Error-spread modelling

An approach to evaluating the impact of speech recognition errors that accounts for how a single misrecognized word degrades comprehension of its neighbouring words, not just the word itself. For example, misrecognizing "kitchen" as "kitten" makes the subsequent word "area" harder to interpret because a contextual cue has been lost. Error-spread modelling is a key component of the ACE2 caption quality metric, which better predicts Deaf and Hard-of-Hearing users' actual comprehension than metrics that treat each error in isolation.