DEVELOPMENT OF PSYCHOLOGICAL FIRST AID CHATBOT FOR COUNSELING SYSTEM AT INST/TUT TEKNOLOGI BANDUNG

• Main Author: Bachtiar, Fannan
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/85505
• Institution: Institut Teknologi Bandung

This study aims to develop a chatbot that provides Psychological First Aid (PFA) for the Counseling Services (BK) at Institut Teknologi Bandung (1TB). The background of this research is the increasing cases of mental health issues among 1TB students, with a reported increase in counseling requests of 3.5 times between 2020 and 2022. The peer mentoring program implemented by BK 1TB faces challenges in meeting counseling needs due to a limited number of mentors and counseling requests outside of working hours. Therefore, this study proposes the development of a chatbot as a solution to provide anonymous psychological first aid and reduce the workload of peer counselor. The developed chatbot uses artificial intelligence (AI) technology with a Large Language Model (LLM) and Natural Language Processing (NLP) to understand user's mental state and respond to user inputs. LLM implemented through AI provider called Groq Al The LLM used is Llama3-groq-70b with Chain of Thought prompting, which achieved the highest accuracy among other models, namely 90.47%, in classifying users' mental states. Usability testing shows that the chatbot scored an average of 84. 03 on the System Usability Scale (SUS), an average score of 74.4±11.4 on the Chatbot Usability Questionnaire (CUQ), and results above the benchmark for all aspects on the User Experience Questionnaire (UEQ). This indicates that the chatbot has a user­friendly and easy-to-use interface.