5th International Conference on Spoken Language Processing

Sydney, Australia
November 30 - December 4, 1998

The CHAM Model of Hyperarticulate Adaptation During Human-Computer Error Resolution

Sharon L. Oviatt

Oregon Graduate Institute, USA

When using interactive systems, people adapt their speech during attempts to resolve system recognition errors. This paper summarizes the two-stage Computer-elicited Hyperarticulate Adaptation Model (CHAM), which accounts for systematic changes in human speech during interactive error handling. According to CHAM, Stage I adaptation is manifest as a singular change involving the increased duration of speech and pauses. This change is associated with a moderate degree of hyperarticulation, which occurs during a low rate of system errors. In contrast, State II adaptations are associated with more extreme hyperarticulation during a high system error rate. It entails change in multiple features of speech - including duration, articulation, intonation pattern, fundamental frequency and amplitude. This paper summarizes the empirical findings and linguistic theory upon which CHAM is based, as well as the model's main predictions. Finally, the implications of CHAM are discussed for designing future interactive systems with improved error handling.

Full Paper

Bibliographic reference.  Oviatt, Sharon L. (1998): "The CHAM model of hyperarticulate adaptation during human-computer error resolution", In ICSLP-1998, paper 0049.