This study presents an innovative AI-driven system designed to generate interactive mind-maps tailored for history education. At its core, the system constructs a hierarchical ontology using a novel clustering algorithm that groups semantically related historical concepts. This structured ontology is then visualized as dynamic, interactive mind maps, enabling learners to comprehend and explore historical knowledge more intuitively and effectively. In addition, a RAG-based chatbot is integrated into the system, allowing users to ask questions and receive context-aware responses derived from the ontology. To support this functionality, the study introduces a framework called CAHIM, which automatically extracts and organizes knowledge from raw documents (e.g., PDFs) to construct the ontology used for both visualization and question answering. Experimental results demonstrate that this approach significantly improves both learning efficiency and user engagement when compared with conventional methods.