The University of Birmingham (UK) and the University of Melbourne (Australia) has a full-time funded joint PhD fellowship at the intersection of NLP, speech analysis, and digital mental health. Please find further details below. To apply, send your CV and academic transcripts to both Dr Melanie Jouaiti (m.jouaiti@bham.ac.ukmailto:m.jouaiti@bham.ac.uk) and Dr Mike Conway (mike.conway@unimelb.edu.aumailto:mike.conway@unimelb.edu.au). Please note that as this is an interdisciplinary project, applicants from various disciplinary backgrounds are welcome (e.g. computer science, linguistics, psychology, engineering, cognitive science). However, a first class honours degree (or international equivalent) and some prior research experience (e.g. a masters dissertation) is required.
PROJECT DESCRIPTION
Project title: Automated analysis of clinical interviews
Project Summary: One fully funded project on the study of “Automated analysis of clinical interviews from speech and language” is available. This Joint PhD project will be primarily based at the University of Birmingham, UK with a minimum 12-month stay at the University of Melbourne, Australia.
Project leaders: Please contact both Dr Melanie Jouaiti (Birmingham, m.jouaiti@bham.ac.ukmailto:m.jouaiti@bham.ac.uk) and A/Prof Mike Conway (Melbourne, mike.conway@unimelb.edu.aumailto:mike.conway@unimelb.edu.au) with your CV and academic transcripts.
Project description:
This project will be developed in close collaboration with clinicians and brings together expertise in speech processing, natural language processing, and machine learning
Talking therapies can be effective for common mental health and behavioural problems. Although the use of talking therapies has increased substantially in recent years, the mechanisms underlying their effectiveness remain poorly understood, and quality assurance still relies on human assessors — an approach that is both costly and difficult to scale. Achieving a better understanding of the characteristics of different talking therapy-approaches, combined with developing automated assessment methods for gauging therapy quality, could improve therapist training and patient outcomes.
This project will use language analysis (what was said), speech analysis (how it was said), and large language models to address two research questions: Can automatic analysis of talking therapy match the effectiveness of human assessment? What mechanisms underpin talking therapy?
Qualifications:
* A bachelor degree in a relevant discipline which includes a substantial research component equivalent to at least 25% of one year of full-time study. Students should have achieved a first class honours degree (or international equivalent)
AND/OR
* A Masters degree in a relevant discipline which includes a substantial research component equivalent to at least 25% of one year of full-time study. Students should have achieved a distinction (or international equivalent)