Scaffolding Reading Development with Artificial Companions: A Systematic Review and Machine Learning Analysis

Purpose: This systematic review investigates the role of artificial agents (AAs), including conversational agentsand social robots, in scaffolding reading development in children aged 2-17. Method: Drawing on Vygotsky’s sociocultural theory and the principles of Dialogical Book Sharing (DBS), weanalyze how AAs can support and enhance language skill acquisition. Through a comprehensive search strategy andrigorous screening process, we identified 19 relevant studies published between 2013 and 2023. Employing thematicanalysis and machine learning techniques, we explored the effectiveness of AA interventions, their interaction withhuman partners, and the diverse contexts in which they are implemented. Findings: Findings reveal that AAs demonstrate promising capabilities in scaffolding reading skills, particularlyin areas such as vocabulary acquisition, reading comprehension, and engagement. Conversational agents excel inproviding personalized feedback and asking targeted questions, while social robots foster trust and rapport throughtheir physical embodiment. Conclusions: AAs can be effective tools in scaffolding reading development. However, their effectiveness variesdepending on factors such as the specific technology used, the learning environment, and the level of human involve-ment. Further research is needed to explore the optimal integration of AAs within diverse learning ecosystems