MuSAMA

MuSAMA was a graduate school running from 2006 to 2015, where young researchers collaborated on the topic of self-organizing appliance ensembles. MuSAMA is based on the hypothesis that ubiquitous machine intelligence, envisioned for our future everyday environments, will be provided by dynamic ensembles: Local agglomerations of smart appliances, whose composition is prone to frequent, unforeseeable, and substantial changes. Members of such ensembles need to be able to cooperate spontaneously and without human guidance in order to achieve their joint goal of assisting the user. The resultant concept of autonomous cooperative assistance poses new challenges for the research on ubiquitous and ambient information technology.

In the context of MuSAMA, I worked in the field of behaviour modelling for activity and intention recognition. I investigated the combination of rule-based and probabilistic methods for recognising the user actions and goals based on imperfect sensor observations.

I started working in MuSAMA in 2010 as a PhD scholarship holder. In 2014 I completed my PhD dissertation entitled “Methods for Engineering Symbolic Human Behaviour Models for Activity Recognition”. Later, I continued my work there, first by receiving a DFG research grant within the context of MuSAMA and later, as a postdoctoral researcher.

Project-related publications

 

2015

  • Kristina Yordanova, Thomas Kirste. A Process for Systematic Development of Symbolic Models for Activity Recognition.  In ACM Transactions on Interactive Intelligent Systems, 2015
  • Martin Nyolt, Frank Krüger, Kristina Yordanova, Albert Hein, Thomas Kirste. Marginal filtering in large state spaces.  In International Journal of Approximate Reasoning, 2015
  • Kristina Yordanova. Evaluation of Requirements for Human Behaviour Modelling for Activity Recognition. , University Library, University of Rostock, 2015
  • Frank Krüger, Albert Hein, Kristina Yordanova, Thomas Kirste. Recognising the Actions during Cooking Task (Cooking Task Dataset). , University Library, University of Rostock, 2015

 

2014

  • Kristina Yordanova, Martin Nyolt, Thomas Kirste. Strategies for Reducing the Complexity of Symbolic Models for Activity Recognition.  In Proceedings of the The 16th International Conference on Artificial Intelligence: Methodology, Systems, Applications (AIMSA 2014), Varna, Bulgaria, 2014
  • Kristina Yordanova, Thomas Kirste. Towards Systematic Development of Symbolic Models for Activity Recognition in Intelligent Environments.  In Proceedings of the The 3rd Workshop on AI Problems and Approaches for Intelligent Environments held at ECAI 2014, Prague, Czech Republic, 2014
  • Albert Hein, Frank Krüger, Kristina Yordanova, Thomas Kirste. Towards causally correct annotation for activity recognition.  In WOAR 2014: Workshop on Sensor-based Activity Recognition, Fraunhofer-Verlag, Stuttgart, 2014
  • Kristina Yordanova. Methods for Engineering Symbolic Human Behaviour Models for Activity Recognition. , Rostock, Germany, 2014. urn:nbn:de:gbv:28-diss2014-0133-5
  • Frank Krüger, Martin Nyolt, Kristina Yordanova, Albert Hein, Thomas Kirste. Computational State Space Models for Activity and Intention Recognition. A Feasibility Study.  In PLoS ONE, 2014

2013

  • Frank Krüger, Kristina Yordanova, Albert Hein, Thomas Kirste. Plan Synthesis for Probabilistic Activity Recognition.  In Proceedings of the 5th International Conference on Agents and Artificial Intelligence (ICAART 2013), Barcelona, Spain, 2013

2012

  • Kristina Yordanova, Frank Krüger, Thomas Kirste. Context Aware Approach for Activity Recognition Based-on Precondition-Effect Rules.  In 9th IEEE Workshop on Context Modeling and Reasoning (CoMoRea’12) at the 10th IEEE International Conference on Pervasive Computing and Communication (PerCom’12), Lugano, Switzerland, 2012
  • André Hoffmeyer, Kristina Yordanova, Stefan Teipel, Thomas Kirste. Sensor based monitoring for people with dementia: Searching for movement markers in Alzheimer’s disease for a early diagnostic.  In Constructing Ambient Intelligence, Springer Berlin Heidelberg, Amsterdam, Netherlands, 2012. presented at the Workshop on User Interaction Methods for Elderly People With Dementia at the AmI 11.
  • Kristina Yordanova, Frank Krüger, Thomas Kirste. Strategies for Modelling Human Behaviour for Activity Recognition with Precondition-Effect Rules.  In Proceedings of the 35th German Conference on Artificial Intelligence, Saarbrücken, Germany, 2012
  • Kristina Yordanova, Frank Krüger, Thomas Kirste. Tool Support for Activity Recognition with Computational Causal Behaviour Models.  In Proceedings of the 35th German Conference on Artificial Intelligence, Saarbrücken, Germany, 2012
  • Frank Krüger, Kristina Yordanova, Thomas Kirste. Tool Support for Probabilistic Intention Recognition using Plan Synthesis.  In Proceedings of the 3rd international joint conference on Ambient Intelligence, Pisa, Italy, 2012
  • Frank Krüger, Kristina Yordanova, Veit Köppen, Thomas Kirste. Towards Tool Support for Computational Causal Behavior Models for Activity Recognition.  In Proceedings of the 1st Workshop: “Situation-Aware Assistant Systems Engineering: Requirements, Methods, and Challenges” (SeASE 2012) held at Informatik 2012, Braunschweig, Germany, 2012
  • Frank Krüger, Kristina Yordanova, Christoph Burghardt, Thomas Kirste. Towards Creating Assistive Software by Employing Human Behavior Models. Davy Preuveneers, Paulo Novais, Juan M. Corchado (eds.). In Journal of Ambient Intelligence and Smart Environments, IOS Press, 2012
  • Proceedings of the Joint Workshop of the German Research Training Groups in Computer Science. Kristina Yordanova, Alexander Gladisch, Kristian Duske, Daniel Janusz, Sebastian Bader (eds.)., Books on Demand GmbH, 2012

2011

  • Kristina Yordanova. Human Behaviour Modelling Approach for Intention Recognition in Ambient Assisted Living.  In Proceedings of 2nd International Symposium on Ambient Intelligence (ISAmI 2011), Springer Berlin Heidelberg, Salamanca, Spain, 2011
  • Kristina Yordanova. Modelling Human Behaviour Using Partial Order Planning Based on Atomic Action Templates.  In The 7th International Conference on Intelligent Environments, Nottingham, UK, 2011
  • Kristina Yordanova. Toward A Unified Human Behaviour Modelling Approach. , Rostock, Germany, 2011. ISSN 0944-5900