Gabriel Murray

Prof. Dr. Gabriel Murray

School of Computing, University of the Fraser Valley, Abbotsford, BC V2S 7M8, Canada
orcid

Research Focus & Metrics

Academic Career Length

22Years (2005-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 September 2022 in JMIR Aging
Verified
Journal Article
Open access
TSThomas Soroski
Thomas Soroski
TVThiago da Cunha Vasco
Thiago da Cunha Vasco
SNSally Newton-Mason
Sally Newton-Mason
SGSaffrin Granby
Saffrin Granby
CLCaitlin Lewis
Caitlin Lewis
HJHyeju Jang
Hyeju Jang
Background: Speech data for medical research can be collected noninvasively and in large volumes. Speech analysis has shown promise in diagnosing neurodegenerative disease. To effectively leverage speech data, transcription is important, as there is valuable information contained in lexical content. Manual transcription, while highly accurate, limits the potential scalability and cost savings associated with language-based screening. Objective: To better understand the use of automatic transcription for classification of neurodegenerative disease, namely, Alzheimer disease (AD), mild cognitive impairment (MCI), or subjective memory complaints (SMC) versus healthy controls, we compared automatically generated transcripts against transcripts that went through manual correction. Methods: We recruited individuals from a memory clinic (“patients”) with a diagnosis of mild-to-moderate AD, (n=44, 30%), MCI (n=20, 13%), SMC (n=8, 5%), as well as healthy controls (n=77, 52%) living in the community. Participants were asked to describe a standardized picture, read a paragraph, and recall a pleasant life experience. We compared transcripts generated using Google speech-to-text software to manually verified transcripts by examining transcription confidence scores, transcription error rates, and machine learning classification accuracy. For the classification tasks, logistic regression, Gaussian naive Bayes, and random forests were used. Results: The transcription software showed higher confidence scores (P<.001) and lower error rates (P>.05) for speech from healthy controls compared with patients. Classification models using human-verified transcripts significantly (P<.001) outperformed automatically generated transcript models for both spontaneous speech tasks. This comparison showed no difference in the reading task. Manually adding pauses to transcripts had no impact on classification performance. However, manually correcting both spontaneous speech tasks led to significantly higher performances in the machine learning models. Conclusions: We found that automatically transcribed speech data could be used to distinguish patients with a diagnosis of AD, MCI, or SMC from controls. We recommend a human verification step to improve the performance of automatic transcripts, especially for spontaneous tasks. Moreover, human verification can focus on correcting errors and adding punctuation to transcripts. However, manual addition of pauses is not needed, which can simplify the human verification step to more efficiently process large volumes of speech data.

17 May 2023 in Proceedings of the ACM on Human-Computer Interaction
Verified
Journal Article
Open access
AHAnuj Harisinghani
Anuj Harisinghani
HSHarshinee Sriram
Harshinee Sriram
CCCristina Conati
Cristina Conati
GCGiuseppe Carenini
Giuseppe Carenini
TFThalia Field
Thalia Field
GMGabriel Murray
Gabriel Murray
There has been increasing interest in non-invasive predictors of Alzheimer's disease (AD) as an initial screen for this condition. Previously, successful attempts leveraged eye-tracking and language data generated during picture narration and reading tasks. These results were obtained with high-end, expensive eye-trackers. Instead, we explore classification using eye-tracking data collected with a webcam, where our classifiers are built using a deep-learning approach. Our results show that the webcam gaze classifier is not as good as the classifier based on high-end eye-tracking data. However, the webcam-based classifier still beats the majority-class baseline classifier in terms of AU-ROC, indicating that predictive signals can be extracted from webcam gaze tracking. Hence, although our results indicate that there is still a long way to go before webcam gaze tracking can reach practical relevance, they still provide an encouraging proof of concept that this technology should be further explored as an affordable alternative to high-end eye-trackers for the detection of AD.

11 April 2017
Verified
Book Chapter
Closed access
PDF file
GMGabriel Murray
Gabriel Murray

1 January 2022
Verified
Conference Paper
Open access
PDF file
HLHang Wun Raymond Li
Hang Wun Raymond Li
WXWen Xiao
Wen Xiao
LXLinzi Xing
Linzi Xing
LWLanjun Wang
Lanjun Wang
GMGabriel Murray
Gabriel Murray
GCGiuseppe Carenini
Giuseppe Carenini
Raymond Li, Wen Xiao, Linzi Xing, Lanjun Wang, Gabriel Murray, Giuseppe Carenini. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. 2022.


Verified
Journal Article
Open access
PDF file
GMGabriel Murray
Gabriel Murray
Based on analyzing verbal and nonverbal features of small group conversations in a task-based scenario, this work focuses on automatic detection of group member perceptions about how well they are making use of available information, and whether they are experiencing information overload. Both the verbal and nonverbal features are derived from graph-based social network representations of the group interaction. For the task of predicting the information use ratings, a predictive model using random forests with verbal and nonverbal features significantly outperforms baselines in which the mean or median values of the training data are predicted, as well as significantly outperforming a linear regression baseline. For the task of predicting information overload ratings, the multimodal random forests model again outperforms all other models, including significant improvement over linear regression and gradient boosting models. However, on that task the best model is not significantly better than the mean and median baselines. For both tasks, we analyze performance using the full multimodal feature set versus using only linguistic features or only turn-taking features. While utilizing the full feature set yields the best performance in terms of mean squared error (MSE), there are no statistically significant differences, and using only linguistic features gives comparable performance. We provide a detailed analysis of the individual features that are most useful for each task. Beyond the immediate prediction tasks, our more general goal is to represent conversational interaction in such a way that yields a small number of features capturing the group interaction in an easily interpretable manner. The proposed approach is relevant to many other group prediction tasks as well, and is distinct from both classical natural language processing (NLP) as well as more current deep learning/artificial neural network approaches.