Connecting Multiple Disciplines to AI Techniques in Interaction-centric Autism Research and
Diagnosis (ICARD)
12th September, 2023
Workshop @ SIGDial-INLG 2023
About this Workshop
In this workshop, we invite researchers and practitioners both in academia and industry with backgrounds across different disciplines including psychology, artificial intelligence, natural language processing, and autism research for a discussion on how to connect these disciplines to different AI techniques for autism research and diagnosis focused on interaction.
Motivation
Children-centered language-based interaction can help understand language development in children, identify deficits, and mediate targeted interventions for neuro-developmental disorders such as autism spectrum disorder (ASD). Developing language-based tools to enhance the diagnosis of autism poses several challenges.
Rich interaction datasets specifically focused on ASD individuals are a precursor for building such tools. Construction of such datasets, collected across different geographical areas, cultures, and races, is difficult and does not currently exist. Data sharing is another challenge due to privacy concerns and institutional constraints.
Doctors use children's behavioral and developmental markers such as impaired social interaction, disordered verbal and non-verbal communications, and repetitive behaviors to make a diagnosis of autism spectrum disorder. Bringing multimodal information to model social interaction during a conversation can provide a nuanced perspective in the diagnosis process.
Researchers working in this research space come from a range of domains including psychology, linguistics, computer science, social science, and human-computer interaction. There is a need for conversation across the disciplines to find agreements about the measures needed for diagnosis and intervention.
Schedule
9:00-9:10 Welcome and introduction
9:10 - 10:00 Keynote by Steven Bedrick: Crossing the computational chasm: Lessons learned (and mistakes made) in interdisciplinary collaboration with clinicians
10:00-10:40 Session 1: Accepted papers oral presentations
10:40-11:00 Coffee/Tea Break
11:00-11:40 Session 2: Accepted papers oral presentations
11:40- 12:30 Panel discussion: Current Challenges in Autism Research: Discussion on obtaining data and separation between disciplines
12:30-13:30 Lunch and Networking
Keynote Speaker
Steven Bedrick
Associate Professor of Medical Informatics and Clinical Epidemiology
Oregon Health and Science University
Talk Title:
Crossing the computational chasm: Lessons learned (and mistakes made) in interdisciplinary
collaboration with clinicians
Abstract:
Computational tools and techniques have the potential to be incredibly useful in the study, diagnosis, and treatment of speech and language disorders. Developing such technologies, and translating them into clinical practice, depends on effective interdisciplinary collaboration between clinicians and computer scientists. The two fields differ profoundly in their scientific and organizational cultures, and as a result interdisciplinary and translational research can be... complicated. In this talk, I will lead a tour through ten years’ worth of mistakes made, and lessons learned, from my collaborations with clinicians across multiple disciplines, including Autism research, aphasiology, and critical care; my goal will be to share what I have learned along the way, and in turn to learn from audience members about their experiences.
Speaker Bio
Dr. Steven Bedrick is an Associate Professor of Medical Informatics and Clinical Epidemiology at Oregon Health and Science University. His research focuses on biomedical applications for speech and language technologies, with a primary focus on speech and language disorders. His ongoing projects in this area include computational assessment of post-stroke anomia and characterization of language features in Autism. His second area of research focus is machine learning and information retrieval in support of secondary use of electronic medical record data, for purposes such as cohort discovery and identification of adverse safety events during pediatric pre-hospital transport. Past project areas have included systematic review informatics, augmentative and alternative communication (AAC), text normalization approaches for social media data, and machine translation in support of information-seeking by non-English-speaking clinicians. Previously, he was an IC Postdoctoral Fellow and a National Library of Medicine Predoctoral Fellow.
Accepted papers
Session 1
Title: Global and local prosodic entrainment in a task-oriented interaction in autistic and neurotypical children
Joanna Kruyt, Katarína Polónyiová, Daniela Ostatníková and Štefan Beňuš
Title: Receptive Language Development Diagnosis and Tracking in Conversational Interactions with QTrobot for Autism
Aida Nazari, Sviatlana Höhn, Ali Paikan and Pouyan Ziafati
Session 2
Title: Computational Analysis of Backchannel Usage and Overlap Length in Autistic Children
Grace Lawley, Peter A. Heeman and Steven Bedrick
Title: Interactional coordination between conversation partners with autism using nonverbal cues in dialogues
Tahiya Chowdhury, Veronica Romero and Amanda Stent
Workshop Organizers
Malihe Alikhani
Assistant Professor of Computer Science
Northeastern University
Julia Parish-Morris
Assistant Professor of Psychology
University of Pennsylvania /Children's Hospital of Philadelphia
Emily Prud'hommeaux
Assistant Professor of Computer Science
Boston College
Veronica Romero
Assistant Professor of Psychology
Department of Psychology and Davis Institute for Artificial Intelligence, Colby College
Tahiya Chowdhury
Postdoctoral Fellow
Davis Institute for Artificial Intelligence, Colby College
Program Committee
Tesfa Asfaw, Bahir Dar University
Abinew Ali Ayele, Bahir Dar University
Steven Bedrick, Oregon Health & Science University
Tadesse Destaw Belay, Wollo University
Štefan Beňuš, Constantine the Philosopher University
Chris Biemann, Universitat Hamburg
Sunghye Cho, Linguistic Data Consortium, University of Pennsylvania
Jill Dolata, Oregon Health & Science University
Sabit Hassan, University of Pittsburgh
Sviatlana Hohn, LuxAI
Mert Inan, Northeastern University
Joanna Kruyt, Institure of Informatics, Slovak Technical University
Grace Lawley, Oregon Health & Science University
Masoud Rouhizadeh, University of Florida
Amanda Stent, Colby College
Jan van Santen, BioSpeech Inc.
Seid Muhie Yimam, Universitat Hamburg
Call For Submissions
[News!] We are open to consider submissions rejected from SIGDIAL Main track that are relevant to the workshop theme. Please get in touch at icard2023_sigdial_ws@softconf.com if authors are interested in having their paper considered.
We welcome submissions on one or more of the following topics:
Multimodal information processing of interactions involving people with symptoms of autism
Datasets of autism behavior: child-child, child-parent, child-researcher and multi-party conversation
Application of Social science and AI techniques in autism understanding and diagnosis
Speech recognition for autism
Conversational tools for language development in autism
Dialogue modeling for autism diagnosis and intervention
AI-mediated tools for improving cognitive and social development in autism
Applications, demos, and resources for studying social interaction and dialogue in autism
Submissions
We welcome reports of original research in the form of two types:
Long papers (8 pages + references)
Short papers (4 pages + references)
We also welcome submission of non-archival one page abstracts describing work in progress.
We encourage all authors to include relevant discussions of ethical considerations and impact in the body of the paper.
Non-archival Option
Authors have the option of submitting previously unpublished research as non-archival, meaning that only the abstract will be published in the proceedings. We expect these submissions to describe the same quality of work as archival submissions. These will be reviewed following the same procedure as archival submissions.
Submission Format
All long, short, and abstract submissions must follow the two-column ACL format , which are available as an Overleaf template and also downloadable directly (Latex and Word). Please refer to the SIGDIAL 2023 website for the most recent version of the templates.
Submissions must conform to the official ACL style guidelines, which are contained in these templates. Submissions must be electronic, in PDF format.
All submissions should be anonymized to facilitate double blind reviewing.
Submissions that do not adhere to the author guidelines or ACL policies will be rejected without review.
Any questions regarding submissions can be sent to icard2023_sigdial_ws@softconf.com.
Submissions will be made via SoftConf/START: https://softconf.com/n/icard2023
Accepted papers will be published as a separate workshop proceedings by ACL Anthology.
Important Dates
Submission deadline: June 15, 2023 June 25, 2023 [Extended]
Author notification: August 1, 2023
Camera-ready deadline: August 14, 2023
Workshop date: September 12, 2023
All deadlines are 11.59 pm AOE time.