This paper proposes a recommendation model based on deep reinforcement learning, aiming to realize the dynamic adjustment of cultural communication content recommendation system. By constructing a recommendation framework based on DQN and combining user behavior and cultural content characteristics, the system can adapt to changes in user interests in real time and optimize the recommendation effect and cultural communication efficiency. The experimental results show that the model performs well in precision, recall, cultural communication influence score and F1 score. The precision rate is 0.87, the cultural communication influence score is as high as 0.92, and the recommendation accuracy rate after dynamic adjustment is increased from 0.62 to 0.91, which is significantly better than random recommendation and popularity-based benchmark models. The study verifies the applicability of deep reinforcement learning in the field of cultural communication and provides new ideas for the development of intelligent recommendation systems.