[automatically translated] This article discusses the cognitive architecture of a humanoid robot based on the paradigm of Latent Semantic Analysis (LSA). The LSA approach allows the creation and use of a multi-dimensional conceptual space and data driven. This paradigm is a step toward a simulated emotional behavior of a robot that interacts with humans. The architecture is organized into three main areas: Subconcettuale, emotional and behavioral. The first area perceptual processes the data from the sensors. The second area is the "conceptual space of emotional states" which is the sub-symbolic representation of emotions. The last a latent semantic behavior active area linked to the humanoid emotional state. The robot generates its overall behavior taking into account his "
|Number of pages||9|
|Journal||ATTI DELLA ACCADEMIA DI SCIENZE, LETTERE E ARTI DI PALERMO, PARTE PRIMA: SCIENZE|
|Publication status||Published - 2011|