2nd CALL FOR PAPERS
SIGIR'07 Workshop: IRGM 2007

Information Retrieval and Applications of Graphical Models
http://irgm.bpiwowar.net
----------------------------------------------------------

MOTIVATION:

Probabilistic models constitute an important kind of Information
Retrieval (IR) model. They have been long and widely used, and offer a
principled way of managing the uncertainty that naturally appears in
many elements within this field. Nowadays, the dominant approach for
managing probability within the field of Artificial Intelligence is
based on the use of Bayesian Networks, and these have also been used
within IR as extensions of classical probabilistic models.

Broadly speaking, a Graphical Model (GM) consists of a qualitative
part, a graph, which may be directed and acyclic in the case of
Bayesian Networks (BN), and a quantitative one, a collection of
numerical parameters, as the conditional probability tables for BNs or
possibility distributions for possibilistic networks. The knowledge
represented in the graphical component is expressed in terms of
dependence and independence relationships between variables. These
relationships are encoded by means of the presence or absence of links
between nodes in the graph. The knowledge represented in the numerical
part quantifies the dependences encoded in the graph, and allows us to
introduce uncertainty into the model.

Therefore, these tools could be very appropriate to model and solve
problems, as those found in the IR field.

GOAL:

As the main goal, this workshop wants to be a common space where
researchers, in general, and young researchers specifically, can show
their innovative GMs applications to the field of IR, in its wide
problem space, opening a new discussion forum. Graphical Models
include Bayesian Networks, possibilistic networks, Markov networks,
dependence graphs, influence diagrams, probability trees, decision
trees, and Fisher Kernel Discriminants, among others.

TOPICS OF INTEREST:

The scope of this workshop will be all the applications of any type of
GMs to IR problems. Possible topics include, but are not limited to:

- Information Retrieval Models for flat or structured document
collections (based on SGML, XML, etc.).
- Cross-language retrieval.
- Web Information Retrieval / Applications to hypertext.
- Indexing and summarization.
- Multimedia information access, as video, audio or image retrieval.
- Document filtering. Recommender Systems: content-based and
collaborative.
- Question answering and extraction.
- Text data mining and machine learning for IR.
- Document classification and clustering (for flat or structured
documents).
- Any other relevant IR topic.

Theoretical or experimental research papers representing original,
previously unpublished works, are invited to be submitted to the
workshop, not exceeding 8 pages. Positioning papers describing
current research are also welcome, as well as opinion papers
discussing GMs in an IR context. Paper should emphasize the strengths
(and weaknesses) of the GMs to solve an IR problem, especially with
respect to other methods. Contributions will be peer-reviewed by two
experts from the related field belonging to the workshop PC and
accepted papers will be included in printed proceedings.

IMPORTANT DATES:

- Deadline for paper submission: May 31th 2007.
- Acceptance / rejection notification: June 19th 2007.
- Deadline for camera ready submission: July 1st 2007.
- Workshop: Amsterdam, July 27th 2007.

PROGRAM COMMITTEE:

- Peter Bruza (Queensland University of Technology, Australia).
- Eric Horvitz (Microsoft Research, USA).
- Gianni Amati (Fondazione Ugo Bordoni, Italy).
- Berthier Ribeiro-Neto (Universidade Federal de Minas Gerais, Brazil).
- Iadh Ounis (University of Glasgow, United Kingdom).
- Luis M. de Campos (Universidad de Granada, Spain).
- Didier Dubois (Université Paul Sabatier, France).
- Pável Calado (Instituto Superior Tecnico, Portugal).
- Ludovic Denoyer (LIP6, University Paris 6, France).
- Yi Zhang (University of California, USA).
- Alfonso E. Romero (Universidad de Granada, Spain).
- Donald Metzler (University of Massachusetts, USA).
- Dunja Mladenic (Jozef Stefan Institute, Slovenia).

PAPER PREPARATION AND SUBMISSION:

The submissions, in PDF format, not exceeding 8 pages and following
the ACM Conference style (see the ACM template page
www.acm.org/sigs/pubs/proceed/template.html using the "Option 2"
style), will be sent by e-mail to the three organisers.

WORKSHOP ORGANIZATION AND CONTACT INFORMATION:

- Benjamin Piwowarski, Yahoo! Research Latin America (bpiwowar at
yahoo-inc.com).
- Juan M. Fernández-Luna, University of Granada (jmfluna at
decsai.ugr.es).
- Juan F. Huete, University of Granada (jhg at decsai.ugr.es).

WORKSHOP WEBSITE:

http://irgm.bpiwowar.net