From Communicative Strategies to Cognitive Modelling

Abstract

The question addressed in this paper is what we can learn from natural language interaction between humans and robots for modelling the robot's cognitive structure. Experiments were carried out which show which communicative strategies human users employ in the interaction with an artificial communication partner. The communicative strategies found in the interaction show a complex mental model of the domain under consideration, that is, spatial instruction. The findings provide insights for future modelling of the robot's cognitive abilities.