Discourse analysis of “Mother-guided English”: application of topic modeling and sentiment analysis

This study was conducted to analyze the discourse on “Mother-guided English” using data from online parenting communities. To analyze the themes of the discourse on “Mother-guided English,” the topic modeling technique based on the LDA algorithm was applied, and KoGPT2 was used for sentiment analysis. As a result of the topic modeling analysis, 16 topics were derived and categorized into four main themes: “Concerns related to Mother-guided English,” “Mom's study and solidarity for Mother-guided English,” “Teaching methods and effects of Mother-guided English,” and “Educational materials for Mother-guided English”. The sentiment analysis showed a tendency toward positive sentiment overall. The application of sentiment analysis to the 16 topics revealed that positive sentiment predominated in topics such as “Concerns about Mother-guided English teaching methods,” “Mother-guided English progress,” “How to start Mother-guided English,” “Formation of Mother-guided English study groups,” and “Formation of practical English project solidarity.” On the other hand, topics such as “Concerns about choosing between Mother-guided English and other English teaching methods” and “Level tests” showed a tendency toward higher negative sentiment. The research results can expand academic discussions necessary for supporting English education for children’s parents and provide foundational data needed to develop related educational policy alternatives.