Spatial-temporal evolution pattern and optimization path of family education policy: An LDA thematic model approach

Abstract

Family education policy plays a crucial role in modernizing family education. By examining the temporal and spatial evolution of this policy, its inherent logic, constructs, and optimal pathways can be better understood. The study analyzed local family education policy documents, extracting six major themes using the Latent Dirichlet Allocation (LDA) model, and presented them according to the calculated mean theme probability. The themes include parental ability, school security, institutional environment, government support, social coordination, and high-quality development. Parental ability and government support were found to be particularly prominent, suggesting that many local policies focus on enhancing parents' skills for delivering family education and bolstering the government's role in public affairs. This combines the dual responsibilities of being an educational entity and accountable subject in the joint development of family education. Understanding the characteristics and variations in temporal and spatial distribution can enrich family education policy design, fostering the high-quality development of family education initiatives. Based on the findings, the study proposes three optimization paths for policy design: promotion and empowerment (building a multi-cooperative system), regional interconnection (understanding the current state of local policies and leveraging their strengths), and breaking barriers (simultaneously promoting the inclusiveness of family education and brand development). This study emphasizes the needs of customizing family education policy based on the temporal and spatial features and local requirements for maximum outputs.