*** Apologies for cross-postings ***


!!! DEADLINE EXTENSION !!!

!!! New Deadline for Submission: 20 June 2007 !!!

"Towards Genre-Enabled Search Engines: The Impact of NLP"
Workshop held in conjunction with RANLP 2007

Workshop Location: Borovets, Bulgaria
Workshop Date: 30 September 2007
Workshop website: http://www.sics.se/use/genre-ws/
RANLP 2007 website: http://lml.bas.bg/ranlp2007/

Organizers: Marina Santini and Georg Rehm


WORKSHOP DESCRIPTION
Broadly speaking, genres are textual categories that streamline
communication by relying on acknowledged conventions and raising
predictable expectations. For instance, the conventions underlying the
BLOG genre are represented by a sequence of daily entries that contain
narratives, opinions, as well as the feelings of the blogger, an
individual who wishes to participate in a discussion on a certain
subject. These entries are public and other bloggers can directly
comment on them by sending their own postings. These conventions are
different from those underlying the EDITORIAL genre, where a single
person presents an argumentative statement of views that are
considered to be representative of a newspaper as a whole. In brief,
genres convey the context of communication. This context is essential
when determining the relevance of the information contained in a text.

The concept of genre has great potential for Information Retrieval
(IR). One application is its integration in a search-engine
architecture, enabling the user to combine topic-based search with
genre-based search.

The distinction between topical and non-topical textual dimensions is
crucial when it comes to features. Traditionally, topics and domains
rely on features based on content words (e.g. in the bag-of-words
approach), while genre classes appear to be more easily identified
through the use of grammatical features (like function words, POS
tags, and syntactic features). As Natural Language Processing (NLP)
provides methods to retrieve grammatical features, an investigation of
the influence of NLP on automatic genre identification appears of
primary importance. For this reason, we wish to investigate to what
extent NLP can help identify genre in an IR scenario.

The main aims of the workshop are as follows:

* To identify the key features that help classify documents by genre,
and their efficiency and effectiveness within an IR framework.
* To explore how genres can improve search engines, and in particular
how it can be integrated with topic-based retrieval.
* To bring together researchers working on genre in different
communities – such as Computational Linguistics, NLP, Information
Retrieval and Extraction, Text and Web Mining, or Summarization – in
order to investigate the extent to which NLP can assist or enable
automatic genre identification.


TOPICS
The topics of interests include but are not limited to:

- Computational modelling of genres for web applications;
- Implementation of genre-based applications for retrieval;
- The impact of lexis, morphology and syntax on automatic genre identification;
- Genre-revealing features and their automatic extraction;
- Genre-driven NLP tools, such as genre-driven taggers, parsers, or
discourse annotation;
- Genre classification schemes accounting for multi-genre and no-genre
documents;
- Test collections and evaluation metrics for genre-enabled applications;
- Clustering and visualization of results based on genre.


CATEGORIES OF PAPERS
Papers can be submitted to one of two categories: REGULAR PAPER and
POSTER. Authors must designate one of these categories at submission
time. REGULAR PAPERS are full-length papers and must NOT exceed 8
pages. POSTERS are submitted as extended abstracts (MAX 4 pages). If
accepted, a poster is presented in the workshop poster session. Both
regular papers and poster papers are included in the workshop
proceedings.


SUBMISSION INSTRUCTIONS
Format. Authors are invited to submit papers and posters on *original
and unpublished* work in the topic area of this workshop. Papers and
posters should be submitted as PDF files, formatted according to the
RANLP 2007 stylefiles, without author name(s) and affiliation(s).
Papers and posters should not exceed the length indicated above. The
RANLP 2007 stylefiles are available at:
http://lml.bas.bg/ranlp2007/submissions.htm

Please, send papers and posters to both [log in to unmask] and
[log in to unmask]

Reviewing. Each submission will be reviewed at least by three members
of the Program Committee. Reviewing will be blind. Reviewers will be
asked to provide detailed comments, and to score submissions on the
following factors:

- Relevance to the workshop
- Significance and originality
- Technical/methodological accuracy
- References to related work
- Presentation (clarity, organisation, English)

Accepted papers policy. Accepted papers (regular papers and posters)
will be published in the workshop proceedings. By submitting a paper
at the workshop the authors agree that, in case the paper is accepted
for publication, at least one of the authors will attend the workshop;
all workshop participants are expected to pay the RANLP-2007 workshop
registration fees.

Dual submissions to the main RANLP 2007 conference and this workshop
are allowed; if you submit to the main session, please DO indicate
this when you submit to the workshop. If your paper or poster is
accepted for the main session, you should withdraw it from the
workshop upon notification by the main session.


REGISTRATION
Information on registration and registration fees are provided at the
conference website (http://lml.bas.bg/ranlp2007/).

IMPORTANT DATES
First Call for Papers: March 20-22, 2007 (passed)
Second Call for Papers: May 2, 2007 (passed)
Final Call for Papers: June, 2007
*** NEW SUBMISSION DEADLINE: June 20, 2007 ***
Workshop paper acceptance notification: July 25, 2007
Camera-ready papers for workshop proceedings due: August 31, 2007
Workshop date: September 30, 2007


PROGRAM COMMITTEE
Shlomo Argamon (Illinois Institute of Technology, USA)
Roberto Basili (University of Rome Tor Vergata, Italy)
Pavel Braslavski (Institute of Engineering Science, RAS, Russia)
Kevin Crowston (Syracuse University, USA)
Aidan Finn (DERI, Ireland)
Jussi Karlgren (Swedish Institute of Computer Science, Sweden)
Alexander Mehler (Bielefeld University, Germany)
Sven Meyer zu Eissen (University of Weimar, Germany)
Alessandro Moschitti (University of Rome Tor Vergata, Italy)
Michael Oakes (University of Sunderland, UK)
Andreas Rauber (Vienna University of Technology, Austria)
Maarten de Rijke (University of Amsterdam, The Netherlands)
Georg Rehm (University of Tübingen, Germany) (co-chair)
Mark Rosso (North Carolina Central University, USA)
Dmitri Roussinov (Arizona State University, USA)
Marina Santini (University of Brighton, UK) (co-chair)
Serge Sharoff (University of Leeds, UK)
Michael Shepherd (Dalhousie University, Canada)
Efstathios Stamatatos (University of the Aegean, Greece)
Benno Stein (University of Weimar, Germany)
John Tait (University of Sunderland, UK)
Ozlem Uzuner (State University of New York, USA)


ORGANIZING COMMITTEE
Marina Santini (University of Brighton, UK)
Email: [log in to unmask]
Personal Home Page: http://www.nltg.brighton.ac.uk/home/Marina.Santini/

Georg Rehm (University of Tübingen, Germany)
Email: [log in to unmask]
Personal Home Page: http://georg-re.hm/


CONTACT INFORMATION
For questions or comments, please contact Marina Santini
([log in to unmask]), or Georg Rehm
([log in to unmask]).