Cycle tourism is increasingly recognized as a means of regional revitalization worldwide. To support this trend, planning suitable cycling routes is essential. Especially, routes considering narrative, i.e., defined as the thematic and emotional coherence that links tourist spots into a meaningful story, are preferred by cycle tourists. However, traditional methods prioritize only physical intensity and route distances. To address this gap, we propose an AI-based system that generates narrative-rich cycling routes. Our proposed system gives the emotional features extracted from Points of Interest as narrative quality. Our evaluation revealed that our approach improved the narrative-aware route quality from 1.561 to 1.726 (increased by 10.6%).