The Workshop on Computational Linguistics and Clinical Psychology: An Overview

April 26, 2019
1:00PM
2:00PM
EDT
Fee: 
Free for AMIA members; $50 for non-members
Presenters: 
Philip Resnik, PhD; Ayah Zirikly, PhD

This is a webinar on applications of NLP to clinical psychology.  It is novel and has been primarily tackled in open domain NLP. This webinar brings together open domain NLP with clinical NLP and is expected to unify two otherwise fairly separate communities.

The Workshop on Computational Linguistics and Clinical Psychology (clpsych.org) was established in 2014 as a forum bringing together researchers on language technology and in-the-trenches clinicians interested in improving the efficiency and reach of mental health treatment. It focuses on: (a) analysis of objective, naturalistic behavioral data to improve upon the assessment of mental health needs provided by self-reported questionnaires (currently the most common method of assessment); (b) providing language researchers a deeper understanding of the field of mental and neurological health and the needs of those currently responsible for diagnosis and treatment; (c) increasing clinicians’ understanding of innovations in language technology and realistic expectations for implementation; (d) co-development of high-value natural language processing tools that can be deployed in the clinical community; and (e) discussion of ethical questions concerning the opportunities and challenges created by human-technology interactions and analysis of the data generated.

In this webinar we will provide a brief overview of the workshop series and what it's accomplished so far, along with a preview of this year's "shared task" exercise on assessment of suicidality from social media postings. Lessons learned from CLPsych 2019 will be presented at the AMIA NLP WG meeting and/or presymposium (pending review) in Fall 2019.

Learning Objectives

After participating in this activity, the learner should be better able to:

  • Understand use of social media for mental health research
  • Understand the use of NLP for social media in identifying folks in need of help
  • Locate new data sets to advance their own NLP work
  • Contribute to mental health literature using social media

Speaker Information

Philip Resnik, PhD 
Professor of Linguistics
University of Maryland, College Park

Ayah Zirikly, PhD
Postdoctoral Fellow
NIH

Philip Resnik is a professor of Linguistics at University of Maryland, College Park. His research focuses on methods for combining machine learning and human insight to solve practical problems, as well as computational models of human language processing. His current work focuses on computational social science, with an emphasis on connecting the signal available in people's language use with underlying mental state. Central applications for this work are computational political science, particularly in connection with ideology and framing, and mental health, where evidence from linguistic behavior shows promise in helping to identify and monitor depression, suicidality, and schizophrenia. Resnik holds two patents and has authored or co-authored more than 100 peer-reviewed articles and conference papers. At various times his work has been highlighted in Newsweek, The Economist, New Scientist, and on National Public Radio, and he has been a repeat organizer and panelist at SXSW Interactive.

Ayah Zirikly is a postdoctoral fellow at the National Institutes of Health. Her work focuses on extracting and classifying information relevant to an individual’s mobility using deep learning models in clinical and self-report data, with the aim to expedite the disability determination process. Additionally, she has been working on building predictive models for suicide ideation in social media in collaboration with Dr. Philip Resnik. During her PhD, under the supervision of Dr. Mona Diab, she worked on building named entity recognizers for low resource languages using transfer learning models, with a focus on dialectal Arabic. She is the author of the Named Entity Recognition component in MADAMIRA toolset.