Comparing phonetic changes in computer-directed and human-directed speech

Abstract

This paper presents a study that examines the difference of certain phonetic features between human-directed speech (HDS) and device-directed speech (DDS) in human-human-computer interactions. The corpus used for the analyses consists of tasks performed by participants in cooperation with a human confederate and/or a computer-based interlocutor. This includes distributional and temporal analyses, examining the similarities and differences of the overall distribution of the measured features and time-based changes throughout the interactions. The features fundamental frequency, intensity, and articulation rate were selected for analysis. Results show significant differences in a majority of the cases for two of the three selected features as well as insights regarding the participants’ speech behavior during the interaction. These outcomes provide a look into further aspects of HDS and DDS and speech-related features in conversation analysis, which may help studies in topics like addressee detection or human-computer interaction (HCI).