Unsupervised Dynamic Bayesian Network model of speech style accommodation

Research Method · All research methods
Evidence · 2 claims (2 for) · 1 study (1 associational), q2 · 1 of 1 report an effect size · 2 claims rest on one study

Description

The model measures speech style accommodation from raw speech without labeled data or hand-crafted indexical features. It uses latent speaking-style states and binary accommodation states over conversational turns, with acoustic-prosodic observation vectors. As the article states, "Our model embodies two premises. First, a person's speech in any turn is a function of his/her speaking style in that turn, which is influenced by their speech style in their previous turn," and style also depends on accommodation to the partner. Parameters are learned with expectation maximization and inference uses the junction tree algorithm; the accommodation score is the fraction of adjacent turns where both partners' most likely style states match.

Accounts

• Unsupervised Dynamic Bayesian Network model of speech style accommodation motivated by Speech Accommodation Theory: The model measures speech style accommodation from raw speech without labeled data or hand-crafted indexical features. It uses latent speaking-style states and binary accommodation states over conversational turns, with acoustic-prosodic observation vectors. As the article states, "Our model embodies two premises. First, a person's speech in any turn is a function of his/her speaking style in that turn, which is influenced by their speech style in their previous turn," and style also depends on accommodation to the partner. Parameters are learned with expectation maximization and inference uses the junction tree algorithm; the accommodation score is the fraction of adjacent turns where both partners' most likely style states match. (Gweon et al. (2013))

Claims

• The DBN accommodation measure scores real conversation pairs higher than lexically matched constructed pairs, validating it as a measure of speech style accommodation [+M]
• An automatic measure of speech style accommodation significantly positively correlates with prevalence of other-oriented transacts in dyadic debates (R = .36, p < .05) [+M]

Related Research Methods

• Bayesian Networks for student modelling, assessment, and collaborative learning theory

Key Sources

• Gweon, G., Jain, M., McDonough, J., Raj, B., & Rosé, C. P. (2013). Measuring prevalence of other-oriented transactive contributions using an automated measure of speech style accommodation. Computer-Supported Collaborative Learning (2013). doi:10.1007/s11412-013-9172-5

Page metadata

Field	Value
type	research-method
id	unsupervised-dynamic-bayesian-network-model-of-speech-style-accommodation
title	Unsupervised Dynamic Bayesian Network model of speech style accommodation
description	The model measures speech style accommodation from raw speech without labeled data or hand-crafted indexical features.
status	review
generated	by process:settle-candidates on 2026-10-18
trust	unverified — no one has confirmed this page's content

Frontmatter is the page's machine-readable record — see the Schema & Guide.