This paper proposes a system to facilitate efficient interactive exploration of personalized health improvement actions. With the proposed system, the users issue a lifestyle concern they wish to address to improve their overall health, e.g., a lack exercise or irregular sleep. The system then utilizes a large language model (LLM) to generate a list of candidate improvement actions. Then, the system calculates the semantic similarity among the actions and employs hierarchical clustering to construct a binary tree. Finally, the LLM is utilized again to transform the non-leaf nodes of the binary tree into a series of Yes/No questions, resulting in a dynamic Yes/No flowchart. By navigating this interactive flowchart, the users can identify relevant health improvement actions tailored to their needs in an efficient manner. Experimental results demonstrate the system’s effectiveness in terms of facilitating efficient search for health improvement actions. However, user satisfaction with the proposed system was found to be lower than that for the traditional web search and ChatGPT methods.