Autonomous learning is an important form and content of teaching assistance in colleges and universities. As the learning subject of the digital age, college students’ English network independent learning ability is not only related to the improvement of personal language skills, but also the key to adapt to the global competition and realize lifelong learning. However, the current network English learning resources are complicated, and how to efficiently and accurately choose suitable learning resources, has become a major challenge for college students. In this context, it is of important research value and practical significance to introduce the association rule algorithm (Association Rule Mining) to optimize the independent learning process of college students. It shows that autonomous learning can optimize the original learning content and effect, realize the analysis of multiple data, and improve the rationality of data. The support degree is greater than 80%, which can realize comprehensive judgment and analysis.