Construction of Mental Images and their Use in a Listener Model

Autor: Stopp, Eva and Blocher, Anselm

Herausgeber:

The research in Artificial Intelligence concerning the connection between vision and natural language systems relies on what is often called reference semantics. In the situation of a listener not able to see a reported scene, the utterances of the speaker must be perceptually anchored and coherent in order to be understandable. Accordingly, the reporter must be able to anticipate the listeners' understanding by means of mental images. In this paper, we demonstrate the construction of mental images and their comparison to visual perception on the level of spatial relations. The results are used to improve the understandability of planned utterances.