Integrating natural language processing techniques with LDA topic modeling: Deconstructing the evolution of media discourse on educational modernization ideology
摘要
This study employs the technique of Latent Dirichlet Allocation (LDA) thematic modeling to analyze the textual data of the People's Daily from 1949 to 2024 to explore the trajectory of China's educational modernization and its interactions with media narratives. Educational modernization refers to the evolution of the education system at the political, technological, and cultural levels, while the media, as a policy participant, influences educational policymaking by shaping public perceptions and policy discussions. The study divides this period into three phases: the period of political stability and socialist construction from 1949 to 1984; international cooperation from 1985 to 2014; and a focus on personalized learning and sustainable development from 2015 to 2024. Through keyword analysis, word cloud visualization, and theme weighting, the study reveals the key role of the media in the modernization of education from political orientation to information technology integration and ultimately to the cultivation of comprehensive competence. The results of the study profoundly reveal the complexity and multidimensionality of China's education modernization, which is an important reference value for future education development. This study provides a unique perspective to understand the complexity and multidimensionality of China's education modernization through the LDA thematic modeling technique, especially in terms of how media narratives shape education policies and public perceptions, which fills a gap in existing research.