the nature of mental phenomena such
as consciousness, memory, and affect. I will discuss the organization
of discourse into topics, subtopics, sentences, and prosodic phrases;
the ways in which such elements are linked through both words and
prosody; and the schematic and interactional factors that keep
language and thought moving forward through time. After introducing
the nature of "activation cost" in terms of given, accessible, and new
ideas, I will show how a hypothesized "one new idea at a time"
constraint can help explain the shape of discourse, and how apparent
counter-evidence can actually broaden our understanding of
lexicalization and the variable weight of ideas. I will mention also
the discourse basis of identifiability or "definiteness". Turning to
the discourse role of grammatical subjects, I will illustrate
ramifications of a hypothesized "light subject" constraint. Finally, I
will discuss and illustrate ways in which prosody expresses emotions
and attitudes, including humor.
The talk will build on ideas set forth in my book Discourse,
Consciousness, and Time (The University of Chicago Press, 1994),
especially chapters 5-11, although chapters 2 and 3 will also be
relevant, and it wouldn't hurt to read the entire book.
____________
STATISTICS SEMINAR
on Tuesday, 9 May 2000, 4:15pm
Sequoia Hall 200
http://www-stat.stanford.edu/seminars/
Estimating Probability of Causation
Judea Pearl
UCLA
According to common judicial standard, judgment in favor of the
plaintiff should be made if and only if it is "more probable than
not" that the defendant's action was "the cause for" the plaintiff's
injury (or death). How can we estimate probabilities of such
hypothetical events as "was the cause for"?
I will propose a formal semantics, based on structural models, for the
probability PN(x,y) that event y would not have occurred if it were
not for event x, given that x and y did in fact occur. Armed with
this semantics, I will then explicate conditions under which PN(x,y)
can be estimated from statistical data, and show how data from both
experimental and nonexperimental studies can be combined to yield
information that neither study alone can provide. Finally, I will
examine when the standard epidemiological criteria (e.g.,
Excess-Risk-Ratio) are adequate for measuring the probability of
causation, and how they can be corrected for confounding bias.
Reference: J.Pearl, Causality, Cambridge University Press, 2000. Also,
Tech Report R-271, http://www.cs.ucla.edu/~judea/
____________
CS548: DISTRIBUTED SYSTEMS RESEARCH SEMINAR
on Wednesday, 10 May 2000, 12:45pm
McCullough 150
http://www.stanford.edu/class/cs548/
Proactive DNS Caching: Addressing a Performance Bottleneck
Edith Cohen
AT&T Labs-Research
The resolution of a host name to an IP-address is a necessary
predecessor to connection establishment and HTTP exchanges between Web
hosts. The Domain Name System, responsible for translating between
hostnames and IP-addresses, is a large distributed database. While its
design accommodated the explosive rise in Web traffic and size of the
domain-name space, DNS resolutions inherently involve communication
with one or more remote name-servers. Hence, when a resolution can not
be processed from the local cache, it often considerably prolongs
user-perceived latency.
DNS lookup times exhibit high variance, even when resolving the same
hostname. For well-connected clients, DNS lookups are a dominant cause
of pathological delays. Hence, they increase the inconsistency in the
quality of Internet-service. Furthermore, DNS lookup time is not
addressed by continuing improvements such as higher-availability Web
sites and increased bandwidth. Thus, as the quality of Internet
service rises, the degrading effect of DNS lookup time becomes even
more noticeable.
We propose enhancements to DNS caching that reduce lookup delays by
increasing local availability of DNS records. Our proposed approaches
operate on top and consistently with DNS, and require only local
deployment.
This is joint work with Haim Kaplan.
____________
BROAD AREA COLLOQUIUM FOR
AI-GEOMETRY-GRAPHICS-ROBOTICS-VISION
on Wednesday, 10 May 2000, 4:00pm
TCseq201 (across from Gates)
http://robotics.stanford.edu/ba-colloquium/
Searching for the Actual Cause
Judea Pearl
UCLA
Many problems in troubleshooting, legal decisions, and natural
language processing require one to identify the "actual cause" of an
event in the context of a given scenario (as in: "Socrates drinking
hemlock was the actual cause of Socrates death.") Following a brief
review of counterfactuals and their structural semantics, I will
propose a formal account of actual causation based on the notion of
"sustenance" -- the capacity of the cause to sustain the effect
despite certain "structural" changes in the model. I will show by
examples how this account avoids problems associated with the
counterfactual-dependence account of Lewis (1986) and how it can be
used both in generating explanations of specific scenarios and in
computing the probabilities that such explanations are in fact
correct.
References
Parts of this talk are based on chapter 10 of CAUSALITY (Cambridge U.
Press, 2000), on my IJCAI-99 Lecture (see
http://www.cs.ucla.edu/~judea/ ) and on recent joint work with J.
Halpern.
Biography: Judea Pearl is a Professor of Computer Science and
Statistics at UCLA where he is the Director of the Cognitive Systems
Laboratory.
He received the B.S. degree in Electrical Engineering from the
Technion, Haifa, Israel, in 1960; a Master degree in Physics from
Rutgers University, New Brunswick, New Jersey, in 1965; and a Ph.D.
degree in Electrical Engineering from the Polytechnic Institute of
Brooklyn, Brooklyn, NY in 1965.
Before coming to UCLA, he worked at RCA Research Laboratories,
Princeton, New Jersey, on superconductive storage devices and magnetic
memory systems. He joined UCLA in 1970, and his current interests
include: knowledge representation, probabilistic and causal reasoning,
constraint processing, nonstandard logics, distributed computation,
and learning .
Pearl has published close to 200 research articles and is the author
of three books: Heuristics (1984), Probabilistic Reasoning in
Intelligent Systems (1988), and Causality: Models, Reasoning and
Inference (2000). He is a Member of the National Academy of
Engineering, a Fellow of the IEEE and the AAAI, and a recipient of the
IJCAI Research Excellence Award in Artificial Intelligence (1999).
____________
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
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/Archive/calendar/current.shtml
and the archives at
http://www-csli.stanford.edu/Archive/calendar/
People on most of the CSLI computers can type 'help csli-calendar' to
see the current issue.
The CSLI Calendar is also posted each week to
news://nntp-csli.stanford.edu/csli.bboard.
and
news://nntp.stanford.edu/su.events
Information about CSLI's research program is available at
http://www-csli.stanford.edu/
For maps to the Stanford University campus see
http://www.stanford.edu/home/visitors/maps.html
____________