Predicting Group Performance in Task-Based Interaction

Murray, Gabriel; Oertel, Catharine

doi:10.1145/3242969.3243027

Cited by 42 publications

(43 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our finding that verbal/linguistic features alone did not yield good prediction performance is, in fact, compatible with the findings of Murray and Oertel [8], who initially had the same result. They subsequently applied domain adaptation and data augmentation and found that the verbal model was amongst the best, second only to the combined verbal+nonverbal model.…”

Section: Modelsupporting

confidence: 91%

“…More recently, Murray and Oertel [8] also used the ELEA corpus to predict group performance. They compared verbal and nonverbal features for this task using the English subset of the corpus.…”

Section: Related Workmentioning

confidence: 99%

“…These include jitter, shimmer, mel-frequency cepstral coefficients (MFCCs), associated delta features, PCM loudness, F0 envelope, F0 contour, and voicing probability. This is the same set of nonverbal features used by Murray and Oertel [8].…”

Section: Nonverbal Featuresmentioning

confidence: 99%

“…In lieu of doing domain adaptation or data augmentation, in this work we combined a newly gathered dataset with the existing ELEA corpus, effectively leading to a two-fold increase in the amount of data. However, the data augmentation method by Murray and Oertel [8] increased the amount of training data four-fold. In future work, we will attempt to replicate the effectiveness of their data augmentation scheme on the merged dataset.…”

Section: Modelmentioning

confidence: 99%

See 3 more Smart Citations

Analyzing Verbal and Nonverbal Features for Predicting Group Performance

2019

Self Cite

View full text Add to dashboard Cite

This work analyzes the efficacy of verbal and nonverbal features of group conversation for the task of automatic prediction of group task performance. We describe a new publicly available survival task dataset that was collected and annotated to facilitate this prediction task. In these experiments, the new dataset is merged with an existing survival task dataset, allowing us to compare feature sets on a much larger amount of data than has been used in recent related work. This work is also distinct from related research on social signal processing (SSP) in that we compare verbal and nonverbal features, whereas SSP is almost exclusively concerned with nonverbal aspects of social interaction. A key finding is that nonverbal features from the speech signal are extremely effective for this task, even on their own. However, the most effective individual features are verbal features, and we highlight the most important ones.

show abstract

Section: Modelsupporting

confidence: 91%

Section: Related Workmentioning

confidence: 99%

Section: Nonverbal Featuresmentioning

confidence: 99%

Section: Modelmentioning

confidence: 99%

See 2 more Smart Citations

Analyzing Verbal and Nonverbal Features for Predicting Group Performance

2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, most SSP work exclusively focuses on nonverbal aspects of interaction, whereas we include linguistic features as well and want to determine the circumstances under which these features are useful for the task at hand. Examples of SSP research on conversational data include automatic prediction of group performance on a task, using nonverbal features of the conversation [12], or a combination of verbal and nonverbal features [13]. Lai and Murray [14] also combine verbal and nonverbal features to predict aspects of group satisfaction and affect.…”

Section: Introductionmentioning

confidence: 99%

Graph-Based Prediction of Meeting Participation

Murray

2019

MTI

Self Cite

View full text Add to dashboard Cite

Given a meeting participant’s turn-taking dynamics during one segment of a meeting, and their contribution to the group discussion up to that point, our aim is to automatically predict their activity level at a later point of the meeting. The predictive models use verbal and nonverbal features derived from social network representations of each small group interaction. The best automatic prediction models consistently outperform two baseline models at multiple time-lags. We analyze which interaction features are most predictive of later meeting activity levels, and investigate the efficacy of the verbal vs. nonverbal feature classes for this prediction task. At long time-lags, linguistic features become more crucial, but performance degrades compared with prediction at short time-lags.

show abstract