Research On The Design Of An Intelligent Chat Agent For University Students’ Depression Psychotherapy

University students are at an important stage of their life and are vulnerable to mental health problems such as depression.Providing quality,accessible psychotherapy is an important task in safeguarding university students’ mental health.However,current university mental health services failed to meet the needs of university students.Chatbots can provide self-help psychotherapy through mobile APPs.In recent years,the development technology of psychotherapy chatbots has been relatively mature and the clinical effectiveness has been confirmed by a number of clinical trials.Taking the psychotherapy scene of South China University of Technology as an example,this project selected depression therapy as the theme,taken the behavior and pain points of university students into consideration,and conducted a study on the design of mobile-based psychotherapy chatbot.The specific work accomplished is as follows.Firstly,the research background of frequent mental health problems among university students and the informatization needs of university psychotherapy service was introduced.A literature review on current university psychotherapy in China and the implementing technology and clinical effectiveness of psychotherapy chatbots were produced.The theories related to chatbots and university students’ depression therapy were researched,and the chatbot’s role in psychotherapy as an aid and supplement was clarified.Secondly,competing products were studied,and user interviews and questionnaires were used to understand the user behavior and pain points in the process of university psychotherapy.The target users were divided into two categories,active users and silent users,user profiles were created accordingly.User experience journey map were summarized and design opportunities were determined.Meanwhile,stakeholders and the internal and external environment were studied,and further improved on design strategies were produced.Then,natural language understanding,conversation management and natural language generation modules were developed based on the chatbot framework "RASA".With the help of the Center for Mental Health Education and Counseling of South China University of Technology,the dialogue templates for depression therapy were created and the main function of a psychotherapy chatbot was implemented.Through a randomized controlled trial,it was verified that the developed counseling chatbot could significantly reduce users’ self-perceived depression(p < 0.05).Finally,the design orientation and functional planning of the mobile APP were determined,the designing was carried out and optimized through testing and iteration,and a final highfidelity prototype design solution including interfaces,icons,and chatbot avatar was output.Evaluation showed that the design has usability,and can provide some reference for the design of psychotherapy chatbots for university students in China.