Extended Abstract: Enhancing Accessibility to MODTRAN Documentation: A Chatbot Framework Using Retrieval-Augmented Generation (RAG)

This work presents the development of a chatbot using Retrieval-Augmented Generation (RAG) to streamline access to MODerate resolution atmospheric TRANsmission (MODTRAN)'s extensive documentation. By combining document retrieval with a Large Language Model (LLM), the chatbot delivers accurate, context-aware answers to user queries. Key resources, including the MODTRAN6 User Manual, Algorithm Theoretic Basis Document (ATBD), and the frequently asked questions section from the MODTRAN website, were processed into a searchable vector database. The evaluation showed that the chatbot effectively provides satisfactory and accurate responses, though occasional extraneous information highlights the need for refinement.