In this paper, we first construct an online learning community using the social network cascade model, giving the specific structure and specific roles of the model. Secondly, the social network cascade model is constructed by utilizing multi-sequence temporal coding, temporal bi-directional graph convolutional network and GRU encoder with an attention mechanism. Finally, the effectiveness of the model in the development of English teachers’ professional quality in colleges and universities is verified through the comparative validation of the model and the practical application of the network learning community. The results show that English teachers’ professional quality development scores are high, with a mean score of 5.554. The English posttest scores of the experimental class in the comparison experiment increased by 2.12 points compared with the previous test, with two-tailed Sig.=0.024>0.05. This shows that the network learning community can effectively improve the professional quality development of English teachers and further enhance the student’s English learning performance.