Statistical Science and the Internet

Mark (
Mon, 18 May 1998 18:41:37 -0400

Statistical Science and the Internet

A mini-conference to explore the impact of emerging Internet
technologies on the field of statistics.

July 12-14, 1998
Drew University
Madison, New Jersey

Organizer: Mark Hansen, Bell Laboratories

--------------- Call for Participants ---------------

Over the past few years, the Internet has changed the way we announce our
research results, primarily by making it easier for us to exchange drafts of
papers destined for journal publication. With the introduction of new computing
languages like Java, however, we are presented with incredible new
possibilities for sharing statistical ideas.

The purpose of this mini-conference is to explore the ability of emerging
Internet technologies to support and possibly shape statistical research
and usage. Sessions will cover:

* Tools for distributed computing (CORBA, PVM, RMI)

Communication of statistical methods, or perhaps more fundamentally
statistical data, is at the heart of current approaches to distributed
computing. Deciding on an architecture is crucial for sharing statistical
methods both conveniently and efficiently.

* Client-server experiments in statistical computation and visualization

A number of vendors have begun developing Web-based interfaces to their
existing computing platforms, including XploRe and S-PLUS. These
experiments have opened up new and interesting directions in distributing
statistical methodology. Related projects like GASP, the Globally
Accessible Statistical Procedures Initiative will also be discussed.

* Statistical applications of Java and VRML

Several packages for computing and visualization have appeared that are
entirely written in Java. Examples include the Graphics Production
Library, WebStat, and a commercial venture Statlets. The creators of each
platform will discuss the issues involved in designing and implementing
these specialized clients.

* Statistical multiuser environments

The goal of this mini-conference is to bring together tool users, tool
developers and system architects, with a focus on sharing statistical
ideas (graphics, computations, documentation, etc.) in a networked
environment like the Internet. Through our discussions, we hope to
integrate existing systems, allowing the user to focus
statistical ideas rather than surrounding software environments.

We can only accommodate 35 participants. This mini-conference
is an opportunity to share ideas, and all participants are expected
to be active in discussions and presentations of new ideas.
If you are interested in joining us, please send email

--------------- Registration Information ---------------

The conference will take place at Drew University in Madison, New Jersey.
A registration fee will cover the cost of food and lodging for 2 1/2 days.
The mini-conference is being held in conjunction with an workshop on
``Data Visualization in Statistics'' organized by AT&T Labs. For more
mini-conference details, see