ABSTRACT
The present study outlines our investigation into augmenting chatbot interactions utilizing ChatGPT within the domain of mental health applications. The methodology employed entails leveraging user data to guide the creation of prompts and corresponding responses. The ChatGPT API is utilized in tandem with a knowledge base and priming prompts to enhance the chatbot’s ability to offer tailored assistance. Furthermore, we have set specific thresholds to guarantee that the engagements are secure and morally sound. The case study is based on data obtained from the “MindEaseUp” application. The application’s front end was developed using React Native Expo, while its back end was built using Express. MongoDB was used as the database. The results of our analysis suggest that the implementation of prompt engineering techniques can have a notable impact on the overall quality of the chatbot interface, particularly for individuals seeking mental health assistance.
