======== Aim and Scope ========

Context is one of the key characteristics of pervasive computing systems.
Using the current context of users and/or devices, such systems adapt
themselves automatically to provide better service quality. As an example, a
system can detect which displays are in the users surrounding and switch an
application's output to one of them. However, these systems are typically
reactive, as they can only react to a given context change after it
occurred. To really create seamless adaptive pervasive systems, the system
must get beyond this restriction and become proactive. It must plan ahead
and prepare for future context changes. This way, the system can, e.g.,
initiate the search for replacement devices before a device becomes
unavailable. It can configure a room, e.g., adapting the light level, before
the user enters it. And it can compute better system configurations that
take into account not only the current state of the system but also when and
how often this state will change, e.g., leading to more stable
configurations.
A major precondition for such proactive adaptive systems is prediction of
the future context. Despite recent work, context prediction still faces a
number of challenges, including a large number of information sources, high
heterogeneity of data, the need to minimize the learning phase, and the need
to work in real time.


======== Topics ========

This workshop addresses the following issues:
1. Context prediction algorithms.
2. Automated decision making, proactive adaptation algorithms.
3. Acting on predicted context.
4. Middleware and context management solutions for context prediction.
5. Middleware solutions and programming models for proactive adaptation.
6. Metrics for evaluating context prediction and proactive adaptation
   approaches.
7. Experiences and lessons learnt from built context prediction and/or
   proactive adaptation systems, including information about approaches that
   did and did not work as well as unexpected results.


======== Paper Submission ========

Authors are invited to submit original manuscripts written in English of
no more than six pages in  ACM double-column format. Submissions must cover
one of the topics listed above and not under review elsewhere. All
submissions will be peer-reviewed and selected based on their originality,
merit, and relevance to the workshop.
Paper submissions will be handled online. Accepted papers must be
presented at the workshop and will be included in the ICPS proceedings and
the ACM Digital Library.


======== Organizers ========

  - Gregor Schiele, Universitaet Mannheim, Germany
  - Andrey Boytsov, Lulea University of Technology, Sweden


======== Technical Program Committee ========

  - Patrick Brezillon, Univerisity Pierre et Marie Curie, Paris, France
  - Josef Hallberg, Lulea University of Technology, Sweden.
  - Carsten Jacob, Fraunhofer FOKUS, Germany
  - Daniele Riboni, University of Milano, Italy
  - Andreas Rieger, Technical University of Berlin, Germany
  - Kare Synnes, Lulea University of Technology, Sweden.
  - Verena Majuntke, Universitaet Mannheim, Germany
  - Claudio Venezia, Telecom Italia
  - Arkady Zaslavsky, Lulea University of Technology, Sweden