Cosit Tutorial Sessions 25 August 1999

organized by: Werner Kuhn and Scott M. Freundschuh
 

T1 (9:00-12:30)

Tom Hewett:
Cognitive Factors in Design:
Basic Phenomena in Human Memory and Problem Solving.
 

Description:

One aspect of the overall goal for this tutorial
is to help develop an approximate, intuitive feel for human memory and
problem solving processes which will serve as a context for use in
designing information systems and supporting decision making in spatial
and other tasks. Another aspect is to develop that
understanding in a context which illustrates some applications to HCI design
and which facilitates further learning or teaching.
 

Biographical Sketch
Tom Hewett is Professor of Psychology at Drexel University
where he teaches courses on Cognitive Psychology, Psychology of Human
Computer Interaction, and Problem Solving and Creativity. He has offered
variants of this tutorial to hundreds of interface designers at both
conferences and in-house training sessions. He is a published courseware
author, has worked on the development and evaluation of several projects,
and is currently working with a group of computer scientists who are
developing a Scientific Problem Solving Environment which integrates
symbolic and numeric computing. Dr. Hewett chaired the ACM SIGCHI
Curriculum Development Group which developed recommendations for
undergraduate curricula and courses for HCI and was one of the general
co-chairs for the CHI '94 conference held in Boston, MA, USA.
 
 
 

T2 (9:00-12:30)

Ivo Duentsch:
Relation algebras and their applications in spatial reasoning

Description:
Recently, composition based reasoning with binary relations has been of
interest to the QSR community and the expressive power, consistency and
complexity of relational reasoning has become an object of study.
 

It has been known for some time, that the expressiveness of reasoning with
basic operations on binary relations is equal to the expressive power of
the three variable fragment of first order logic (see Tarski/Givant, 1987,
and the references therein. Thus, it seems worthwhile to use methods of
relation algebras, initiated by Tarski (1941), to study part of and
contact relations in their own right, and then explore their expressive
power with respect to various spatial domains.
 

In this tutorial, we present the basic theory and main results of
relation algebras, and present their expressive power of part of and contact
relations on well known spatial domains.

Ivo Duentsch
School of Information and Software Engineering,
Faculty of Informatics, University of Ulster
Ph: (+44)(0) 1232 368976
http://www.infj.ulst.ac.uk/staff/I.Duentsch
 
 
 
 

T3 (14:00-17:30)

Daniel R. Montello:
Errors in Human Spatial Knowledge: Empirical and Analytic Approaches

Description:
In this tutorial, we will discuss the study of errors in spatial knowledge in
larger-scale spaces, places and environments such as rooms, buildings,
campuses, neighborhoods, cities, landscapes, and regions. Different types of
spatial knowledge may be metric, nonmetric, or fuzzy metric. Various meanings
of error will be considered, but the focus will be on differences between
estimated and actual ("objectively correct") spatial relations. We will go
over methods for the empirical study of spatial error, including their
strengths and limitations. Several issues involved in the analysis and
interpretation of spatial cognitive data will be discussed, including
distinctions between absolute, constant, and variable error; within-subject
vs. between-subject aggregation; special aspects of directional and distance
estimates; and the use of multidimensional scaling (MDS). Some common
patterns
of errors in the scientific literature will be presented. Attendees will
perform some demonstration exercises during the tutorial.

Biographical Sketch:
Daniel R. Montello is Associate Professor in the Department of Geography, and
Affiliated Faculty in the Department of Psychology, at the University of
California, Santa Barbara. He has a Ph.D. in Psychology from Arizona State
University and was a Postdoctoral Affiliate at the Institute of Child
Development, University of Minnesota. His research interests are in the areas
of spatial perception, cognition, and behavior; cognitive issues in
cartography and GIS; spatial aspects of social behavior; and environmental psychology and
behavioral geography. He may be reached at: montello@geog.ucsb.edu
 
 
 

T4 (14:00-17:30)

Thomas Roefer:
Cognitive Robotics

Description:
During the recent years, a new range of application has appeared for robots:
the so-called "Service Robotics". Service robots are artificial autonomous
mobile systems that work in the immediate vicinity of human beings. They,
e.g., deliver mail in an office building, transport food in a hospital, or
guide persons through a museum. In contrast to robot arms in factories, this
new kind of robot has to work in places not specially designed for it, i.e.
it has to navigate in a dynamic, partially unknown environment. Therefore,
it needs sensors to perceive its surroundings, and it must accumulate the
readings of these sensors to a - maybe only local - model of the world to be
able to react in an appropriate way. This link from perception to action is
the major topic in the research field of "Cognitive Robotics".
The tutorial will try to give a rough survey about this area, and it will
attempt to arouse interest for further studies, i.e. to work with autonomous
robots. Mainly, it will focus on navigation. Robot navigation consists of
two basic tasks: A robot has to determine its current position
(self-localization), and it must be able to perform at least one step
towards the goal location (perform an action). The combination of both
generates a robot's "behavior". As some navigation approaches in cognitive
robotics are inspired biologically or psychologically, the tutorial will
give some brief examples of navigation techniques of natural systems. An
important means for self-localization and planning a way to the goal are
maps - either metrical or topological. It will be shown how maps can be
learned, and how they are used in robot navigation.

Dr. Thomas Röfer
Bremer Institut für Sichere Systeme,
Fachbereich 3 - Mathematik und Informatik,
Universität Bremen, Postfach 330 440,
28334 Bremen, Germany