CALL FOR PAPERS/PARTICIPATION

    Document Layout Interpretation and its Applications (DLIA2003)

                  August 2, 2003 (before ICDAR2003)

                         Edinburgh, Scotland

              http://www.science.uva.nl/events/dlia2003

                 Submission deadline: April 18, 2003
         (one month after anticipated ICDAR acceptance date)

---------------------------------------------------------------------------
WORKSHOP CO-CHAIRS:

Andrew D. Bagdanov, University of Amsterdam, The Netherlands
Bertin Klein, DFKI, Germany

---------------------------------------------------------------------------
PROGRAM COMMITTEE (invited members):

Apostolos ANTONACOPOULOS, University of Liverpool, U.K.
Henry BAIRD, Palo Alto Research Center (PARC), USA
Larry BAUM, The Boeing Company, USA
Thomas BREUEL, Palo Alto Research Center (PARC), USA
Andreas DENGEL, DFKI, Germany
David DOERMANN, University of Maryland, USA
Jianying HU, Avaya Labs Research, USA
Hiromichi FUJISAWA, Hitachi CRL, Japan
Rolf INGOLD, University of Fribourg, Switzerland
Yasuto ISHITANI, Toshiba Corporation, Japan
Koichi KISE, Osaka Prefecture University, Japan
Donato MALERBA, Universita degli Studi di Bari, Italy
Simone MARINAI, Universita di Firenze, Italy
Lawrence O'GORMAN, Avaya Labs Research, USA
Larry SPITZ, Document Recognition Technologies, USA
Luc VINCENT, LizardTech, USA
Marcel WORRING, University of Amsterdam, The Netherlands

---------------------------------------------------------------------------
OVERVIEW:

We would like to build on the two successful DLIA workshops, held in
conjunction with ICDAR 1999 in Bangalore and ICDAR 2001 in Seattle,
and bring together researchers in document layout analysis again for
an exchange of ideas, discussion of applications and available tools,
and discussions about future directions.

The techniques of document layout segmentation have reached such a
level of maturity that researchers are beginning to explore the wealth
of possibilities enabled by the availability of accurate layout
information.  The purpose of this international workshop is to bring
together researchers, developers, and users of document understanding
technology to explore the following question:

 "Assuming accurate layout information is available, what now?"

Of course, before approaching such a broad, open-ended question, it is
essential to determine a task-dependent definition of accuracy for
layout information.  Once the interpretation purpose is fixed, the
result of a specific layout analysis technique can be poor or very
good.

While conventional research about layout segmentation is still a very
active area, the emphasis of the DLIA workshop is on the
interpretation of extracted layout information.

In order to stimulate discussion during the next DLIA workshop, we
would like to especially encourage submissions from people with
interesting application scenarios to present.  The emphasis of such
reports should be on the (not necessarily solved) application of
document layout interpretation.

We invite the submission of original, previously unpublished work.  We
also welcome, with some restrictions, submissions which are closely
related to work submitted also to this ICDAR.  Please use our workshop
to present work that differs materially from your ICDAR presentations,
in any of several ways:


  -- recent results too late for the ICDAR deadline;

  -- the motivations and broader implications of work presented in
     more detail at ICDAR;

  -- methodological issues facing the DLIA community;

  -- proposals for community-wide data sets, experiments,
     competitions, web sites, etc.

---------------------------------------------------------------------------
TECHNICAL FOCUS:

The workshop will explore document layout interpretation and its
application in document conversion to structured formats (XML, MS
Word, etc.), genre recognition, document databases and retrieval,
multimedia documents, and automatic hypertext creation.

We invite contributions from individuals or groups working in the
following areas:

  -- Application scenarios involving a strong document layout
     interpretation component;

  -- Layout analysis and page segmentation based either on general
     layout principles, document-specific grammars or models, or
     application specific constraints;

  -- Abstraction principles and models of the document layout
     generation and analysis cycle;

  -- Table understanding, including the detection, segmentation and
     interpretation of tabular data, as well as extraction and
     interaction with tables in all media (e.g., printed documents,
     ASCII documents, HTML, and other standard document formats);

  -- Methods and systems that take advantage of document layout
     information for conversion to structured formats, hypertext
     creation, document management, and document databases;

  -- Acquisition of document models and parameter settings using
     techniques from statistics or machine learning;

  -- Application of document layout interpretation concepts to broader
     classes of documents, such as multimedia, hypertext, or other
     structured document formats;

  -- Use of standardized methods for representation, interchange, and
     visualization of document layout interpretation results
     (e.g. XML, XSL/XSLT, and SVG);

  -- Standardization and dissemination of resources that have already
     reached maturity and broad application within the DLIA research
     community (e.g. databases for training and evaluation,
     representation/interchange formats, and software).

---------------------------------------------------------------------------
WORKSHOP FORMAT:

This will be a one-day workshop for a maximum of about 50 participants.
Each participant will submit an extended abstract.  All participants
are expected to contribute to the discussions.

Abstracts will be published before the workshop.  These abstracts
should be position papers representing a specific view of the
researcher, or application scenario papers describing an application
of document layout interpretation for which no solution currently
exists.

Participants will present summaries of their abstracts at the
beginning of the workshop.  Each session will begin with an
application scenario presentation selected as appropriate to set the
tone and context for the other summary presentations.  Summaries
should identify current needs, open problems, and overall direction of
their research in layout interpretation.  After the summary
presentations we will split up into several discussion groups.  The
afternoon session will consist of a summary of results from the
morning session, followed by additional group discussions, and a
plenary session at the end of the day.

After the workshop, the organizing committee will compile a Workshop
Summary, based on the working group notes, and make it available on
the Web.  It is hoped that the workshop will stimulate cooperative
follow-on activities that will accelerate the pace of research in
document layout interpretation and its applications.

---------------------------------------------------------------------------
EXTENDED ABSTRACT SUBMISSION:

Each potential participant or group of participants should submit an
extended abstract which should arrive no later than April 18, 2003.
Abstracts may be submitted in one of the following forms:

 -- PDF (preferred) or Postscript through the Web at
    http://www.science.uva.nl/events/dlia2003/

 -- PDF or Postscript by sending E-mail to dlia2003@science.uva.nl

 -- In paper form (least preferred), by sending mail to
    DLIA 2003
    Faculty of Science
    Kruislaan 403, 1098 SJ Amsterdam
    The Netherlands

Papers should be formatted according to the ICDAR style guidelines,
and should be limited to four printed pages including references.  The
first page should include the name, address, telephone, fax, and email
address of the corresponding author and possible co-authors.

---------------------------------------------------------------------------