

School Notes
Date posted: Jul 30, 2018
BC Computer Science Asst. Prof. Emily Prud'hommeaux has received an R21 research grant from the National Institutes on Deafness and Other Communication Disorders in the National Institutes of Health (NIH), for her work Automated analysis of pragmatic language development in Autism Spectrum Disorder. Approximately $500K is expected over three years, with $175K in the first year..
More information below.
(from the award abstract)
In this project, we propose using computational language analysis methods to convert speech into text transcripts and to automatically identify specific areas of pragmatic deficit in those transcripts. We explore these methods using spoken language data we will collect from high-functioning verbal adults with ASD and with typical development. In the course of this project, we will investigate: (1) the precise pragmatic functions that are affected in high- functioning adults with ASD; (2) the extent to which these and other reported pragmatic deficits are observed in individuals with ASD regardless of age; and (3) the accuracy of our novel computational methods for extracting these metrics from spontaneous spoken language samples. While achieving these aims, we will explore the utility of automated analysis of spoken language, setting the stage for a future proposal to develop an automated software tool for analyzing spoken language to identify specific areas of deficit in pragmatic expression. A tool for automatically identifying strengths and weaknesses in pragmatic expression could offer utility not only to researchers investigating language functioning in ASD but also to clinicians tasked with diagnosis and health professionals working to develop targeted therapies and interventions. Such interventions could lead to more favorable social outcomes for individuals of all ages with ASD and higher rates of employment among adults with ASD, potentially yielding reductions in costs associated with and resources allocated for serving the needs of this special population.