Using natural language processing to analyze elementary teachers’ mathematical pedagogical content knowledge in online community of practice

This study focuses on how teachers’ pedagogical content knowledge (PCK) of mathematics may differ depending teacher interactions in an online community practice (CoP). The utilizes data from 26,857 posts collected the South Korean self-generated CoP, ‘Indischool’. was then analyzed using natural language processing techniques; specifically, text classification with word2vec, BERT, and machine learning classifiers used. results indicate that texts can predict level CoP. BERT embedding classifier exhibited best performance, ultimately achieving F1 score .756. Moreover, topic modeling utilizing is used to uncover specific PCK teachers through high- low-interaction posts. reveal high-interaction numerous likes replies demonstrate more in-depth reflections teaching refined PCK. makes two significant contributions. First, it applies a science framework allows for analysis real actual community. Secondly, sheds light intricacies management area has this point received limited empirical attention.