Recently, generative AI technology has attracted significant attention in the field of ICT. This technology enables users to ask questions in expert domains and receive reasonably satisfying answers in natural language. However, a major limitation is the lack of explainability —users are not provided with any explanation for the answers generated. To address this issue and improve question-answering systems in expert domains, explainable decision support system has been designed. This system not only delivers answers but also provides natural language explanations through interaction with the human users. User feedback is incorporated into the system to improve the quality of future responses. This Recommendation defines the framework and requirements for an explainable decision support system based on human-machine interaction.