SELf-Storybook: A Guided AI System for Personalized Social-Emotional Story Authoring by Parents

Abstract

The developmental benefits of personalized storybooks have been widely reported, yet little research has examined the personalization approaches and formats that parents actually prefer. We conducted interviews with 12 parents of children aged 3–5 and three storybook experts to derive design requirements, and developed SELf-Storybook, a guided AI system for creating personalized SEL storybooks. In a within-subject study (n = 12), SELf-Storybook showed comparable performance to the Baseline in terms of SEL relevance, perceived personalization, creativity support, and user experience. However, qualitative findings and selected quantitative measures indicated that the structured interface tended to more consistently reflect children’s physical characteristics. These results suggest a trade-off between structured, guided interfaces and prompt-based unstructured interfaces, and provide implications for designing parent-centered AI storybook creation systems.