| Popis: |
In this study, we submit short silent videos of ironic and sincere speakers to judges who where asked to describe the speakers' mental state with one word, freely. We vary the level of context: either the judges see only the videos (i.e. base their response on facial expressions only), or the videos with information about what the speakers are saying (facial expressions + semantics), or the videos with information about the utterance context, positive or negative (facial expressions + context), or the videos and both of the above cues. We measure (i) whether judges describe speakers as ironic or not and (ii) whether judges use emotional words or not. |
