Chair for Theoretical Computer Science and Foundations of Artificial Intelligence
 

 

AI/Cognition Group -- Research and Development Foci

The Artificial Intelligence/Cognition Group at the Technical University Munich (headed by Dr. Gerhard Weiß) is concerned with research and development in the following fields:


Intelligent Agents and Multiagent Systems
In the past decade, intelligent agents and multiagent systems (MASs) have become one of the most active areas of research and development in information technology. Focussing on computational entities called agents that act and interact pro-actively, autonomously and flexibly and employ semantically rich forms of communication, agent and multiagent technology is particularly suited for building computer-based applications that are open, distributed and tightly embedded in complex socio-technical surroundings.
Such applications are relevant in a variety of industrial, commercial and scientific domains, including, for instance, e-commerce, telecommunications, knowledge management, and simulation of biological and social processes. Moreover, agent and multiagent technology has the potential to play a key role in putting novel models and paradigms of information processing such as Autonomic Computing, Grid Computing, P2P Computing, Mobile Computing and Pervasive/Ubiquitous Computing into practice.
Our current research in the area of agent and multiagent systems primarily focuses on the following topics:
Agent Communication: Adaptive Models of Communication, Communication Languages and Semantics, Reasoning about Communication.

Contact: Matthias Nickles
more...
Agent and Multiagent Learning

Contact: Achim Rettinger and Gerhard Weiß
coming
soon
Social AI: Researching and Modeling of Artificial Sociality

Contact: Matthias Nickles
more...
Computational Trust

Contact: Achim Rettinger
coming
soon
Agent-Oriented Software Engineering (AOSE):

Contact: Matthias Nickles and Gerhard Weiß
more...


Knowledge and Ontologies in Open Environments
Our research in the field of computational ontologies and knowledge primarily focuses on the semantic representation, emergence, evolution and integration of knowledge from autonomous sources, and on its use by autonomous recipients. Starting from the fundamental insight that in complex, open information environments (like the Semantic Web, Enterprise Knowledge Management Systems, Semantic Grids and Peer-to-Peer Systems), knowledge emerges from (potentially indefinite, oblique or conflictive) communication processes, these objectives are pursued along the following lines of research:
Open Ontologies and Open Knowledge Bases

Contact: Matthias Nickles
more...


Spatial Cognition and Neural Networks
The perception and processing of spatio-temporal patterns is a fundamental part of visual cognition. In order to learn more about the principles behind these biological processes, we analyse and model the representation of spatio-temporal structures on different levels of abstraction. For the low-level processing of motion information we have argued for the existence of a spatio-temporal memory in early vision. The basic properties of this structure are reflected in a neural network model we develop to gather a deeper understanding of motion perception. Our recurrent neural network model is based on the self organizing map (SOM) proposed by Teuvo Kohonen. In order to enable the representation, processing and prediction of spatio-temporal pattern on different levels of granularity and abstraction, the SOMs are organized in a hierarchical manner. The constraints for the neural modeling and the data sets for training the neural network are obtained from psychophysical experiments where human subjects' ability of dealing with spatio-temporal information is investigated.
Motion Perception and Prediction

Contact: Volker Baier
coming
soon
Hierarchically Structured Recurrent Neural Network Models

Contact: Volker Baier
coming
soon


Information about past projects can be found on the group's old homepage.

 
News
No current news. You can look at the archive for older news items by clicking here.
to top >>
old homepage >>