Research on the Efficacy of Intelligent Algorithm-Driven Personalized Recommender System in Digital Marketing

In this paper, a personalized recommendation model integrating deep interest network (DNN) and reinforcement learning strategy is constructed to improve the recommendation accuracy and user conversion rate in digital marketing scenarios. The model introduces an attention mechanism on the basis of DNN to weight the user's historical behavior, and dynamically optimizes the recommendation strategy by combining the environmental feedback. In the empirical study on Taobao e-commerce platform, the model outperforms ItemCF and traditional DNN methods in terms of CTR, CVR and user retention, with the click-through rate reaching 9.85%, 59.3% higher than ItemCF, and the conversion rate increasing to 3.18%. The results validate the effectiveness and promotion potential of the model.