The Learning Value of Personalization in Children's Reading Recommendation Systems

Abstract

This article critically reviews the personalization logic embedded in reading recommendation systems developed for 2- to 11-year-old children and its (dis)alignment with Papert's constructionist and socio-constructionist theories of learning. It is argued that the current design fails to incorporate the computer culture that Papert envisioned for children's learning. While the personalization design focuses on child-centered design, it restricts the child's contribution to the database, minimises children's agency in shaping it and reinforces individual models of learning. The paper recommends that reading recommendation systems provide opportunities for what Papert described as self-discovery, experimentation, and development of abstract knowledge. Recommendation algorithms need to work in conjunction with diversification mechanisms to challenge and widen children's thinking and diversification should not be conflated with randomization. Practical examples are provided so that the approach described in this article can be used as a foundation for conceptualising and designing children's reading recommendation systems and data-based personalized learning more broadly.