Today's information technology is rapidly moving small computerized consumer devices and hi-tech personal appliances from the desks of research labs into sales shelves and our daily life. Various platforms from low performance PDAs, embedded computers in cameras, cars, or mobile phones, up to high performance wearable computers will become essential tools in many situations for private and professional use. This development will continue and computers will start to vanish into the environment, computational power and networking capabilities are then becoming ubiquitous.
Systems will by then be organised in intelligent environments that require new interaction metaphors and methods of control. Well-known explicit interaction devices, such as mouse and keyboard are not necessarily available, rendering user interfaces that rely on them inappropriate. Beside more natural interaction techniques (e.g. speech and gesture) mobile and static sensors will be used in intelligent environments to support implicit user interaction.
One of the main challenges for the success of ubiquitous computing is the design of personalised user interfaces and software that allows easy access to relevant information and that is flexible enough to handle changes in user context and availability of resources.
Currently, issues relating to user modelling in ubiquitous computing are gaining more and more attention from research groups world wide including those representing the User Modelling, HCI and the Ubiquitous Computing communities
The goal of this workshop is to bring together researchers of these communities to enhance the exchange of ideas and concepts, and to go one step further towards personalisation in Ubiquitous Computing.
Format:
We believe that the topic is rich enough for a one day workshop. We will accept letter of interests of potential participants as well as classical paper-style submissions, both of which will be reviewed by members of the program committee prior acceptance.
Each submission has to explicitly address one or more of the following questions.
Questions:
What are the different levels of abstractions of context that are needed to build up suitable user models for Ubicomp applications?
How can users edit and control their models in Ubicomp environments?
How should we modify and extend existing UM Languages and representations to address specific Ubicomp issues?
How can user models be made mobile while ascertaining persistency across sessions and locations?
What Ubicomp applications require or would profit from advanced user modelling?
How can the acquisition of user models be adapted to limited resources of the Ubicomp environment?
All submissions should include an abstract, list of keywords and full
author contact information. All submissions must be in English. Papers
should not exceed 5000 words (excluding figures and tables). Authors
should use the APA style of citations, two-column format and a font no
smaller than 10 points. Submissions should be made in PDF or postscript
formats.
Submissions should be sent electronically to ubium03@di.uniba.it by March 3, 2003.
Altogether the format will consist of four elements:
This question addresses the problem of deriving higher level contexts from raw data Ubicomp sensors. Closely related is the question how to handle and represent uncertainty that comes along with those sensors and if the different context levels are domain independent.
The role of sensor and context history is also of major interest.
In Ubicomp scenarios lots of information is collected about the users and their context. This opens up major questions regarding security and inspectability of the inferred user models.
Submissions should address appropriate techniques to enable users to interrogate and/or amend their user model.
The answer to this question would help to understand if the existing
UM Languages are powerful enough to handle user modelling issues in a
Ubicomp scenario. Of major interest would be to look at the combination of group/community modelling and private user models.
We are interested in collecting good examples of applications in Ubicomp scenarios that make use of user modelling. Are there good examples where user modelling helps consumer guidance and recommendation systems to have the "right" level of proactiveness and adaptivity (i.e. not to confuse or annoy users). We are interested to hear about experiences with systems, this includes but is not restricted to intelligent tourist and navigation systems, intelligent shopping guides, restaurant recommendation systems, and news services.
In Ubicomp environments computational and sensor resources are highly variable. We are interested to see approaches that are resource adaptive to these contextual factors. For example, processing confidential and private user data might be possible only on a machine that is under private control und thus with very limited computational resources (e.g. a PDA). In other scenarios certain sensors might be not available and have to be compensated by others to obtain significant inferences.
Submission details and Contact Information: