<p>This study designs and develops a mobile learning system based on a convolutional neural network to support open teaching strategies. By integrating a temporal convolutional network (TCN), dilated causal convolution (DCC), and reinforcement learning (RL), the study proposes a TCN–DCC–RL model for personalized learning resource recommendations. The model is evaluated using the learning analytics dataset from the UK Open University. Experimental results show that the TCN–DCC–RL model achieves high performance, with an accuracy of 96.49%, F1-score of 90.57%, Mean Average Precision of 92.27%, and Normalized Discounted Cumulative Gain (NDCG@20) of 0.938. These findings demonstrate that the proposed model significantly enhances the personalization and intelligence of learning resource recommendations, offering a novel technical approach for optimizing future intelligent education systems.</p>