UAI Applications Workshop 2007

Call for Contributions

Objective

As in previous years, the workshop will  provide a focused but informal forum for exchanges among theorists, practitioners and tool developers. Discussions may cover research questions and insights, methodologies, techniques, and experiences with applications of Bayesian models to
particular domains.  Proceedings of the workshop will be published in the CEUR Workshop Proceedings series.  In addition, the intention is to publish a subset of the papers in a special issue of an appropriate journal.

Workshop Theme

This year we are especially encouraging contributions that address the theme:

            Bayesian Model Views

In software and information systems engineering, it is often useful to make an explicit separation between the model and the user interface.  Separation facilitates system maintenance and extensibility, because changes can be made to the model without affecting the user interface, and vice versa.  Another reason to separate the model from the interface is that different categories of users may interact with the model in different ways.  Examples include users with different degrees of technical sophistication, users having different access privileges, users in different geographic areas, and users who are responsible for different aspects of the problem domain.  It is often desirable to tailor the user interface to these different categories of user. In many domains, there are common domain-specific communication patterns, such as maps, equipment diagrams, or standard tabular charts.  In such cases, it may be desirable to provide model views based on these standard patterns.  Finally, model views that are useful for building, testing and maintaining a model may be very different from the most appropriate views for end users.

We encourage papers that focus on issues relating to model views in applications of Bayesian networks.  Examples include, but are not restricted to, the following issues.

Different uses of a model may require different levels of granularity and resolution.  Papers are encouraged that discuss how model views can be used to present results and/or accept inputs at different granularities and resolution.

Metadata and ontologies may be used to support definition and and management of model views.  Examples of useful metadata may include information about who is invoking the model, what is their role/task, what is their background, where are they
located, or what is the degree of time pressure.  A formal ontology of tasks and users may be useful for reasoning about the most appropriate model view in the given circumstances.

Applications of Bayesian networks are usually tailored to a particular real-world setting, or in some cases, a small collection of real-world settings.  Users must invoke the model, provide data (e.g., evidence, queries, contextual data), examine outputs, evaluate the quality of outputs, and make use of the outputs in some decision context. How do model views support the process of interacting with and making use of a model?  Which model views are most appropriate for which kinds of interactions?

Model views may also be appropriate for knowledge engineering. Papers are encouraged that address model views that are useful for acquiring information from experts.

General Focus

The emphasis of this year's workshop is on issues of model views in developing and using Bayesian models.  Submissions
are encouraged that address fundamental issues, present concrete solutions, describe experiments, or analyze open problems,
within a broad range of application areas involving diagnosis, optimization, temporal reasoning, spatial reasoning, and forecasting.

In addition to submissions addressing model views, we also welcome contributions relevant to the overall focus of Bayesian modelling. Submissions addressing novel applications are particularly encouraged.

Format

The workshop will take place on Thursday, July 19th, the day of the UAI tutorials.  The workshop will begin with each participant giving a short introduction to themselves and their work.  These brief introductions will be followed by sessions devoted to
issues in developing and using Bayesian models, consisting of presentations and open discussions.  We will conclude with a plenary panel discussion which will summarize the issues raised during the day and consider plans for next year.

Arrangements

The workshop will be co-located with the UAI conference in Vancouver, BC, Canada.  A registration fee for the workshop will be required in addition to the main conference registration.

To Contribute

Submissions of 3-8 pages, preferably in the same format as for the overall conference) are solicited, along with a short statement
of interest. The statement of interest should explain the contributors' backgrounds, the application on which they have
worked, and the characteristics of that application domain. The contribution should raise questions and offer results that
participants can discuss. All participants will be expected to have material to present for the workshop.  Participants are
encouraged to apply jointly with members of other disciplines with whom they have collaborated. Submissions and inquires should be addressed to UAIWorkshop-L [at] listserv.gmu.edu.

Participation will be limited to 30 individuals.  All members of the workshop committee who would like to attend are expected to qualify as participants.  Contributors will be selected by the committee to achieve a mix of individuals who are likely to generate the most constructive presentations and discussions.

Workshop Committee

Kathryn Blackmond Laskey, George Mason University, Co-chair
Suzanne M. Mahoney, Innovative Decisions, Inc., Co-chair
Judy Goldsmith, University of Kentucky, Co-chair

John Mark Agosta, Intel Corporation
Russell Almond, Educational Testing Service
Marek Druzdzel, University of Pittsburgh
Sean Gaurino, Charles River Analytics
James H. Jones, Ferris State University
Oscar Kipersztok, The Boeing Company
Linda van der Gaag, Utrecht University, Chair emeritus
Edward J. Wright, Information Extraction and Transport, Inc.

Important Dates

Submission of contributions:                                                       April     6, 2007
Notification of selections by the organizing committee:          May    11, 2007
Deadline for contributions to the proceedings:                        June     8, 2007
Date of the workshop:                                                                  July    19, 2007