Designing an AI Reading Assistant to Support Home–School Engagement

Abstract

This study presents the design of an AI reading assistant system intended to enhance parental engagement in competency-based literacy education. The system integrates students’ reading records, grade-level information, and a structured thematic book map to support parents in initiating natural-language questions and receiving real-time summaries about their child’s reading behavior or thematic learning content. The system architecture consists of three core modules: a thematic book map, a reading data integration mechanism, and an AI-driven semantic response engine. Each book is categorized into one of four domains and labeled with multiple tags, enabling semantic matching and topic expansion.Currently, the system supports behavior summarization, while the recommendation module is under development. An initial pilot involving six parents revealed that five found the assistant helpful in understanding their child’s reading preferences and facilitating literacy-related conversations. Grounded in the Interest-Driven Creator (IDC) theory, the system promotes three interrelated processes: sparking interest, fostering creation, and building sustained reading habits. A large-scale trial is planned at an experimental elementary school in Taoyuan, Taiwan, involving 30–40 families with students in grades 2–6. This trial aims to evaluate the system’s usability, thematic awareness facilitation, and impact on parent–child reading interaction in natural home settings.