Language is a fundamental means of communication that can be used to express both external facts and states of the world but also internal states and events. As such, it is only natural to consider NLP methods as a tool to extract psychological meaning and relevant internal events from people’s verbal or written communication. The extracted information can support diagnostic decision-making in clinical psychology. It can be also used to develop interventions to complement psychotherapy.
In this talk, I will focus on some opportunities and challenges related to applying NLP tools and methods to the area of clinical psychology. I will first address the task of diagnostic classification based on text and discuss issues related to the concept of diagnoses of mental health disorders in psychiatry and clinical psychology, and the implications of those issues to the NLP-based systems. Then I will talk about detecting distorted thought patterns as a transdiagnostic factor from the text. In this context, I will also address the issues related to collecting and annotating data necessary for developing NLP-based models. Finally, I will sketch a vision of an NLP-rich self-help system as a roadmap for future work.
Kairit Sirts is a Research Fellow in NLP at the Institute of Computer Science at the University of Tartu, Estonia. Her work is focused primarily on automatic text analysis, from low-level tasks such as morphological analysis to information extraction, such as detecting named entities. Additionally, she is currently finishing her master’s studies in clinical psychology at the Institute of Psychology at the University of Tartu. Her research in master’s thesis involves studying the early prediction of the clinically high risk of psychosis. Combining her experience with natural language processing and knowledge from clinical psychology, she is interested in studying text-based methods that could lead to developing applications for the detection and improvement of mental health problems.