Smart environments are playing an increasing role not only in the research community but also in our everyday's life. Consequently, it comes to no surprise that there is an interest in current research dealing with ways for supporting users in smart environments such as meeting rooms. In the scope of proactive user assistance, two major challenges have been identified:

1) the recognition of users needs and wishes; and
2) ways for controlling the environment in a way that fulfills the requirements for the user.

One research direction for coping with the first challenge are model-based approaches for intention recognition. They encode domain knowledge about users and promise good results in recognizing the activity of human users. Due to noisy sensor information these approaches utilize probabilistic models such as dynamic bayesian networks. To approach challenge (2), different approaches have been proposed, most of them relying on some formal description of potential actions to allow for ad-hoc planning of supporting actions.

Different approaches to bridge the gap between challenge (1) and (2) have been proposed for certain fixed use-cases. This workshop aims at combining model based approaches to recognize the intention of users, and methods for controlling the environment according to user wishes and needs in a systematic manner, which allows the re-usage of developed models and components. The goal of the M2A workshop is to provide a forum for researchers from both areas -- model-based intention recognition and strategy synthesis for smart environments -- where current methods for recognizing the user intentions and for providing assistance in smart environments will be discussed. As an outcome of the discussion we intend to assemble a list of successful as well as unsuccessful approaches for the above mentioned challenges: intention recognition and proactive user assistance. Such collection will provide a guideline for best practices and a possible roadmap to success in this research field. Finally, the list will also contain open problems and research suggestions for further investigation of model based intention recognition, proactive assistance and the bridge between the two areas.

We solicit papers on the following topics:

Reasoning under uncertainty
Probabilistic modeling and inference
Causal reasoning
Model-based reasoning and diagnosis
Graphical models
Situation awareness
Model-based activity recognition
Model-based intention analysis
Strategy synthesis
Proactive assistance in smart environments

As usual, the above list is not an exhaustive list but an indicative one. Submitted papers (up to 12 pages) must be written in English and adhere the submission guidelines posted on the submission page (http://mmis.informatik.uni-rostock.de/m2a/submission.html).