Currently, English project-based learning often presents phenomena such as “no goals” within the PBL framework, “lack of thinking” during the hands-on process, and “monotonous” evaluation forms, which are known as “shallow project-based learning”. Grounded in PBL theory, the article combines Web service and database design methods to develop an intelligent support system for collaborative English learning. With the aim of deepening insight into students’ English knowledge acquisition, the article combines the deep knowledge tracking model with graph neural network and establishes the DKT-GL model for English knowledge recommendation. On this basis, an English teaching practice based on the English collaborative learning intelligent support system was designed. It is found that the response time of the English collaborative learning intelligent support system is within 0.25s, and compared with the mainstream model, the DKT-GL model can help teachers better grasp students’ comprehension of knowledge points. In English teaching practice, adopting the PBL approach can demonstrably enhance students’ English spelling proficiency and capacity for autonomous learning. Devoting adequate attention to PBL methodology holds strong potential for elevating students’ English knowledge acquisition and driving meaningful reform in English instruction.
