[Prev][Next][Index]

CSLI Calendar, 14 Mar 1996, vol.11:20




        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
______________________________________________________________________________

14 March 1996                    Stanford                      Vol. 11, No. 20
______________________________________________________________________________

      A weekly publication of the Center for the Study of Language and
Information (CSLI), Stanford University, Ventura Hall, Stanford, CA 94305-4115
                               ____________

                CSLI ACTIVITIES DURING 14 -- 22 MARCH 1996

  THURSDAY, 14 MARCH
         4:15 - SSP Forum
                Building 60, Room 61-F
                The Last Forum of the Quarter
                Penelope Eckert, Stanford Linguistics

  FRIDAY, 15 MARCH
        12:30 - HCI Seminar
                Skilling Auditorium
                Magic Cap Human Interface: Adventures in Design
                Kevin Lynch, General Magic
                Abstract below

         3:30 - Linguistics Department Colloquium
                Building 460, Room 146
                The Human Sentence Processor Memory Structure and
                Accessibility: Can You Tell the Difference? 
                Charles Lee, Stanford Linguistics
                Abstract below

  WEDNESDAY, 20 MARCH
         4:15 - Seminar on Computational Learning and Adaptation
                Gates Hall, Room 104
                Pruning with Generalization-based Weight Saliencies
                Morten Pedersen, Technical University of Denmark
                Abstract below

                               ____________

The CSLI Calendar appears weekly on Wednesdays throughout the academic year.
Announcements, abstracts, and other information to appear in the Calendar can
be submitted to [mailto:incalendar@csli.stanford.edu].

Information about CSLI's research program and past issues of the CSLI Calendar
are available at [http://www-csli.stanford.edu/csli/].  The CSLI Calendar is
also posted each week to [news://nntp-csli.stanford.edu/csli.bboard].

                               ____________

                         COGLUNCH SPRING SCHEDULE
                           Theme: Consciousness

The Thursday noon CogLunch series on consciousness will continue
through the Spring Quarter, starting in April.  Here is a tentative
schedule of speakers and titles of the talks:

   April 4: MARLEEN ROZEMOND (Philosophy, Stanford U.)
            "Descartes and Consciousness"
        11: TEED ROCKWELL (Berkeley, CA) 
            "Awareness, Mental Phenomena, and Consciousness:
             A Synthesis of Dennett and Rosenthal"
        18: ROGER SHEPARD (Psychology, Stanford U.)
            "My Experience, Your Experience, and the World We Experience"
        25: JOHN GABRIELI (Psychology, Stanford U.)
            "Consciousness as the Gatekeeper of Memory"
   May   2: BRIAN SMITH (Xerox PARC & Philosophy, Stanford U.)
            "Who's on Third?  The Physical Bases of Consciousness"
         9: BOB ZAJONC (Psychology, Stanford U.)
            "Unappraised Affect"
        16: MICHAEL CORNER (Netherlands Institute for Brain Research)
            "Prologomena to Any Future Mind-Brain Synthesis, and
             a Theory About the Nature of Self-Consciousness: 
             Phenomenological and Physiological Constraints"
        23: JOHN PERRY [Commentary: G. Guzeldere]
            "Indexicality and the Knowledge Argument"
        30: KEN TAYLOR (Philosophy, Stanford U.)
            "The Hard Problem of Consciousness may not be
             Solvable, but Dualism is Still False"
   June  6: Dead Week

NOTE: The next issue of the CSLI Calendar will appear at the start of Spring
Quarter.

                               ____________

                          SYMBOLIC SYSTEMS FORUM
                          on Thursday, 14 March
                    4:15 p.m., Building 60, Room 61-F
                      The Last Forum of the Quarter
                             Penelope Eckert
                           Stanford Linguistics
                        [eckert@csli.stanford.edu]

                               ____________

                  SEMINAR ON HUMAN-COMPUTER INTERACTION
                           on Friday, 15 March
                     12:30 p.m., Skilling Auditorium
             Magic Cap Human Interface: Adventures in Design
                               Kevin Lynch
                              General Magic
                        [kevin_lynch@genmagic.com]

Our goal in designing Magic Cap was to create a personal communicator that
people would welcome into their lives, rather than be a source of stress.  In
order to achieve this, we worked to come up with designs that people could not
only use successfully but also enjoy doing so.

This mission resulted in a several year design adventure, including living
through our first rather interesting usability tests, struggling with
touchscreens, coming up with solutions, relentlessly iterating, and gradually
zeroing in on a successful design.

This fast-paced presentation will use video clips from usability tests as well
as early design sketches to review some of these adventures.

At the conclusion, the speaker will accept challenges to remember why any
aspect of Magic Cap is the way it is and will either tell the story behind it
or be forced to make up something convincing.

KEVIN LYNCH has been designing human interfaces for over a decade and is
currently Director of the Magic Cap development team at General Magic.  He
pioneered the development of Magic Cap's navigation metaphor and has been
responsible for unleashing several features in Magic Cap including Sniffy the
searching dog.  The Magic Cap human interface can be seen in the Sony Magic
Link and Motorola Envoy personal communicators.

Kevin studied computer graphics at the University of Illinois at Chicago ,
where he focused on interactive graphics, working with artists and engineers
in the Electronic Visualization Laboratory.  While in school, Kevin helped
start Challenger Software to create Macintosh applications in 1984, including
Legacy, a graphic adventure game, and Mac 3D, a three-dimensional modeling
package.  He served as the company's Vice President of Product Development.

Kevin later joined Frame Technology , where he designed the human interface
for the first Macintosh version of FrameMaker in 1989, then managed Frame's
Core Technology Group, directing the creation of FrameMaker 4.  Kevin came to
General Magic in 1992 and has helped lead the design of the Magic Cap human
interface, doing his best to create interfaces that succeed at being both
functional and enjoyable.

                               ____________

                    LINGUISTICS DEPARTMENT COLLOQUIUM
                           on Friday, 15 March
                    3:30 p.m., Building 460, Room 146
            The Human Sentence Processor Memory Structure and
               Accessibility: Can You Tell the Difference? 
                               Charles Lee
                           Stanford Linguistics
                         [clee@csli.stanford.edu]

There have been numerous models proposed to model human sentence parsing
performance (Frazier & Fodor 1978, Marcus 1980, Church 1982, Abney 1989,
Gibson 1991, Jurafsky 1993, Stevenson 1994).  A brief discussion of these
models and their limitations will be presented.  This talk will introduce a
model which accounts for parsing preferences, the ease of processing sentences
with certain types of local ambiguity, as well as accounts for the gradedness
and multi-dimensionality in the difficulty of processing center-embedded, high
attachment, and strong and weak garden path sentences.  Examples of sentences
which the processor can account for are shown below:

  (1) a. Mary watched the Olympic trials on TV.
      b. Mary watched the Olympic trials on TV on his new color TV.
      c. Tom believed Bill thought Mary took out the cat on Monday.
  (2) a. John believes Bill.
      b. John believes Bill died.
      c. Susan put the book on the shelf into her backpack.
  (3) a. The mouse the cat the dog barked at chased is mine.
      b. men women children dogs bark at adore love are rare.
  (4) a. John gave the claim that Bill thought Mary died no credibility.
      b. I gave the girl whom you thought Bill liked a book.
      c. I gave the boy who you wanted to give the books to three books.
  (5) a. Without her donations to the charity that won failed to appear.
      b. Without her donations failed to appear.
      b. After the man drank the water proved to be poisoned.
  (6) a. The man gave the girl a ring impressed a watch.
      b. The horse raced past the barn fell.

The proposed model has a memory architecture which consists of a workspace of
3D-tree fragments for which the accessibility and immediate accessibility of
its nodes are limited.  An accessible node requires a unit of short-term
memory, and the total number of short-term memory units is bounded.  The
parsing algorithm uses an ordered non-deterministic depth-first search
algorithm, constructing tree-fragments using a modified head-corner parsing
algorithm.  Because of the unique memory architecture, the model requires, for
a whole parse, only constant time per transition.

The proposed human parsing model was implemented as a computer program and
will be demonstrated at this talk, illustrating how this model processes the
above types of sentences.

                               ____________

             SEMINAR ON COMPUTATIONAL LEARNING AND ADAPTATION
                          on Wednesday, 20 March
                     4:15 p.m., Gates Hall, Room 104
           Pruning with Generalization-based Weight Saliencies
                             Morten Pedersen
                     Technical University of Denmark

The purpose of most schemes for optimizing the architectures of neural
networks is to reduce their generalization error. In this talk I suggest
estimating the weight saliency as the associated change in generalization
error if the weight is pruned. We detail the implementation of a storage
scheme with O(N) complexity and another with O(N^2) complexity that,
respectively, extend Optimal Brain Damage and Optimal Brain Surgery, two
earlier methods for pruning neural net architectures. I ilustrate the
viability of both methods on the prediction of chaotic time series.

This talk describes work done jointly with Lars Kai Hansen and Jan Larsen at
the Technical University of Denmark.

                               ____________