Jan Miksatko, Research

About me

I am a researcher and PhD candidate at the EMBOTS group researching multimodal human-computer interfaces with embodied agents. We are a member of the Cluster of Excellence Multimodal Computing and Interaction (MMCI) and the group is based at German Research Center for AI (DFKI), Intelligent User Interfaces Lab, Saarbrücken, Germany. My PhD work is supervised by Dr. Michael Kipp and Prof. Wolfgang Wahlster.

My research interests are development of multimodal user interfaces with embodied agents, evaluation of the innovative user interfaces and application of artifical intelligence techniques into real systems in general.

I obtained a M.Sc. in Teoretical Computer Science from Charles University in Prague, Czech Republic (2008) and a M.Sc. in Computer Science from Kansas State University in Manhattan, KS, USA (2004). My previous research at both institutions focused on application of machine learning techniques to pattern recognition in online student discussions and prediction of execution times in a load balancing task. Parallel to my research and study occupation, I worked at industrial projects for several companies in the USA and the Czech Republic.



Persuasive agents in the car

Persuasive agents in an eco driving scenario

Embodied agents are known to invoke social reactions similar to those seen in human-human communications, for instance a student feels an obligation to learn harder in front of a teacher. Persuasion is a set of strategies aimed at changing behavior (or even an attitude) of the receiver by means of communication. We combine both areas and develop a generic persuasive engine with embodied agent output. We evaluate the engine in an eco driving scenario - the agent persuades the driver towards ecological behavior while respecting the competing goal of reaching the target fast and being unbotrusive. For instance, consider the following scenario: A car drives uphill and then reaches a traffic light that turned red. An intelligent persuasive engine tries to convince the driver to proceed with medium speed uphill, freewheel to the traffic lights (because they are going be red anyway) and turn off the engine while waiting at the lights.

The engine first computes a series of future driving decisions with optimal balance of being ecological and being fast (usually contradictory goals). Then the persuasive module creates an abstract FML message based on selected future driving actions, user model and persuasive strategies. The strategies are based on the model of Fogg and they are encoded as combination of if-then rules and finite state machines. Finally, an output realizer converts the FML message into a multimodal utterance in BML. It decides on the time scheduling and realization form of the abstract message and passes the output to the Embots framwork. The utterance is rendered by an icon resembling a human face on the dashboard while driving or by an embodied agent on the navigation display while standing (e.g. at a traffic light).


ITeach: Evaluating virtual character benefits on a learning task with repeated interactions

Embodied agents have the potential to become a highly natural human-computer interaction device. However, it remains an open question whether adding an agent to an application has a measurable impact, positive or negative, in terms of motivation and learning performance. Prior studies are very diverse with respect to design, statistical power and outcome; and repeated interactions are rarely considered. We present a controlled user study of a vocabulary trainer application that evaluates the effect on motivation and learning performance. Subjects interacted either with a no-agent and with-agent version in a between-subjects design in repeated sessions. As opposed to prior work (e.g. Persona Effect), we found neither positive nor negative effects on motivation and learning performance, i.e. a Persona Zero-Effect.



INTAKT - Interaktive Avatar Kommunikations-Technologie

INTAKT is a project funded by the German ministry of research and education to foster cooperation between small/medium-sized enterprises and research. In INTAKT we explore the tight coupling of character animation technology and intelligent authoring tools. We cooperate with the character animation company Charamel GmbH and are working on extensions of our Scenemaker tool. A first milestone application consisted of an instrumented supermarket (cooperation with the Innovative Retail Laboratory, St. Wendel) where multiple virtual shop assistants give advise and talk to the user and the user's personal (mobile) agent.



I co-supervised a master thesis by I. Gregor: "IVAN: A plan-based approach for affective sports commentary in real-time"

I co-taught a proseminar on Human-Computer Interaction: "Proseminar "Mensch-Computer-Interaktion (SS 2009)"



Journal Articles

Miksatko, M., Kipp, K.H., Kipp, M. (submitted) Evaluating the benefits of pedagogical agents in repeated interactions: Comparing motivation, distraction and memory performance.

McLaren, B.M., Scheuer, O., & Miksatko, J. (2010). Supporting collaborative learning and e-discussions using artificial intelligence techniques. In International Journal of Artificial Intelligence in Education (IJAIED) 20(1).

Wegerif, R., McLaren, B.M., Chamrada, M., Scheuer, O., Mansour, N., Miksatko, J., & Williams, M. (2010). Exploring creative thinking in graphically mediated synchronous dialogues. Computers & Education, 54(3), 613-621 doi:10.1016/j.compedu.2009.10.015

Conference Papers

Endres C., Miksatko J., Braun D. (to appear) Youldeco - Exploiting the power of Online Social Networks for Eco-Friendly Driving. In Adjunct proceedings of the 2nd International Conference on Automotive User Interfaces and Interactive Vehicular Applications (AutomotiveUI 2010).

Miksatko, J., Kipp, K.H., Kipp, M. (2010) The Persona Zero-Effect: Evaluating virtual character benefits on a learning task. In: Proceedings of the 10th International Conference on Intelligent Virtual Agents (IVA-10), Springer.

Miksatko, J., Kipp, M. (2009) Hybrid Control for Embodied Agents Applications. In: Proc. of the 32nd Annual Conference on Artificial Intelligence (KI 2009), LNAI 5803, pp. 524-531, Springer.

Gregor, I., Kipp, M. and Miksatko, J. (2009) IVAN - Intelligent Interactive Virtual Agent Narrators. In: Proceedings of the 9th International Conference on Intelligent Virtual Agents (IVA-09), Springer, pp. 560-561.

McLaren, B.M., Wegerif, R., Miksatko, J., Scheuer, O., Chamrada, M., & Mansour, N. (2009). Are your students working creatively together? Automatically recognizing creative turns in student e-Discussions. In V. Dimitrova, R. Mizoguchi, B. du Boulay, & A. Graesser (Eds.), Proceedings of the 14th International Conference on Artificial Intelligence in Education (AIED-09), Artificial Intelligence in Education: Building Learning Systems that Care: From Knowledge Representation to Affective Modelling. (pp. 317-324). IOS Press.

Wegerif, R., McLaren, B.M., Chamrada, M., Scheuer, O., Mansour, N. & Miksatko, J. (2009). Recognizing creative thinking in graphical e-Discussions using artificial intelligence graph-matching techniques. In C. O’Malley, D. Suthers D., P. Reimann, & A. Dimitracopoulou (Eds.) Computer Supported Collaborative Learning Practices, Proceedings of the 9th International Conference on Computer Supported Collaborative Learning (CSCL-09), Vol. I (pp. 108-112). International Society of the Learning Sciences, Inc. ISBN978-1-61584-137-0.

Miksatko, J. & McLaren, B.M. (2008). What’s in a cluster? Automatically detecting interesting interactions in student e-Discussions. In B. Woolf, E. Aimeur, R. Nkambou, S. Lajoie (Eds.), Proceedings of the 9th International Conference on Intelligent Tutoring Systems (ITS-08), Lecture Notes in Computer Science, 5091 (pp. 333-342). Berlin: Springer.

Miksatko J., Andresen D. (2005). A parameter-based dynamic scheduling system for J2EE server clusters. In M. H. Hamza (Ed.), Proceedings of the IASTED International Conference on Web Technologies, Applications and Services (WTAS 2005), ACTA Press, ISBN 0-88986-485-3.


Miksatko, J. (2007). Using Machine Learning Techniques to Analyze and Recognize Complex Patterns of Student E-Discussions. M.Sc. thesis, Charles University, Prague, Czech Republic. The work was done at DFKI and supervised by Dr. Bruce McLaren (DFKI and Carnegie Mellon University).

Miksatko, J. (2004). Dynamic Load Balancing Of Fine-Grain Services Using Prediction Based On Service Input. M.Sc. thesis, Kansas State University, Manhattan, KS, USA. Supervised by Dr. Daniel Andresen (Kansas State Univeristy).