[Date Prev][Date Next][Date Index]
CSLI Calendar, Wednesday, 22 November 2006, vol. 22:12
CSLI CALENDAR OF PUBLIC EVENTS
______________________________________________________________________
22 November 2006 Stanford Vol. 22, No. 12
______________________________________________________________________
A weekly publication of the
Center for the Study of Language and Information (CSLI)
Stanford University, Cordura Hall, Stanford, CA 94305-4101
http://www-csli.stanford.edu/
____________
ACTIVITIES FROM 22 NOVEMBER 2006 TO 1 DECEMBER 2006
WEDNESDAY, 22 NOVEMBER 2006
THURSDAY, 23 NOVEMBER 2006 - University Holiday
FRIDAY, 24 NOVEMBER 2006 - University Holiday
MONDAY, 27 NOVEMBER 2006
12 noon UC Berkeley Psychology Job Talk [27-Nov-06]
3105 Tolman (Berkeley)
"Social Neuroscience - to be announced"
James Morris
Duke-UNC Brain Imaging and Analysis Center
http://psychology.berkeley.edu/news/colloquia.html
3:30pm Social Lab [27-Nov-06]
Bldg. 200:105
"Social Psychology in the the Media"
Hazel Markus
Stanford University
http://www-psych.stanford.edu/events_social.html
4:00pm UC Berkeley Linguistics Department Colloquium [27-Nov-06]
182 Dwinelle (Berkeley)
"The origins of vowel length neutralization patterns"
Scott Myers
University of Texas, Austin
http://linguistics.berkeley.edu/events/
Abstract below
4:00pm Berkeley Intel Research Seminar [27-Nov-06]
Intel Research Berkeley, 2150 Shattuck, Ste. 1300
"How Simply and Understandably Could The 'Personal Computing'"
Alan Kay
Viewpoints Research Lab
http://www.intel-research.net/berkeley/Seminars.asp
Abstract below
4:15pm CS528: Broad Area Colloquium [27-Nov-06]
Hewlett Teaching Center 200
Title to be announced
Frank Dellaert
Computing, Georgia Tech
http://www.cc.gatech.edu/~dellaert/
http://graphics.stanford.edu/courses/cs528/
TUESDAY, 28 NOVEMBER 2006
12 noon Berkeley Redwood Center for Theoretical Neuroscience [28-Nov-06]
3105 Tolman (Berkeley)
"Learning Invariant Features Using Inertial Priors, or
'Why Google might want to be in the neocortex business?'"
Thomas Dean
Brown University / Google
http://redwood.berkeley.edu/seminars.php
Abstract below
3:30pm SRI CSL Seminar Series [28-Nov-06]
EK255, SRI International
"Peptide discover:
From mass spectrometry to bioinformatics to function"
Lloyd D. Fricker
Pharmacology, Albert Einstein College of Medicine, Bronx, NY
http://www.csl.sri.com/
Abstract below
4:00pm Berkeley Future of Scholarly Communication Seminar [28-Nov-06]
CSHE Library, South Hall Annex (Berkeley)
"Academic Amnesia: Who Is Preserving Our Data?"
Abby Smith
Independent Consultant
http://cshe.berkeley.edu/events/
Abstract below
4:00pm UC Berkeley Working Group in the Philosophy of Mind [28-Nov-06]
location unknown
Title to be announced
Tom Griffiths
Psychology, Berkeley
http://psychology.berkeley.edu/faculty/profiles/tgriffiths.html
http://neurophilosophy.berkeley.edu/meetings.htm
4:00pm SRI AI Seminar Series [28-Nov-06]
EJ228, SRI International
"Data Delivery in a Service-Oriented World:
The BEA AquaLogic Data Services Platform"
Michael Carey
BEA Systems, Inc.
http://www.ai.sri.com/seminars/
Abstract below
6:45pm SULUG Meeting [28-Nov-06]
Gates 104
"Starting a technology based company"
special speaker
http://sulug.stanford.edu/
WEDNESDAY, 29 NOVEMBER 2006
12 noon Psychology Developmental Brownbags [29-Nov-06]
Jordan Hall 420:102
Title to be announced
Anne Fernald
Stanford
http://www-psych.stanford.edu/events_developmental.html
12 noon UC Berkeley CITRIS Research Exchange [29-Nov-06]
290 Hearst Memorial Mining Building (UC Berkeley)
"Discovering design principles in bacteria and using them to
design tumor-killing bacterium"
Adam Arkin
Bioengineering, UC Berkeley
http://www.lbl.gov/~aparkin/
http://www.citris-uc.org/
12 noon UC Berkeley IPSR colloquium [29-Nov-06]
5101 Tolman Hall (Berkeley)
"Power, Cognition, and Behavior"
Ana Guinote
University of Kent, UK
http://ls.berkeley.edu/dept/ipsr/colloquium.htm
4:15pm EE380: Computer Systems Lab Colloquium [29-Nov-06]
Gates B03 (NEC Auditorium)
"NVIDIA"
Ian Buck
NVIDIA
http://www.nvidia.com/
http://ee380.stanford.edu/contents.html
6:00pm Silicon Valley Web Guild [29-Nov-06]
Google (900 Alta Avenue, Mountain View)
"SEO for Web 2.0"
Panel
http://www.webguild.org/
(Registration and fee)
Information below
6:00pm Berkeley History and Philosophy of Logic Mathematics, and Science
234 Moses (Berkeley) [29-Nov-06]
"The Future of Proof"
Dana Scott
Carnegie Mellon
http://hplms.berkeley.edu/
Abstract below
THURSDAY, 30 NOVEMBER 2006
11:00am SRI AI Seminar Series [30-Nov-06]
EJ228, SRI International
"Machine Learning in Web 2.0: Analyzing Dynamic Content-driven
Social Networks"
Sugato Basu
SRI
http://www.ai.sri.com/seminars/
Abstract below
12 noon CSLI CogLunch [30-Nov-06]
Cordura Hall 100
"The Relationship Between Consciousness and Self-Awareness"
Johannes Brandl
University of Salzburg
http://www-csli.stanford.edu/events/Coglunch/
Abstract below
12:45pm Stanford Networking Seminar [30-Nov-06]
Gates 104
Title to be announced
Nick Feamster
Computing, Georgia Tech
http://www-static.cc.gatech.edu/~feamster/
http://netseminar.stanford.edu/
3:00pm Stanford Software Seminar [30-Nov-06]
Gates 104
"What's Next for Software Verification?"
Rupak Majumdar
UCLA
http://www.cs.ucla.edu/~rupak/
http://theory.stanford.edu/~mhn/sss.html
Abstract below
4:00pm Stanford Phonology Workshop [30-Nov-06]
Margaret Jacks Hall 460:126
"Discussing 'A Maximum Entropy Model of Phonotactics and
Phonotactic Learning' by Bruce Hayes and Colin Wilson"
http://roa.rutgers.edu/files/858-0806/858-HAYES-0-0.PDF
http://www-linguistics.stanford.edu/pinterest/
4:00pm PARC Forum
George Pake Auditorium at PARC
"The race for 21st century fuels"
Alex Farrell
UC Berkeley
http://www.parc.com/forum/
4:00pm Berkeley Center for Studies in Higher Education [30-Nov-06]
175 Barrows Hall (Berkeley)
"Issues in Higher Education Access for English Language Learners:
Thoughts on Pathways through the Master Plan"
Kenji Hakuta and George C. Bunch
Stanford University and UC Santa Cruz
http://cshe.berkeley.edu/events/
Abstract below
4:15pm SSP10: Symbolic Systems Forum [30-Nov-06]
Bldg. 380:380C (Math Corner)
"Selecting Among Paraphrases"
Tom Wasow
Linguistics, Stanford
http://symsys.stanford.edu/ssp_events
Abstract below
4:15pm Information Systems Seminar [30-Nov-06]
Packard 101
Title to be announced
Young-Han Kim
UCSD
http://isl.stanford.edu/colloquium.html
4:15pm Stanford Algorithms Seminar (AFLB) [30-Nov-06]
Gates 498
"Incentive-Based Ranking Mechanisms"
Rajat Bhattacharjee
http://www.stanford.edu/~rajatb/
http://theory.stanford.edu/~aflb/
Abstract below
4:15pm Fundamental Themes in Neuroscience Seminar [30-Nov-06]
Munzer Auditorium, Beckman
"The brain's own cannabis - recent developments"
Brad Alger
Physiology and Psychiatry, University of Maryland
http://neuroscience.umaryland.edu/faculty/default.asp?ID=3
http://nis-seminars.stanford.edu/
Abstract below
FRIDAY, 1 DECEMBER 2006
11:00am SRI AI Seminar Series [1-Dec-06]
EJ228, SRI International
"In-network PCA and anomaly detection"
XuanLong Nguyen
UC Berkeley
http://www.ai.sri.com/seminars/
Abstract below
12:30pm CS547: Human-Computer Interaction Seminar [1-Dec-06]
Gates B01 (HP Auditorium)
"Technology for Developing Regions"
Eric Brewer
UC Berkeley
http://www-pcd.stanford.edu/cs547/
3:00pm Berkeley Information Access Seminar [1-Dec-06]
107 South Hall (Berkeley)
"Math People, a Distributed Name Authority System"
Jim Pitman
Math & Statistics, Berkeley
http://www.sims.berkeley.edu/courses/is296a-1/f06/schedule.html
Abstract below
3:15pm Philosophy Department Colloquium [1-Dec-06]
Bldg. 90:92Q
"Quine"
Peter Hylton
University of Illinois at Chicago
http://www.uic.edu/depts/phil/bios/hylton.htm
http://www-philosophy.stanford.edu/ce.html
3:15pm Friday Cognitive Seminar [1-Dec-06]
Jordan Hall 420:050
"Value or salience? Competing theories of nucleus accumbens function"
Jeff Cooper
http://www-psych.stanford.edu/events_cognitive.html
3:15pm ME394: Design Forum [1-Dec-06]
Terman 556
"my research"
Ken Waldron
http://me.stanford.edu/faculty/facultydir/waldron.html
http://www.stanford.edu/class/me394/
3:30pm Linguistics Department Colloquium [1-Dec-06]
Margaret Jacks Hall 460:126
Title to be announced
John T. Hale
Michigan State University
http://www.msu.edu/~jthale/
http://www-linguistics.stanford.edu/colloq/
____________
Stanford Blood Center: Shortage of O, A, and AB-. For an appointment:
http://bloodcenter.stanford.edu/ or call 650-723-7831. It only takes
an hour of your time and you get free cookies.
____________
UC BERKELEY LINGUISTICS DEPARTMENT COLLOQUIUM
on Monday, 27 November 2006, 4:00pm
182 Dwinelle (Berkeley)
http://linguistics.berkeley.edu/events/
"The origins of vowel length neutralization patterns"
Scott Myers
University of Texas, Austin
In languages with contrasts in vowel length, there are recurring
patterns in which those contrasts are neutralized under the same
conditions across a range of unrelated languages. Many of these
recurring patterns correspond directly to phonetic patterns in vowel
duration (closed syllable shortening, stressed syllable lengthening,
monosyllabic lengthening). In these cases, the phonological pattern
can be attributed to language learners misinterpreting a durational
pattern as a pattern in category distribution (phonologization).
Other recurring patterns have no such clear correspondence to
durational patterns. There is a widespread pattern in which word-final
vowels are required to be short. This is in contrast with a robust
duration pattern in which final vowels are longer than nonfinal
ones. I argue that phonological final shortening arises as a result of
utterance-final devoicing, which renders the end of final vowels
nonsalient. This hypothesis is supported by a series of perception
experiments with Finnish listeners in which final devoicing increased
the probability that a final vowel would be identified as short.
Another recurring pattern without a clear counterpart in duration is
the lengthening of vowels next to vowel-like sonorants. I present more
evidence from Finnish speakers in support of the claim that this
lengthening results from listeners' responses to the long ambiguous
formant transitions in these sequences.
These patterns correspond, then, not to clear durational patterns, but
to contexts where it is difficult to determine the acoustic boundaries
of the vowel. Patterns of this sort are related to a general model of
phonologization, interpreted as the core of a theory of recurring
phonological patterns.
____________
BERKELEY INTEL RESEARCH SEMINAR
on Monday, 27 November 2006, 4:00pm
Intel Research Berkeley, 2150 Shattuck, Ste. 1300
http://www.intel-research.net/berkeley/Seminars.asp
"How Simply and Understandably Could The 'Personal Computing'"
Alan Kay
Viewpoints Research Lab
The "Personal Computing Experience" is a difficult to define large
range of interactive dynamic media that includes user interfaces,
desktop publishing, spreadsheets, email, web media, hypercard-like
authoring, and quite a bit more, all supported by systems software of
various kinds including programming languages, storage systems,
TCP/IP, etc. The conglomeration of commercial and most open source
software consumes in the neighborhood of several hundreds of millions
of lines of code these days. We wonder: how small could be an
understandable practical "Model T" design that covers this
functionality? 1M lines of code? 200K LOC? 100K LOC? 20K LOC? Other
scientific fields try to sum up what they think they know by devising
mathematics that can describe the theories "in an eyeful" (e.g.
Maxwell''s Equations). The computing sciences rarely do this (but
should). Such an "eyeful" for "personal computing" might be a book
instead of the current "Library of Congress", and it will likely
require new kinds of programming and systems design principles to make
it cover the desired range of functions, user interfaces, small size,
practicality, and understandability. It perhaps would not constitute a
"reinvention of programming", but it would likely be "serious steps
towards a reinvention of programming". This talk will discuss a just
started and funded research project to tackle these questions, and
will cover some of about a dozen "principles" that have been invented
in CS over the last 50 years that we think can be the basis of the
needed compact but expressive architecture.
____________
BERKELEY REDWOOD CENTER FOR THEORETICAL NEUROSCIENCE SEMINAR
on Tuesday, 28 November 2006, 12 noon
3105 Tolman (Berkeley)
http://redwood.berkeley.edu/seminars.php
"Learning Invariant Features Using Inertial Priors, or
'Why Google might want to be in the neocortex business?'"
Thomas Dean
Brown University / Google
We address the technical challenges involved in combining key features
from several theories of the visual cortex in a single computational
model. The resulting model is a hierarchical Bayesian network factored
into modular component networks implementing variable-order Markov
models. Each component network has an associated receptive field
corresponding to components in the level directly below it in the
hierarchy. The variable-order Markov models account for features that
are invariant to naturally occurring transformations in their inputs.
These invariant features support efficient generalization and produce
increasingly stable, persistent representations as we ascend the
hierarchy. The receptive fields of proximate components on the same
level overlap to restore selectivity that might otherwise be lost to
invariance. Technical jargon aside, we believe there is enough known
about the primate cortex to enable engineers to build systems that
approach the pattern-recognition capability of human vision. Moreover,
we believe that such a capability can be implemented using the
distributed computing infrastructure that Google has today.
____________
SRI CSL SEMINAR SERIES
on Tuesday, 28 November 2006, 3:30pm - 5:00pm
EK255, SRI International
http://www.csl.sri.com/
"Peptide discover:
From mass spectrometry to bioinformatics to function"
Lloyd D. Fricker
Pharmacology, Albert Einstein College of Medicine, Bronx, NY
Neuropeptides perform a large variety of functions as intercellular
signaling molecules. It has been estimated that there are a large number
of undiscovered neuropeptides based on the finding of over 100 orphan
receptors in the mouse genome, many of which have homology to known
neuropeptide receptors. Traditionally, peptides have been discovered by
one of two general methods. One approach starts with a crude tissue
extract and a bioassay involving whole animals, isolated tissues, or
cultured cells. The endogenous factor present in the crude extract that
stimulates the bioassay is then purified to homogeneity and identified.
The other general approach is to first identify peptides present in a
tissue extract, without concern for function, and then test these
peptides in various bioassays. A modern version of this approach, termed
"peptidomics," uses mass spectrometry and bioinformatics approaches to
identify peptides. Quantitative peptidomics studies generally use
isotopic tags to label two sets of extracted peptides, as done with
proteomic studies, using isotopic tags that react with amine groups. The
advantage of quantitative peptidomics over non-quantitative studies is
that knowledge of the conditions that regulate levels of a neuropeptide
can provide clues as to the neuropeptide's function. The discovery and
putative functions of novel mouse brain peptides will be discussed. In
addition, current limitations of the peptidomics approach that could be
improved by better computational analysis will be described.
About the Speaker: Lloyd Fricker received a Ph.D. from the Department
of Pharmacology, Johns Hopkins University School of Medicine
(Baltimore, MD) in 1983, where he trained with Dr. Solomon Snyder. He
received post-doctoral training in the laboratory of Dr. Edward
Herbert at the Vollum Institute of the Oregon Health Sciences
University (Portland, OR). In 1986 he started at Albert Einstein
College of Medicine (Bronx, NY) as an Assistant Professor of Molecular
Pharmacology, with promotion to Associate Professor in 1991 and full
Professor in 1996. Dr. Fricker has published 135 scientific articles,
reviews, and book chapters, edited one book, and authored another.
(Information for attending is probably similar to those for the SRI AI
seminars, see http://www.ai.sri.com/visiting/ )
____________
BERKELEY FUTURE OF SCHOLARLY COMMUNICATION SEMINAR
on Tuesday, 28 November 2006, 4:00pm
CSHE Library, South Hall Annex (Berkeley)
http://cshe.berkeley.edu/events/
"Academic Amnesia: Who Is Preserving Our Data?"
Abby Smith
Independent Consultant
Scholarship is built on the cumulative record of the past and the
well-tended, authentic, and readily accessible data of the present.
Current federal efforts to build a digital information preservation
infrastructure at the Library of Congress and the National Archives
assume that research institutions responsible for producing large
quantities of research data, such as the University of California,
will take responsibility for ensuring its long-term access. Is that a
reasonable expectation? What is at risk if they do not?
About the Speaker: Abby Smith is a historian and consulting analyst
with special interest in the creation, preservation, and use of the
cultural record in a variety of media; the impact of digital
information technologies on cultural heritage institutions; and the
evolving role of information as a public good. Until her recent
relocation to the Bay Area, she was director of programs at the
Council on Library and Information Resources in Washington, DC. Prior
to that, she worked at the Library of Congress managing programs
relating to preservation of and access to cultural heritage
collections.
She currently works with the Library of Congress National Digital
Information Infrastructure and Preservation Program (NDIIPP) in
development of its national strategy to identify, collect, and
preserve digital content of long-term value. She is an advisor to the
ACLS Commission on the Cyberinfrastructure for the Humanities and
Social Sciences and serves as Senior Advisor to the Scholarly
Communications Institute at the University of Virginia. She consults
with several universities on identifying content of long-term value,
understanding various risk factors to its persistence, and analyzing
organizational strategies for its long-term access.
She holds a doctoral degree in history from Harvard University and has
taught at Harvard and Johns Hopkins Universities. She has published
most recently on the Library of Congresss digital preservation
initiative (Distributed Preservation from a National Perspective:
NDIIPP at Mid-Point, in D-Lib Magazine
http://www.dlib.org/dlib/june06/smith/06smith.html ). Other recent
publications include: Access in the Future Tense; New-Model
Scholarship: How Will It Survive?; Strategies for Building Digitized
Collections; The Evidence in Hand: Report of the Task Force on the
Artifact in Library Collections; and Authenticity in the Digital
Environment.
____________
SRI AI SEMINAR SERIES
on Tuesday, 28 November 2006, 4:00pm - 5:30pm
EJ228, SRI International
http://www.ai.sri.com/seminars/
"Data Delivery in a Service-Oriented World:
The BEA AquaLogic Data Services Platform"
Michael Carey
BEA Systems, Inc.
"Wow. I fell asleep listening to SOA music, and when I woke up, I
couldnt remember where Id put my data. Now what?" Has this happened to
you? With the new push towards service-oriented architectures and
process orientation, data seems to have been lost in the shuffle. At
the end of the day, however, applications are still about data, and
service-oriented architecture (SOA) applications are no different. In
this talk, we will present BEAs approach to serving up data to SOA
applications. Last year, BEA introduced a middleware product called
the AquaLogic Data Services Platform (ALDSP). The purpose of ALDSP is
to make it easy to design, develop, deploy, and maintain a data
services layer in the world of SOA. ALDSP provides a new, declarative
foundation for building SOA applications and services that need to
access and compose information from a range of enterprise data
sources.
The talk will attempt to cover both the foundation and the key
features of ALDSP, including its underlying technologies, its overall
system architecture, and its interesting capabilities. Time
permitting, a brief demo may be included.
About the Speaker: Michael J. Carey is a Sr. Engineering Director on
the AquaLogic side of BEA. Mike manages BEAs XQuery engineering team
and is the chief architect for the AquaLogic Data Services Platform
product. Prior to joining BEA in 2001, Mikes experiences include a
dozen years as a University of Wisconsin-Madison Computer Science
faculty member, five years as an IBM Almaden database
researcher/manager, and a year and a half working under various
inflated titles at an e-commerce software startup.
____________
SILICON VALLEY WEB GUILD
on Wednesday, 29 November 2006, 6:00pm
Google (900 Alta Avenue, Mountain View)
http://www.webguild.org/
"SEO for Web 2.0"
Panel
Web 2.0 is making the entire internet a giant community by connecting
communities of users together. User-generated content can quickly grow
a web site to vast dimensions in little time. Web 2.0 companies depend
heavily on communities and viral marketing to drive traffic to their
sites. However, Web2.0 sites need to rely on search engines to attract
visitors as well. But getting a Web 2.0 site to rank "well" is not as
easy as it seems because there generally is less editorial control and
various structural issues can cause lack of indexability. These issues
can represent insurmountable hurdles for online visibility but they
can be overcome.
This event will focus on:
* How Web 2.0 companies can increase their traffic and search engine
rankings.
* Why all content can't be tucked away behind a membership login?
* How to open up internal data to the search engines without
compromising user privacy?
* Structural barriers limiting indexability and maximizing the
spiderability of web sites.
About the Speakers:
Adam Lasnik, SEO Strategist, Google:
Before there was a public Internet, Adam was e-mailing. Before there
was Netscape or Internet Explorer, he was surfing the Web. He's
written comprehensive search engine optimization reports, managed
sponsored ad campaigns for Fortune 500 companies, and provided broad
communications consulting to successful startup companies. Adam earned
MBA and law degrees -- focused on Global Electronic Communications and
Commerce issues -- and then moved to Germany to serve as an
entrepreneurial consultant to a multinational IT company. Grateful for
the international experience but fascinated by the burgeoning American
dot.com scene, he hopped over to San Francisco and joined the high
tech PR firm Niehaus Ryan Wong as an Interactive Strategist, helping
clients understand and leverage the power of online communities. When
the dot.com boom turned to bust, Adam spent the next years broadening
his online communications and advertising chops with a mix of small
and large companies. In 2006, Adam became Google's first Search
Evangelist, dedicated to building stronger relationships between
Google and Webmasters.
David Hahn, Director of Advertising, LinkedIn:
David directs LinkedIn's advertising business and manages several
strategic product initiatives, including public profiles. LinkedIn's
public profile system helps users publish their profiles on the web,
so that they can be found by search engines and referenced by a custom
url. David also designed and implemented LinkedIn's profile-targeted
advertising system, creating a new targeting paradigm for advertisers.
Prior to LinkedIn, David served as an aide to Senator Dianne Feinstein
in Washington, D.C. and a fellow at The Carter Center in Atlanta.
Markus Hoevener, Chief Visionary, Bloofusion:
Markus is a pioneer in the search engine industry. During his studies
in computer science and marketing at the University of Dortmund,
Germany, he started developing a search engine specifically designed
for the needs of the German marketplace. By 1996 he had co-founded
evision, a company that specialized in the development of search
engine technology. His first commercially available search engine
project, Nathan, was successfully demonstrated in 1997 at Cebit, the
world's largest trade show. A year later, he developed the meta search
engine Apollo.7 that filled the niche for German web content. In 1999
he developed the first wireless access protocol (WAP) search engine
WapUP! and co-founded his next company to market this technology in
Munich, Germany. This VC-backed company, WAP Communications, licensed
its wireless search engine technology to AltaVista. In 2001, he
conceived a search engine focused on the medical world, Mediwarp.com.
Finally, in 2002, he designed his latest search engine that gathers
high tech articles from the Web: TrooBloo.com.
Joelle Gropper Kaufman, Vice President of The Experience, Engage.com:
Joelle joined Engage Corporation in 2006 as Vice President of The
Experience. She is the Engage concierge, responsible for creating and
delivering the most natural, compelling and useful online community
experience that enables members to discover new possibilities for
serious relationships. Over ten years ago, Joelle was a Director of
the Firefly Networks product and account management teams, led
licensing and integration with BarnesandNoble.com for Firefly, was CEO
of Lifespire, and led marketing and sales development for technology
companies such as Reactivity and RSA Securitys Developer Solutions
Group. At Engage, she is refocusing on enabling people to communicate
and collaborate in the journey of creating meaningful relationships
and love. Joelle earned her BA in Organizational Psychology (Studies)
from the University of Michigan and her MBA as a Baker Scholar from
the Harvard Business School.
Moderator: Research Analyst, JupiterKagan
____________
BERKELEY HISTORY AND PHILOSOPHY OF LOGIC, MATHEMATICS, AND SCIENCE
on Wednesday, 29 November 2006, 6:00pm - 7:30pm
234 Moses (Berkeley)
http://hplms.berkeley.edu/
"The Future of Proof"
Dana Scott
Carnegie Mellon University
Goedel showed us many things. Among others he showed us the
possibility of proof (via the Completeness Theorem for First-Order
Logic); and then quite soon thereafter he showed us the impossibility
of proof (via the Incompleteness Theorem for (suitable) Higher-Order
Logics). These results are well known and famous, but their impact on
the practice of mathematics has perhaps not been very noticeable. To
be sure, related recursive unsolvability results have a clear
explanatory value in keeping people from searching for algorithms
where none can exist. And modern developments in complexity theory
show that many easily stated problems have -- in general -- no quick
solutions. But again, many commentators agree that there has not been
a big shift in main-stream mathematics as a consequence of Goedel's
fundamental work. However, the insight into formalization sparked by
Goedel's original work is now having major payoffs in mechanized
mathematics and proof systems. The lecture will survey some
developments, but it will also bring up the questions of what we
should now regard as a proof and of how new proof methods develop.
____________
SRI AI SEMINAR SERIES
on Thursday, 30 November 2006, 11:00am
EJ228, SRI International
http://www.ai.sri.com/seminars/
"Machine Learning in Web 2.0:
Analyzing Dynamic Content-driven Social Networks"
Sugato Basu
SRI
In the last decade, machine learning and data mining techniques have
seen widespread successful application to different Internet
technologies, including web search, product recommendation, spam
detection, spelling correction, and news clustering. However, the web
is fast undergoing a paradigm shift, moving from being a mechanism for
delivering static web-content in the existing Web 1.0 model to a
platform facilitating dynamic collaborative content creation in the
emerging Web 2.0 paradigm. This trend is reflected in the growing
popularity of new social web-services, for example, tagging (Flickr)
compared to photo editing (Ofoto), and blogging (Blogger) compared to
homepage hosting (Geocities). This talk will outline how this new
emphasis on rapid creation and sharing of consumer-generated data
(CGM) over large social networks has given rise to dynamic
content-driven social networks, and a new set of challenging machine
learning problems in this context. Focusing on a project (iLink) that
the speaker is currently working on, the talk will discuss research
problems like online learning of topic models over streaming text,
large-scale topic analysis over social networks, and learning to route
messages in a social query model.
About the Speaker: My main areas of research interest are machine
learning, data mining, information retrieval, statistical pattern
recognition and optimization, with applications to analysis of text
and biological data. I am currently working as a research scientist on
the CALO project.
____________
CSLI COGLUNCH
on Thursday, 30 November 2006, 12 noon - 1:00pm
Cordura Hall 100
http://www-csli.stanford.edu/events/Coglunch/
"The Relationship Between Consciousness and Self-Awareness"
Johannes Brandl
University of Salzburg
What is the relation between having an experience and being self-
aware? This question has played an important role in the
phenomenological tradition before it resurfaced in the recent debate
between higher-order and first-order representational theories of
consciousness. In addressing this question it will be useful to
distinguish two different concerns here: One concern is the
extensional relation between the two terms: Does consciousness occur
with and without self-awareness, or is self-awareness involved in
every conscious experience? A second concern is the explanatory
relation between these terms: Does an explanation of consciousness
have to proceed via an explanation of self-awareness, or can we
explain self-awareness via an explanation of consciousness? My thesis
will be that it is not the task of a theory of self-awareness to
explain what conscious experience is, but we can (and in fact must)
use a theory of consciousness to explain in which sense one is
minimally self-aware whenever one has a conscious experience.
____________
STANFORD SOFTWARE SEMINAR
on Thursday, 30 November 2006, 3:00pm - 4:00pm
Gates 104
http://theory.stanford.edu/~mhn/sss.html
"What's Next for Software Verification?"
Rupak Majumdar
UCLA
http://www.cs.ucla.edu/~rupak/
Over the last few years, software verification based on predicate
abstraction and counterexample-guided refinement has been a successful
technique for performing automatic and precise static analysis of
programs. Other than device driver protocols, though, software
verifiers have so far mostly been applied to simple properties of
systems. For these programs and properties, we show a simple syntactic
algorithm to construct approximate program invariants that is
surprisingly effective, suggesting perhaps that software verification
tools are overkill for these applications. For deeper properties, we
describe our recent attempts to verify a memory management subsystem
and outline several challenges that must be met before the
verification will succeed. We describe some partial progress. In
particular, we describe a technique to infer predicates in the
presence of data structures such as lists and sets, and a type system
that extracts word-level relationships from code containing bitwise
operations.
____________
BERKELEY CENTER FOR STUDIES IN HIGHER EDUCATION
on Thursday, 30 November 2006, 4:00pm
175 Barrows Hall (Berkeley)
http://cshe.berkeley.edu/events/
"Issues in Higher Education Access for English Language Learners:
Thoughts on Pathways through the Master Plan"
Kenji Hakuta and George C. Bunch
Stanford University and UC Santa Cruz
English Language Learners and students with English as a second
language now comprise a substantial proportion of California's K-12
schools. Very little is known about their progress into the higher
education system. In this talk, we will bring together what we know
about the academic performance of the second language population in
K-12, and propose a research agenda for better understanding their
pathways into and through the higher education system.
About the Speakers: Kenji Hakuta is Professor in the School of
Education at Stanford University. He was previously founding Dean of
the School of Social Sciences, Humanities and Arts at UC
Merced. George C. Bunch is Assistant Professor of Educational
Linguistics at UC Santa Cruz. His research interests include education
of language minority students and transition of underrepresented
students to higher education.
____________
SSP10: SYMBOLIC SYSTEMS FORUM
on Thursday, 30 November 2006, 4:15pm
Bldg. 380:380C (Math Corner)
http://symsys.stanford.edu/ssp_events
"Selecting Among Paraphrases"
Tom Wasow
Linguistics, Stanford
What leads speakers to select one way of saying something over another
way of expressing the same thought? This lecture proposes four
general strategies of utterance production that influence the choice
among alternative formulations:
- Contiguity - Minimize interruptions internal to syntactic and
semantic units;
- Procrastination - Postpone producing complex units;
- Brevity - Keep what is predictable short; and
- Audience Design - Let your audience know when you are having
difficulty.
It also argues against the widely held assumption that ambiguity
avoidance is a major factor in the choice among syntactic
alternatives.
Evidence for these strategies will be drawn from corpus studies and
psycholinguistic experiments, with special attention to the following
five alternations in English:
- Heavy NP Shift: "They take too many dubious assumptions for
granted" vs. "They take for granted too many dubious assumptions"
- Dative Alternation: "We gave a bone to a dog" vs. "We gave a dog a
bone"
- Verb Particle Placement: "You figured out the problem" vs. "You
figured the problem out"
- Relativizer Optionality: "This is the book that I was reading"
vs. "This is the book I was reading"
- Complementizer Optionality: "I think that it is raining" vs. "I
think it is raining"
While each strategy is quite simple and intuitive in itself, the
interactions among them lead to some subtle and surprising results.
____________
STANFORD ALGORITHMS SEMINAR (AFLB)
on Thursday, 30 November 2006, 4:15pm
Gates 498
http://theory.stanford.edu/~aflb/
"Incentive-Based Ranking Mechanisms"
Rajat Bhattacharjee
Stanford
http://www.stanford.edu/~rajatb/
We consider ranking/recommendation systems based on user feedback. We
make a case for sharing with users the revenue generated by such
systems. Our main contribution is a mechanism which rewards useful
feedback and is resistant to selfish/malicious behavior (eg. click
spam). The mechanisms are designed to reward discovery of high quality
entities. Misclassified pages represent an arbitrage opportunity for
the discerning user. The mechanisms are simple enough for use with
existing technology. Joint work with Ashish Goel.
____________
FUNDAMENTAL THEMES IN NEUROSCIENCE SEMINAR
on Thursday, 30 November 2006, 4:15pm
Munzer Auditorium, Beckman
http://nis-seminars.stanford.edu/
"The brain's own cannabis - recent developments"
Brad Alger
Physiology and Psychiatry, University of Maryland
http://neuroscience.umaryland.edu/faculty/default.asp?ID=3
Marijuana and related cannabinoids affect the CNS by binding to CB1R
cannabinoid receptors. When released from cells in many areas of the
brain, including the hippocampus, cerebellum, striatum and neocortex,
endogenous cannabinoids (endocannabinoids) also activate CB1Rs.
Typically, endocannabinoids mediate retrograde signaling: they are
released from postsynaptic cells, travel backwards across the synapse,
and inhibit conventional transmitter release from nerve terminals. An
endocannabinoid, often 2-arachidonyl glycerol, 2-AG, is released
following an influx of Ca2+ into principal cells, or by activation of
mGluRs or mAChRs on the principal cells. The most prominent
endocannabinoid-mediated actions in the hippocampus are exerted on
GABAergic interneurons. Short-term suppression of inhibition is
initiated by depolarization-induced Ca2+ influx (DSI), or relatively
brief activation of the G-protein coupled receptors. Long-term
depression of inhibition (iLTD) is caused by activation of mGluRs
lasting for minutes. This talk will focus on the initiation of
endocannabinoid actions by Ca2+, mGluRs and mAChRs. Particular
attention will be devoted to the mechanisms by which endocannabinoids
are mobilized and to the endocannabinoids as agents and subjects of
neuronal plasticity.
Recent Papers:
[1]Edwards, D.A., Kim, J. and Alger, B.E. (2006) Multiple mechanisms
of endocannabinoid response initiation in hippocampus. J.
Neurophysiol. 95:67-75.
http://nis-seminars.stanford.edu/reprints/Alger1.pdf
[2]Heinbockel, T., Brager, D.H., Reich, C.G., Zhao, J., Muralidharan,
S., Alger, B.E. and Kao, J.P. (2005) Endocannabinoid signaling
dynamics probed with optical tools. J. Neurosci. 25: 9449-9459.
http://nis-seminars.stanford.edu/reprints/Alger2.pdf
____________
SRI AI SEMINAR SERIES
on Friday, 1 December 2006, 11:00am
EJ228, SRI International
http://www.ai.sri.com/seminars/
"In-network PCA and anomaly detection"
XuanLong Nguyen
UC Berkeley
We consider the problem of network anomaly detection in large
distributed systems. In this setting, Principal Component Analysis
(PCA) has been proposed as a method for discovering anomalies by
continuously tracking the projection of the data onto a residual
subspace. While successful empirically in moderate sized networks,
this approach has serious scalability limitations. To overcome these
limitations, we develop a PCA-based anomaly detector in which adaptive
local data filters send to a coordinator just enough data to enable
accurate global detection. Our method is based on a stochastic matrix
perturbation analysis that characterizes the tradeoff between the
accuracy of anomaly detection and the amount of data communicated over
the network. This is joint work with Ling Huang, Minos Garofalakis,
Michael I. Jordan, Anthony Joseph and Nina Taft.
About the Speaker: XuanLong Nguyen is a PhD candidate in EECS, UC
Berkeley. His research interests lie in machine learning,
computational statistics, optimization and their applications to
distributed/adaptive systems and sensor networks.
____________
BERKELEY INFORMATION ACCESS SEMINAR
on Friday, 1 December 2006, 3:00pm - 5:00pm
107 South Hall (Berkeley)
http://www.sims.berkeley.edu/courses/is296a-1/f06/schedule.html
"Math People, a Distributed Name Authority System"
Jim Pitman
Math & Statistics, Berkeley
Math People is a Web 2.0 application which leverages multiple sources
of personal name data to provide a distributed name authority system
for people in the mathematical sciences. For users browsing a suitable
webpage related to mathematical science, the system will provide links
back to a name server which will match names and identifiers in the
text and url of the page with name and identifier data gleaned from
various sources including homepages and open access digital
repositories, to provide authoritative links to these sources whenever
possible. The system also allows authorized users to make
identifications and disambiguate the name data.
It is designed to be an open navigation system to allow users to pass
through the walls which currently separate various information
resources. If successful in achieving this purpose in mathematical
sciences, there appears to be no obstacle to propagation of the system
to provide a distributed name authority network spanning any branch of
human knowledge with enough people and professional organizations
willing to support it.
The proposed business model for long term maintenance of the system is
that data providers with adequate financial resources support the
system to enhance the appeal to users of their electronic offerings,
with either open or gated access: with a small tax on that income from
gated resources to support the linking infrastructure. In determining
the extent of its support of such an open system, each data provider
will have to balance its interest in the open flow of academic
information against its instinct to keep users away from competing
sites, or to restrict navigation by a closed linking system such as
CrossRef which does not acknowledge open access digital repositories
or professional homepages. For more see
http://bibserver.berkeley.edu/projects/mathpeople.html
About the Speaker: Jim Pitman is Professor of Statistics and
Mathematics at U.C. Berkeley, and President of the Institute of
Mathematical Statistics http://imstat.org/ . Over the last few years he
has been working towards improving the quality and quantity of open
access content in the mathematical sciences, by promoting and
launching open access journals for expository and survey material, and
by creating bibliographic software to encourage distributed
alternatives to centrally controlled indexing systems. More at
http://www.stat.berkeley.edu/users/pitman/
____________
END MATERIAL
The CSLI Calendar appears weekly on most Wednesdays throughout the
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 many of the CSLI computers can type 'help csli-calendar' to
see the current issue.
The CSLI Calendar is also posted each week to the su.events usenet
newsgroup (only available from computers on the Stanford network)
Information about CSLI's research program is available at
http://www-csli.stanford.edu/
For maps to the Stanford University rooms see
http://www-csli.stanford.edu/events/locations.shtml
____________