[Date Prev][Date Next][Date Index]
CSLI Calendar, Wednesday, 23 April 2003, vol. 18:30
CSLI CALENDAR OF PUBLIC EVENTS
______________________________________________________________________
23 April 2003 Stanford Vol. 18, No. 30
______________________________________________________________________
A weekly publication of the
Center for the Study of Language and Information (CSLI)
Stanford University, Ventura Hall, Stanford, CA 94305-4115
http://www-csli.stanford.edu/
____________
ACTIVITIES FROM 23 APRIL 2003 TO 2 MAY 2003
WEDNESDAY, 23 APRIL 2003
3:00pm Reuters Foundation Digital Vision Program Seminar
email for location mailto:digitalvision@csli.stanford.edu
Beau Takahara
ZeroOne - The Art and Technology Network, Palo Alto
http://reuters.stanford.edu/seminar_speakers.html
Abstract below
3:45pm Psychology Department Colloquium
Jordan Hall 420:041
"Gender Nonconformity and Sexual Orientation"
Michael Bailey
Northwestern University
http://www-psych.stanford.edu/news.html
4:00pm ** CANCELED **
XX UC Berkeley Psychology Colloquium
XX Sue Mineka
XX Northwestern University
XX http://psychology.berkeley.edu/admin/colloquia.html
4:15pm EE380: Computer Systems Lab Colloquium
Gates B03 (NEC Auditorium)
"Real-Time Watermarking System for
Audio Signals Using Perceptual Masking"
Michael Lustig
Electrical Engineering
http://ee380.stanford.edu/contents.html
Abstract below
THURSDAY, 24 APRIL 2003
12 noon RNI/Stanford Seminar on Theoretical Neuroscience
Arrillaga Alumni Center/Fisher, 326 Galvez St.
"Top-down and Bottom-up Models of Selective Visual Attention"
Christof Koch
California Institute of Technology
http://www.klab.caltech.edu
http://www.rni.org/seminar.html
Abstract below
12:15pm CSLI CogLunch
Ventura Hall, Room 17 * SPECIAL LOCATION *
"A (Slightly) Different Account of Folk Psychology"
Peter Godfrey-Smith
Philosophy, Stanford
http://www-csli.stanford.edu/events/Coglunch/
3:00pm Dialogue Systems Talk
Ventura Hall, room 17
"Multimodal Event Parsing for Intelligent User Interfaces"
Will Fitzgerald
Mathematics and Computer Science, Kalamazoo College
http://henson.kzoo.edu/~wfitzg/index.html
Abstract below
4:00pm Personality Seminar
Jordan Hall 420:100
Title to be announced
Ying Wong, Anda Gershon
Stanford
http://www-psych.stanford.edu/news.html#person_lab
4:00pm PARC Forum
George Pake Auditorium at PARC
"Open Innovation: the New Imperative for Creating
and Profiting from Technology"
Henry Chesbrough
Harvard Business School
http://www.parc.com/forum/
Abstract below
4:00pm ** CANCELED **
XX UC Berkeley CIS Seminar
XX Brian Wandell
XX http://www.cs.berkeley.edu/~eyal/cis-seminar
4:15pm SSP10: Symbolic Systems Forum
Bldg. 380:380C (Math Corner)
"Bigger than Chaos: Understanding Complex Systems
Using Probability"
Michael Strevens
Philosophy Department
http://symsys.stanford.edu:8081/ssp-dynamic/servlet/ssp_events
Abstract below
4:15pm US-Asia Technology Management Seminar
Terman Auditorium
"The Era of the First Foray of Dense Parallel Optical
Communication into the Commercial Arena - a Technology
Perspective"
Keith Goossen
Electrical & Computer Engineering
University of Delaware
http://asia.stanford.edu/events/Spring03/index.html
Abstract below
5:00pm School of Engineering Graduate Research Expo
Arrillaga Alumni Center, McCaw Hall
Contact Kate Barry <kate.barry@stanford.edu>
or Marge Kastner <marge.kastner@stanford.edu> for information.
Registration is required at:
http://soe.stanford.edu/alumni/Grad_Expo_Attendees.html
FRIDAY, 25 APRIL 2003
11:00am Berkeley Institute of Cognitive and Brain Seminar
Tolman 5101 (Berkeley)
"From transient patterns to persistent structures:
Episodic memory formation via cortico - hippocampal interactions"
Lokendra Shastri
International Computer Science Institute, Berkeley
http://socrates.berkeley.edu:4247/seminar.html
12 noon Logical Methods in the Humanities
Bldg. 380:383N (math corner)
"Is the Euclidean algorithm optimal among its peers?"
Yiannis N. Moschovakis
UCLA and University of Athens
http://www-logic.stanford.edu/seminars.html
Abstract below
12:30pm CS547: Human-Computer Interaction Seminar
Gates B01
"Croquet: A Collaboration Architecture"
Alan Kay
HP Labs
http://www-pcd.stanford.edu/cs547/
Abstract below
12:30pm UC Berkeley Database Seminar
606 Soda Hall, Berkeley
"The BINGO! System for Expert Web Search and Information
Portal Generation"
Gerhard Weikum
University of the Saarland, Saarbruecken, Germany
http://buffy.eecs.berkeley.edu/Seminars/
2:15pm NLP Reading Group
Wallenberg Hall, Room 323
"The Role of Syntagmatic Prototypes in Mapping Meaning
onto Use"
Patrick Hanks
Brandeis University
http://www-nlp.stanford.edu/nlp/nlpgroup.html
Abstract below
3:30pm Special University Oral Examination
Packard 101
"Block-Type-Decodable Encoders for Constrained Systems"
Panu Chaichanavong
Information Systems Laboratory, Electrical Engineering
Abstract below
MONDAY, 28 APRIL 2003
4:15pm Broad Area Colloquium in AI,
Geometry, Graphics, Robotics, and Vision
TCSeq 200
"Processing Natural Language without Natural
Language Processing"
Eric Brill
Microsoft
http://robotics.stanford.edu/ba-colloquium/
Abstract below
TUESDAY, 29 APRIL 2003
2:45pm CS548: Internet and Distributed Systems Seminar
Gates B03
"What Middleware Administrators Do"
Paul Maglio
IBM Almaden Research Center
http://cs548.stanford.edu/schedule.shtml
4:15pm Engineering 200: Research Universities: Stanford, A Case Study
Jordan 420:040
"Research Policies, Human Subjects, Intellectual Property"
Charles Kruger/Kathy Ku
http://www.stanford.edu/dept/DoR/101.html
4:15pm Networking Lecture
Packard 101
"Capacity of Wireless Networks and Design Principles"
Stavros Toumpis
Abstract below
WEDNESDAY, 30 APRIL 2003
12 noon Psychology Developmental Brownbags
Bldg. 380:381U
"The Development of Career Plans"
John Krumboltz
School of Education
http://www-psych.stanford.edu/news.html#dev_brownbag
12 noon UC Berkeley Institute of Personality and Social Research Colloquium
5101 Tolman (Berkeley)
Title to be announced
Margaret Kemeny
Psychiatry, UC San Francisco
http://psychology.berkeley.edu/admin/colloquia.html
3:00pm Reuters Foundation Digital Vision Program Seminar
email for location mailto:digitalvision@csli.stanford.edu
Bruno Lanvin and Samia Melhem
The World Bank
http://reuters.stanford.edu/seminar_speakers.html
Abstract below
3:45pm Psychology Department Colloquium
Jordan Hall 420:041
"The Rational and Social Paths to Creating Madness in Normal
People"
Phil Zimbardo,
Stanford
http://www-psych.stanford.edu/news.html
4:15pm EE380: Computer Systems Lab Colloquium
Gates B03 (NEC Auditorium)
"Internet Security: An Optimist Gropes for Hope"
Bill Cheswick
Lumeta Internet Security:
http://ee380.stanford.edu/contents.html
Abstract below
THURSDAY, 1 MAY 2003
4:00pm Personality Seminar
Jordan Hall 420:100
Title to be announced
Becky Ray
Stanford
http://www-psych.stanford.edu/news.html#person_lab
4:00pm UC Berkeley CIS Seminar
Soda Hall 380 (UC Berkeley)
Title to be announced
Richard Voyles
University of Minnesota
http://www.cs.berkeley.edu/~eyal/cis-seminar
4:15pm SSP10: Symbolic Systems Forum
Bldg. 380:380C (Math Corner)
"Learning to listen for meaning:
Spoken Language Processing in the Second Year of Life"
Anne Fernald
Psychology Department
http://symsys.stanford.edu:8081/ssp-dynamic/servlet/ssp_events
4:15pm US-Asia Technology Management Seminar
Terman Auditorium
"Silicon-based Light Emitting Devices for Optical Links"
Communication into the Commercial Arena - a Technology
Perspective"
Don Gardner
Senior Research Engineer, Microprocessor Research Lab
Intel Corporation
http://asia.stanford.edu/events/Spring03/index.html
Abstract below
FRIDAY, 2 MAY 2003
11:00am Berkeley Institute of Cognitive and Brain Seminar
Tolman 5101 (Berkeley)
Title to be announced
Ed Munnich
Graduate School of Education, UC Berkeley
http://socrates.berkeley.edu:4247/seminar.html
12:30pm CS547: Human-Computer Interaction Seminar
Gates B01
Title to be announced
Will Wright
Maxis
http://www-pcd.stanford.edu/cs547/
3:15pm Philosophy Department Colloquium
Bldg. 90:92Q
Title to be announced
Jim Woodward
CalTech
http://www-philosophy.stanford.edu/ce.html
____________
Stanford Blood Center status: Critical shortage of O+; shortage of O-,
B-, and B+. For an appointment: http://bloodcenter.stanford.edu/ or
call 650-723-7831. It only takes an hour of your time.
____________
REUTERS FOUNDATION DIGITAL VISION PROGRAM SEMINAR
on Wednesday, 23 April 2003, 3:00pm - 4:30pm
email for location mailto:digitalvision@csli.stanford.edu
http://reuters.stanford.edu/seminar_speakers.html
Beau Takahara
Director and CEO
ZeroOne - The Art and Technology Network, Palo Alto
Beau comes to the Network after five years at The Tech Museum of
Innovation in San Jose where she worked as a fund-raiser on the Capital
Campaign and the Annual Fund. While at The Tech, Beau curated an art
and technology lecture series and arranged for two art and technology
exhibits, one from Interval Research and one from Xerox PARC, each to
be installed for a six-month period in the new Tech. Prior to her
time at The Tech, Beau worked with a number of not-for-profit art
organizations including Yerba Buena Center for the Arts, San
Francisco; and George Coates Performance Works, San Francisco, where
she was executive director for three years and involved in
large-scale, leading-edge productions of new music theater utilizing
state-of-the-art technologies. During her nine-year tenure at the San
Francisco Museum of Modern Art, she curated video and media exhibits,
produced a large computer graphics festival in conjunction with
SIGGRAPH, and organized the first museum screening of HDTV, as part of
a large seven- week festival: Tokyo Form and Spirit, which she
conceived of and produced. Beau is past Board Chair of The Lab, an
alternative arts organization in San Francisco that presents
experimental work by visual and performance artists. She received a
BFA in art and art history from The University of Utah and did
graduate studies in art history at the University of Iowa.
____________
EE380: COMPUTER SYSTEMS LAB COLLOQUIUM
on Wednesday, 23 April 2003, 4:15pm - 5:30pm
NEC Auditorium (B03), Gates Computer Science Building
http://ee380.stanford.edu/contents.html
"Real-Time Watermarking System for
Audio Signals Using Perceptual Masking"
Michael Lustig
Electrical Engineering
mlustig@stanford.edu
This work was the winner of Texas Instruments DSP and Analog Challenge
worldwide competition, receiving a prize of 100,000$.
Recent development in the field of digital media raises the issue of
copyright protection. Digital watermarking offers a solution to
copyright violation problems. The watermark is a signature, embedded
within the data of the original signal, which in addition to being
inaudible to the human ear, should also be statistically undetectable,
and resistant to any attempts to remove it. In addition, the
watermark should be able to resolve multiple ownership claims (known
as the deadlock problem), which is achieved by using the original
signal (i.e., the unsigned signal) in the signature detection process.
In order to meet the above demands, a frequency-masking scheme using a
psycho-acoustic model is used to ensure a maximal, yet inaudible,
additive signature. This scheme was implemented on the Texas
Instruments TMS320C5410 DSP, and achieves real-time capabilities.
The talk will have two parts. In the first part Michael will talk
about digital audio watermarking, the real-time DSP implementation and
the system design considerations. There will be a limited yet
enjoying demo. The second part is about wining the DSP Challenge, he
will try to share some of the moments from the competition. And
finally, some tips on what can be done with a 100,000$.
This is a joint work with Yuval Cassuto and Shay Mizrachy. The
project was done in the Signal & Image processing lab, Technion IIT
About the speaker: Michael is a EE graduate student at Stanford
University. Prior to his studies Michael worked for Gigami
(Net2Wireless) trying to compress images over a cellular channel, then
doing research in tracking algorithms and computer vision for
Prof. Ron Kimmel in the CS department, Technion IIT. Michael holds a
Bsc. In Electrical Engineering from the Technion IIT.
____________
RNI/STANFORD SEMINAR SERIES IN THEORETICAL NEUROSCIENCE
on Thursday, 24 April 2003, 12 noon - 1pm
Arrillaga Alumni Center/Fisher, 326 Galvez St.
http://www.rni.org/seminar.html
"Top-down and Bottom-up Models of Selective Visual Attention"
Christof Koch
Division of Biology & Division of Engineering and Applied Science
California Institute of Technology
http://www.klab.caltech.edu
Although brains possess a paradigmatically massively parallel
architecture, sensory systems employs a serial computational strategy
to select "interesting" objects in any scene for further processing,
including access to short-term memory, planning and awareness. In
the visual system of human and monkeys, selective visual attention is
guided by a rapid, task-independent, stimulus-driven saliency-based
form of selection process as well as by a slower, volitional
controlled top-down selection process. I will describe two detailed
computational accounts of both selection mechanisms and discuss their
implications for the underlying neuronal processes in the primate's
visual system.
About the Speaker: Christof Koch obtained his Ph.D. in Tuebingen,
Germany in 1982 in (bio)-physics with a minor in Philosophy. After
a four year sojourn at MIT, he joined the California Institute of
Technology, where he is the Lois and Victor Troendle Professor of
Cognitive and Behavioral Biology. His laboratory focuses on
experimental and computational research pertaining to the biophysics
and neurophysiology of neurons, and the neuronal correlates of
selective visual attention, awareness and consciousness in the
mammalian brain. With Dr. Francis Crick at the Salk Institute, he has
been working on discovering the neuronal correlates of consciousness
in the primate brain.
____________
DIALOGUE SYSTEMS TALK
Ventura Hall, room 17
on Thursday, 24 April 2003, 3pm
"Multimodal Event Parsing for Intelligent User Interfaces"
Will Fitzgerald
Mathematics and Computer Science, Kalamazoo College
http://henson.kzoo.edu/~wfitzg/index.html
Many intelligent interfaces must recognize patterns of user activity
that cross a variety of different input channels. These multimodal
interfaces offer significant challenges to both the designer and the
software engineer. The designer needs a method of expressing
interaction patterns that has the power to capture real use cases and
a clear semantics. The software engineer needs a processing model
that can identify the described interaction patterns efficiently while
maintaining meaningful intermediate state to aid in debugging and
system maintenance. In this talk, I will describe an input model, a
general recognition model, and a series of important classes of
recognition parsers with useful computational characteristics; that
is, we can say with some certainty how efficient the recognizers will
be, and the kind of patterns the recognizers will accept. Examples
illustrate the ability of these recognizers to integrate information
from multiple channels across varying time intervals.
About the Speaker: Will Fitzgerald obtained his Ph.D. in computer
science at Northwestern University in 1994. Until 2002 he expanded
his work on embedded conceptual parsers into multimodal dialogue
systems as VP for research of Intell/Agent Systems and Neodesic
Corporation and as CTO of I/NET. Since joining Kalamazoo College,
his research includes multimodal event detection and interpretation.
____________
PARC FORUM
on Thursday, 24 April 2003, 4:00pm - 5:00pm
George Pake Auditorium, PARC
http://www.parc.com/forum/
"Open Innovation: the New Imperative for Creating
and Profiting from Technology"
Henry Chesbrough
Harvard Business School
While the key to successful innovation once lay in the controlled
environment of the corporate laboratory, today the widespread
distribution of useful knowledge makes such control infeasible.
Competitive advantage now often comes from leveraging the discoveries
of others.
Rather than relying entirely on internal ideas to advance the
business, an "open" approach to innovation leverages internal and
external sources of ideas. Rather than restricting innovation to a
single path to market, open innovation inspires companies to find the
most appropriate business model to commercialize a new offering -
whether that model exists within the firm or must be sought through
external licensing, partnering or venturing.
About the speaker: Henry Chesbrough is an assistant professor of
business administration at the Harvard Business School, teaching the
management of technology. This fall he will join the faculty of the
Haas Business School at UC-Berkeley. He has consulted to leading
companies in both the U.S. and Japan on issues of technology
management and technology venturing and is the author of numerous
papers, including an article on Xerox's management of its technology
spinoff organizations in the Winter 2002 issue of Business History
Review. His academic research has been published in Research Policy,
Industrial and Corporate Change, Business History Review, and the
Journal of Evolutionary Economics. He has also published managerial
articles in the Harvard Business Review, California Management Review,
and Sloan Management Review.
____________
SSP10: SYMBOLIC SYSTEMS FORUM
on Thursday, 24 April 2003, 4:15pm
Bldg. 380:380C (Math Corner)
http://symsys.stanford.edu:8081/ssp-dynamic/servlet/ssp_events
"Bigger than Chaos: Understanding Complex Systems Using Probability"
Michael Strevens
Philosophy, Stanford
Many complex systems behave in surprisingly simple ways. In spite of
the of complexity of animal behavior, for example, most ecosystems
exhibit a fairly simple and stable behavior at the population level.
By examining the foundations of statistical reasoning about complex
systems such as gases, ecosystems, and certain social systems, I aim
to provide an understanding of how macrolevel simplicity emerges from
microlevel complexity. The talk will outline the main arguments and
methods of my forthcoming book on this topic.
____________
US-ASIA TECHNOLOGY MANAGEMENT CENTER PUBLIC LECTURE SERIES
on Thursday, 24 April 2003, 4:15pm - 5:30pm
Terman Auditorium
http://asia.stanford.edu/events/Spring03/index.html
"The Era of the First Foray of Dense Parallel Optical Communication
into the Commercial Arena - a Technology Perspective"
Keith Goossen
Electrical & Computer Engineering
University of Delaware
Dense parallel optics, more specifically 2 dimensional arrays of light
communication, even more specifically the natural technological format
of such as a chip-on-chip integrated array of photonics on a VLSI
chip, has in the last few years gone private, so to speak. Three
companies, Xanoptix, Teraconnect, and Aralight, were formed to
commercialize this technology.
The chosen product in each case was a higher channel count E/O
converter than currently available twelve channel versions, which are
based on a linear array of non-integrated photonics. The technology
and issues with this application will be discussed.
About the speaker: Keith Goossen, Associate Professor, University of
Delaware, Department of Electrical and Computer Engineering, has 14
years professional experience in the fields of optoelectronics and
optical fiber communication, resulting in over a hundred refereed
journal papers, numerous invited talks and conference proceedings, and
67 patents.
____________
LOGICAL METHODS IN THE HUMANITIES
on Friday, 25 April 2003, 12 noon - 1:15 pm
Math Corner 380:383N
http://www-logic.stanford.edu/seminars.html
"Is the Euclidean algorithm optimal among its peers?"
Yiannis N. Moschovakis
UCLA and University of Athens
In the first (main) part of this talk, I will describe recent joint
work with Lou van den Dries in arithmetic complexity, including the
following result:
If alpha is any recursive program from +, -, and division with
remainder which decides co-primeness for pairs of numbers, and if
c_alpha(a,b) is the (strict, parallel, time-) complexity of alpha,
then for all co-prime a,b ,
|a/b-sqrt(2)| < 1/b^2 --> c_alpha(a,b) >= 8 log log b
It follows that any recursive program which computes the greatest
common divisor gcd(a,b) from +, -, and division with remainder
requires at least 8 log log b steps on infinitely many inputs,
which is one log away from a natural formulation of the optimality
of the Euclidean algorithm; this remains open.
Here, recursive programs from given operations are relativized,
McCarthy-style recursive definitions which ``spend'' one time unit to
call a given operation. In the second part of the talk, I will argue
(briefly) that the proofs of these results support the proposal to
define first order algorithms using their natural expression by
relativized McCarthy-style recursive programs.
____________
CS547: HUMAN-COMPUTER INTERACTION SEMINAR
on Friday, 25 April 2003, 12:30 - 2:00pm
Gates B01
http://www-pcd.stanford.edu/cs547/
"Croquet: A Collaboration Architecture"
Alan Kay
HP Labs
Alan.Kay@squeakland.org
http://www.opencroquet.org
Croquet is a computer software architecture built from the ground up
with a focus on deep collaboration between teams of users. It is a
totally open, totally free, highly portable extension to the Squeak
programming system, a modern variant of Smalltalk. Croquet is a
complete development and delivery platform for doing real
collaborative work. There is no distinction between the user
environment and the development environment.
Croquet is also a totally ad hoc multi-user network. It mirrors the
current incarnation of the World Wide Web in many ways, in that any
user has the ability to create and modify a "home world" and create
links to any other such world. But in addition, any user, or group of
users (assuming appropriate sharing privileges), can visit and work
inside any other world on the net. Just as the World Wide Web has
links between the web pages, Croquet allows fully dynamic connections
between worlds via spatial portals. The key differences are that
Croquet is a fully dynamic environment, everything is a collaborative
object, and Croquet is fully modifiable at all times.
Croquet is a joint project being developed by David A. Smith, Alan
Kay, David P. Reed, and Andreas Raab. More information is available
at: http://www.opencroquet.org
About the speaker: Alan Kay is one of the pioneers of personal
computing. In 1966 he helped invent "object-oriented programming" In
1967-9 he and Ed Cheadle invented the FLEX Machine, a very early
modern desktop machine they called a "personal computer" which led to
his design of the Dynabook, "a personal computer for children of all
ages," in the form of a very portable notebook, with a flat-screen,
stylus, wireless network, and local storage. At Xerox PARC in the 70s
he invented Smalltalk, which was the first complete dynamic object
oriented language, development, and operating system, and was one of
the instigators for the first bitmap displays and the main inventor of
the now ubiquitous overlapping windows, icons, point-click-and-drag
user interface.
____________
NLP READING GROUP
on Friday, 25 April 2003, 2:15pm - 3:30pm
Wallenberg Hall, Room 323
http://www-nlp.stanford.edu/nlp/nlpgroup.html
"The Role of Syntagmatic Prototypes in Mapping Meaning onto Use"
Patrick Hanks
Brandeis University
Recent work in corpus analysis shows that everyday usage is both
highly patterned -- more so than was once thought -- and highly
creative. Moreover, the patterns are not always as we expect them to
be. Distinguishing uses that conform to patterns -- syntagmatic norms
-- from uses that represent creative exploitations of those patterns
is a necessary preliminary for text understanding. It is, moreover, a
necessary preliminary for procedures such as meaning coercion. (The
meaning of an abnormal sentence cannot be coerced if we do not know
the normal uses of at least some of the words involved.)
It is now over a quarter of a century since Fillmore proposed that
meanings should be computed in terms of resemblance to a prototype
rather than satisfaction of necessary conditions, and yet no
satisfactory inventory of prototypes exists. At the same time "word
sense disambiguation" is perceived as a continuing problem for NLP.
Current work on a "Context Dictionary of English Verbs" shows that
semantically motivated syntagmatic norms can be teased out of corpus
evidence, but that identifying relevant contextual elements that
distinguish one meaning from another is a non-trivial task. There are
many surprises. The relationship between the context dictionary and
other projects such as Framenet is discussed.
____________
SPECIAL UNIVERSITY ORAL EXAMINATION
on Friday, 25 April 2003, 3:30pm
Packard 101
"Block-Type-Decodable Encoders for Constrained Systems"
Panu Chaichanavong
Information Systems Laboratory, Electrical Engineering
In most recording channels, arbitrary user data is encoded into
constrained sequences to improve performance of storage systems. In
order to avoid error propagation in the decoding process, many
practical applications use block encoders. Although they are
conceptually simplest, we may be able to achieve higher rates using
block decodable finite-state encoders and error propagation is still
limited to one block. However, the optimal rate is difficult to
compute and an achieving block decodable encoder is hard to design.
Nevertheless, for some constraints including runlength limited
RLL(d,k) and maximum transition run MTR(j,k) constraints, this problem
turns out to be equivalent to designing a so-called deterministic
encoder, which is much more tractable. In this talk, we shall focus
on these three classes of encoders which we call block-type-decodable
encoders.
It is known that the construction of a block-type decodable encoder is
equivalent to choosing a subset of states in the graph, called a set
of principal states. Thus our goal is to choose an optimal set of
principal states that yields the highest code rate. In many
circumstances, an optimal set of principal states at asymptotically
large block length is easier to find than one at finite block length.
We shall demonstrate how the asymptotic case helps the computation of
the finite case. An asymmetric runlength-limited constraint is used
as an example. The complexity of block-type-decodable encoder design
is presented.
Finally, we discuss some recent results on bounded-delay-encodable
block-decodable (BDB) encoders. These are encoders that are based on
look-ahead encoders rather than finite-state encoders.
Refreshments at 3:15pm.
____________
BROAD AREA COLLOQUIUM FOR
AI-GEOMETRY-GRAPHICS-VISION-ROBOTICS
on Monday, 28 April 2003, 4:15pm
TCSeq 200
http://robotics.stanford.edu/ba-colloquium/
"Processing Natural Language without Natural Language Processing"
Eric Brill
Microsoft
Despite decades of research and development, we can still only create
machines with the most rudimentary natural language processing
capabilities. One of the greatest barriers to advanced natural
language processing is our inability to overcome the linguistic
knowledge acquisition bottleneck. Language appears to be extremely
complex and idiosyncratic. Over the years, there has been an ongoing
debate as to how best to overcome this bottleneck: via better
linguistics or more powerful machine learning. While we have been
debating, the amount of on-line text has ballooned from the ubiquitous
million-word Brown corpus to close to a trillion words accessible on
the Web. Does this change everything? We will describe recent work
in a number of areas, including automatic question answering,
automatic training of grammar checkers, and language modeling, where
state of the art accuracy is achieved using very simple methods whose
power comes entirely from the plethora of text currently available to
these systems.
About the speaker: Eric Brill is a Senior Researcher in the Machine
Learning and Applied Statistics Group, of Microsoft Research. His
research interests include machine learning, string algorithms,
natural language processing and information retrieval.
____________
NETWORKING LECTURE
on Tuesday, 29 April 2003, 4:15pm
Packard 101
"Capacity of Wireless Networks and Design Principles"
Stavros Toumpis
We define and study capacity regions for wireless ad hoc networks with
an arbitrary number of nodes and topology. These regions describe the
set of achievable rate combinations between all source-destination
pairs in the network under various transmission strategies, such as
variable-rate transmission, single hop or multihop routing, power
control and successive interference cancellation. Multihop cellular
networks and networks with energy constraints are studied as special
cases. With slight modifications, the developed formulation can
handle node mobility and time-varying flat fading channels.
Using this framework, we then concentrate on the joint-layer design of
wireless ad hoc networks. We show that using single routes for each
data stream might significantly reduce the capacity of the network.
We then study CSMA/CA, and find that its performance strongly depends
on the choice of the accompanying routing protocol. We introduce two
protocols that outperform CSMA/CA, both in terms of energy efficiency
and achievable throughput. The *Progressive Back Off Algorithm
(PBOA)* performs medium access jointly with power control. The
*Progressive Ramp Up Algorithm* sacrifices energy efficiency in favor
of higher throughput. Both protocols slot time, and are integrated
with queuing disciplines that are more relaxed than the First In First
Out (FIFO) rule. They are totally distributed and the overhead they
require does not increase with the size and node density of the
network.
We conclude with a brief discussion on the capacity of ad hoc networks
with an asymptotically large number of nodes. We establish a lower
bound on the aggregate throughput as a function of the number of
nodes, under a general model for fading, and assuming the nodes to be
immobile. We then assume node mobility, showing a fundamental
trade-off between the packet delay and the aggregate throughput.
Refreshments at 4:00.
____________
REUTERS FOUNDATION DIGITAL VISION PROGRAM SEMINAR
on Wednesday, 30 April 2003, 3:00pm - 4:30pm
email for location mailto:digitalvision@csli.stanford.edu
http://reuters.stanford.edu/seminar_speakers.html
Bruno Lanvin and Samia Melhem
The World Bank
At the World Bank, Bruno Lanvin is the Manager of the Information for
Development Program (infoDev), a multi-donor program focusing on
extending digital opportunities for all (www.infodev.org); so far,
infoDev has financed over 170 projects in some 85 countries around the
world.
In 2000, Dr. Lanvin was appointed Executive Secretary of DOT Force,
the G-8 initiative launched by the Okinawa Summit of July 2000 to
bridge the Digital Divide (www.dotforce.org). He was recruited by the
World Bank in September 2000 as senior advisor for e-commerce and
e-government. Until then, he was Head of Electronic Commerce in the
United Nations Conference on Trade and Development (UNCTAD) in Geneva.
During his twenty years with the United Nations, he has occupied
various senior positions including Chief of Cabinet of the Director
General of the United Nations in New-York, Head of Strategic Planning
and later Chief of the SME Trade Competitiveness Unit of UNCTAD/SITE.
He has been the Deputy Executive Secretary of UNISTE (UN International
Symposium on Trade Efficiency) in 1994 and General Manager of GETUP
(Global Electronic Trade UN Partnership) in 1998. From 1994 to 1998,
he was the World Coordinator of the United Nations Trade Point
Programme, a global e-commerce network operating in over 100
countries. In 1999 and 2000, he managed UNCTAD's component of the UN
'Development Account', devoted to e-commerce and development, and
organized a series of regional and interregional workshops on
e-commerce, allowing more than 2000 representatives from governments
and enterprises to exchange experiences and best practices in the area
of e-commerce.
A frequent keynote speaker and participant in international
conferences on the `New Economy,' he has published a large number of
articles and books on ICT and development. He was the main drafter,
team leader, and editor of 'Building Confidence: electronic commerce
and development', published in January 2000. In 2003, he co-authored
the Global Information Technology Report, 2002 and 2003 (INSEAD, The
World Economic Forum, infoDev).
Bruno Lanvin has worked in more than 60 countries. He holds a BA in
Mathematics and Physics from the University of Valenciennes (France),
an MBA from Ecole des Hautes Etudes Commerciales (HEC) in Paris, and a
PhD in Economics from the University of Paris I (La Sorbonne) in
France. He speaks and writes French (mother tongue), English, Spanish
and has a working knowledge of Italian, Portuguese, Russian, and some
Chinese.
In her capacity at the World Bank, Samia Melhem coordinates the
regional projects for the Middle East and North Africa (MNA) region in
Telecoms Reform and private sector development for the ICT sector.
She also has been working on ICT in development projects, putting
together with governments national E-readiness strategies where ICT is
used as a tool to improve literacy, knowledge, and employment, and
ultimately to close the gap of the digital divide.
Her expertise is mostly in planning, developing, and implementing
large-scale information systems for governments, and for education
agencies with institutional strengthening, capacity-building, and
change management programs. She managed a large tax computerization
and customs automation project in the Philippines and several distance
education projects in MNA and sub-Saharan Africa.
She also manages the ICT training program of the World Bank's GICT
department, by providing courses on telecoms, software development,
media, ecommerce, team building, and coaching.
She held several other positions at the WBG, which she joined in 1988.
She holds degrees in electrical engineering (BSEE), computer sciences
(MS), and international business (MBA).
____________
EE380: COMPUTER SYSTEMS LAB COLLOQUIUM
on Wednesday, 30 April 2003, 4:15pm - 5:30pm
NEC Auditorium (B03), Gates Computer Science Building
http://ee380.stanford.edu/contents.html
"Internet Security: an Optimist Gropes for Hope"
Bill Cheswick
Chief Scientist, Lumeta Corp.
ches@cheswick.com
By all accounts the Internet has grown more dangerous since its
inception. Most of the expected attacks have appeared and become
commonplace. Increasingly sophisticated malware has learned to hide
in the deep bushes of verdant, wild software. Users can't keep up
with these dangers, and it is hard enough for the professionals.
And yet, there are indications that things can get better. Many
important web sites get security right enough to support large
business models. Those who run our most secure networks report that
they repeatedly pass the pop-quizzes of the attacks du jour. We can
use crypto when we want to, and many do.
We can do better, and many of us are starting to.
About the speaker: Ches has been out and about in the Internet
security field since the late 1980s. He is known for his early work
in firewalls and proxies, and for the book he has co-authored with
Steve Bellovin and now Avi Rubin. In summer 2000 Ches helped spin off
the Internet cartography he did at Bell Labs with Hal Burch into a
startup, Lumeta Corp, which explores the extent and perimeter hosts of
corporate and government intranets.
____________
US-ASIA TECHNOLOGY MANAGEMENT CENTER PUBLIC LECTURE SERIES
on Thursday, 1 May 2003, 4:15pm - 5:30pm
Terman Auditorium
http://asia.stanford.edu/events/Spring03/index.html
"Silicon-based Light Emitting Devices for Optical Links"
Don Gardner
Senior Research Engineer, Microprocessor Research Lab, Intel Corp.
Recently, there has been increasing interest in light emission from
silicon for optoelectronic circuits and chip-to-chip interconnections.
In this work, the requirements for creating a chip-to-chip optical
link were determined. Different silicon-based devices were studied
for their potential use in an optical link including silicon diodes
and waveguides with silicon nanocrystals and rare earth ions. The
luminescence efficiency and performance of silicon diodes fabricated
using both standard CMOS and solar-cell process technologies were
analyzed. Photoluminescence (PL) and electroluminescence were
measured from both implanted and gas-source diffused devices. Peak PL
intensity from implanted-annealed junctions is approximately an order
of magnitude lower than for unprocessed silicon. Silicon
light-emission efficiency was improved by reducing the surface/contact
recombination, and bulk recombination (Auger and Shockley-Read-Hall).
This can be done by using gas-source diffusion, passivating and
texturing the surfaces, and depositing anti-reflective coatings.
The challenge is that the radiative recombination lifetime is long
because silicon is an indirect bandgap material and carriers are
removed by slow spontaneous emission and diffusion rather than
stimulated emission and drift. The result is slow switching speeds.
In addition, the light is incoherent preventing its use with many
high-speed refractive or phase modulators, photonic crystals, etc.
Conversely, waveguides with silicon nanocrystals and optical or
preferably electrical pumping show promise as a source of coherent
light. Modeling parameters for the optical power and optimal
wavelength required for an optical link were determined by the
photodetector, transimpedance amplifier, coupling losses, noise, and
frequency of operation. The prospects for creating a silicon-based
optical link look promising.
____________
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 (i.e., no abstracts). 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://news.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
____________