INVESTIGADORES
BRUSSINO Silvina Alejandra
congresos y reuniones científicas
Título:
A protocol for collecting speech data with varying degrees of trust
Autor/es:
GAUDER, LAURA; GRAVANO, AGUSTIN; FERRER, LUCIANA; RIERA, PABLO; BRUSSINO, SILVINA ALEJANDRA
Lugar:
Viena
Reunión:
Workshop; Proc. SMM19, Workshop on Speech, Music and Mind; 2019
Resumen:
This paper describes a novel experimental setup for collectingspeech data from subjects induced to have different degrees oftrust in the skills of a conversational agent. The protocol consistsof an interactive session where the subject is asked to respondto a series of factual questions with the help of a virtualassistant. In order to induce subjects to either trust or distrustthe agent?s skills, they are first informed that the agent waspreviously rated by other users as being either good or bad;subsequently, the agent answers the subjects? questions consistentlyto its alleged abilities. These interactions will be speechbased,with subjects and agents communicating verbally, whichwill allow for the recording of speech produced under differenttrust conditions. Ultimately, the resulting dataset will be usedto study the feasibility of automatically predicting the degreeof trust from speech. This paper describes a preliminary experimentusing a text-only version of the protocol in ArgentineSpanish. The results show that the protocol effectively succeedsin influencing subjects into the desired mental state of eithertrusting or distrusting the agent?s skills. We are currently beginningthe collection of the speech dataset, which will be madepublicly available once ready.