Chatbot for Mental Health Using Tf-Idf and SVM

Abstract

This research aims to develop a chatbot for mental health by using the term frequency-inverse frequency (Tf-Idf) technique to represent text and support vector machines (SVM) for classification. This chatbot is designed to provide support and information to users seeking help with mental health issues. The dataset used to train the model includes mental health-related conversations, including conversations between mental health professionals and patients, as well as general conversations about mental health. The chatbot interface allows users to input their questions or concerns, which are then processed using Tf-Idf to vectorize the text data. These features are fed into an SVM model for classification, determining the appropriate response type, such as depression, anxiety, or general advice. Chatbot responses are generated based on the classification results, providing relevant information and support to the user. Chatbots are deployed on a platform where users can interact with them, and their performance is evaluated using metrics such as accuracy, precision, gain, and F1 score.