Social Patterns: Modeling and Analysis
S-PATTERNS DIT-PRJ-10-060
Status active project
DISI role Coordinator
Project type Research Project
Dimension National
Acquisition date 2010-08-03
Start date 2010-12-01
End date 2014-12-01
Project details
Project astract Social cognition (SC) is the main channel through which human beings access the social world, very much like vision and hearing are channels through which people<br/>access the physical world. For a long time, researchers have addressed the computational implementation of vision and hearing in domains like computer vision and<br/>speech recognition, but only early attempts have been made to do the same with SC. In other words, machines are beginning to be effective in handling perceptual<br/>aspects of the physical world, but up to now are devoid of social context.<br/>We believe SC to be the missing link between humans and computers, and so, our ultimate goal is to put together a set of key disciplines related to the new field of<br/>Artificial Social Cognition (ASC) in a cohesive scientific framework. S-PATTERNS constitutes the first step towards this long-term goal and focuses on the main<br/>physical evidence of SC: social signals. These are a vast array of unconscious nonverbal behaviors that people use during social interactions like conversations,<br/>speeches, presentations, or interviews. A wealth of information is conveyed nonverbally in parallel to the spoken words. Social signals include body gestures and<br/>posture, variations in voice characteristics (pitch, speaking rate, prosody, etc.), facial expressions, and eye gaze, among others. The humans analyze these signals and<br/>are able to extract social patterns which are regulating their social behavior. The project makes use of two fundamental features of social signals: they are social acts<br/>- these signals exist for social interaction and are mainly manifested while interacting - and are measurable.
Fundings 307000 €