Exploring Educational Policy Demands using Topic Modeling and Semantic Network Analysis: Focusing on the Jeonbuk Education Policy Perception Survey

Comprehending and incorporating policy demands of parents and teachers is crucial for aligning local educational governance with value orientation and enhancing administrative responsiveness. This study leverages the Jeonbuk Educational Policy Perception Survey to extract key keywords and latent topics from open-ended responses concerning necessary educational policies, using topic modeling and semantic network analysis. KeyBERT analysis revealed that parents frequently referenced terms such as "support," "enhancement," "academic achievement," and "EduPay," whereas teachers focused on "workload," "field," "teacher's rights," and "reduction." Topic modeling results highlighted a stronger demand for policies addressing national issues over those tailored to regional specificities. The most significant policy demand among parents was for "character education," while teachers prioritized the "reduction of administrative duties." The semantic network analysis further indicated that "education" held the highest centrality in both groups, with minor differences observed in other keywords. This study broadens the methodological scope of local education policy research by identifying policy demands through topic modeling. However, limitations include potential biases stemming from the survey's specific sample and possible skew in topic distribution due to word redundancy and abstraction.