Fluency Profiling System: An automated system for analyzing the temporal properties of speech

Abstract

The temporal characteristics of speech can be captured by examining the distributions of the durations of measurable speech components, namely speech segment durations and pause durations. However, several barriers prevent the easy analysis of pause durations: The first problem is that natural speech is noisy, and although recording contrived speech minimizes this problem, it also discards diagnostic information about cognitive processes inherent in the longer pauses associated with natural speech. The second issue concerns setting the distribution threshold, and consists of the problem of appropriately classifying pause segments as either short pauses reflecting articulation or long pauses reflecting cognitive processing, while minimizing the overall classification error rate. This article describes a fully automated system for determining the locations of speech-pause transitions and estimating the temporal parameters of both speech and pause distributions in natural speech. We use the properties of Gaussian mixture models at several stages of the analysis, in order to identify theoretical components of the data distributions, to classify speech components, to compute durations, and to calculate the relevant statistics.

Fluency in Second Language Testing: Insights From Different Disciplines

“…The obvious reason for this change is that the very time-consuming measuring of objective fluency, which had to be done by hand, has been helped along by technological advances. For instance, there are now ways to calculate measures, such as speech rate and silent pauses, automatically, without the need to carry out an orthographic transcription (De Jong & Wempe, 2009;Dekens, Martens, Van Nuffelen, De Bodt, & Verhelst, 2014;Little, Oehmen, Dunn, Hird, & Kirsner, 2013). Even when researchers base their measures on manual transcripts and manual measures of pauses (e.g., Kahng, 2014;Révész et al, 2016), technological advances have made it easier to measure pauses and syllables (semi-)automatically.…”

Empirical studies of the construction of discourse

“…Pause duration is not included in this analysis given that any threshold (for instance between "short" or "long" pauses) would require to take into account each speaker's average speaking rate, following Little et al (2013). This first positional cue to a more global scope of the topic-shift function is, however, not confirmed by co-occurring pauses, where we can observe a similar preference for the [pause+DM] pattern in 44% and 52% of turnmedial topic-resuming and topic-shifting DMs, respectively (turn-initial and turn-final DMs were excluded from this analysis since they are, by definition, less prone to co-occurring with pauses).…”

Local vs. global scope of discourse markers

“…Pause duration is not included in this analysis given that any threshold (for instance between "short" or "long" pauses) would require to take into account each speaker's average speaking rate, following Little et al (2013).…”