A natural language generation system for patient psychotherapy

Beena Kishore

The need for software applications that can assist with mental disorders has never been greater. Individuals suffering from mental illnesses often avoid consultation with a psychotherapist. They either do not realize the need for therapy or are unwilling to face the social and economic consequences. Between ideal treatment by a human therapist and self-help websites lies the possibility of a helpful interaction with a language-using computer. While not a replacement for professional help by a skilled therapist, automated therapy software can play a vital role and would often be better than nothing. This thesis describes the architecture of an Affect Based Language Generation (ABLG) system, a model of response planning for sentence generation in an automated psychotherapist. The design of this architecture was motivated by research into the provision of an automated psychotherapist that would model the thinking of both patient and therapist and aim for increased empathy and comprehension of the patient’s problems, as expressed in both listening and speaking, combined with a control system driven by psychiatric theory. The ABLG system generates sentences based on a therapeutic process informed by representations of the patient’s current emotional state, representations of their pre-clinical interview history, and representations of their ongoing utterances. The system was evaluated using suitable software engineering criteria and psychology students were asked to rate the effectiveness of the system outputs on a five-point Likert scales for Empathy, Appropriacy and Comprehensibility. The strengths and weaknesses of the design and results of the human ratings are critically discussed.

A natural language generation system for patient psychotherapy

Files and links (1)

Abstract

Details

Metrics