Neuro-cognitive mechanisms and computational modelling of spatial belief revision
In order to integrate new spatial information, intelligent agents have to revise what they belief about the arrangements of objects or about the way objects are arranged in a room. Starting point of the project is the assumption that spatial beliefs result from construction and inspection of mental models. Mental models are cognitive representations of information given in spatial descriptions.
Investigations of spatial belief revisions are based on three levels: In psychophysical experiments we investigate how consistency is regain, after inconsistent spatial information that conflicts with current spatial beliefs is encountered. What information is retracted, which is maintained in order to regain consistency? In imaging studies we collect evidence what is the neuronal basis of spatial belief revision. Formal computational methods are used to model spatial reasoning and spatial belief revision. The aim of the project is to develop a theory of spatial belief revision that encompasses all three levels: behaviour, biological implementation, and mechanisms underlying the information processing steps.
The project is conducted in collaboration with the SFB/TR 8 "Spatial Cognition" at the Universities Bremen and Freiburg and with the Chair for Foundation of Artificial Intelligence, Prof. Dr. Bernhard Nebel, at the University of Freiburg.
Supported by Deutsche Forschungsgemeinschaft under grant AZ: KN-465/6-2
The project is funded from 2011 till 2013.
Contact: Dr. Leandra Bucher, Dipl.-Psych. Jelica Nejasmic, Prof. Dr. Markus Knauff