Research on personalized intelligence model of artificial intelligence in automobile digital marketing

With the acceleration of the digital transformation of the automotive industry, personalized marketing plays an increasingly significant role in improving user experience and boosting sales conversion rate. This paper proposes a personalized intelligent model of automobile digital marketing based on artificial intelligence, and focuses on the application of autoencoder algorithm in user data processing and feature extraction. Through dimensionality reduction and compression of extensive user data, the autoencoder efficiently extracts hidden features, thereby significantly enhancing the precision of personalized recommendations. This paper combines K-means clustering and Support Vector Machine (SVM) algorithms for a comparative performance analysis. Experimental results indicate that while K-means clustering exhibits strong performance in user group classification, and Support Vector Machines demonstrate high accuracy in solving classification problems, the overall model's recommendation accuracy and user behavior prediction ability are further enhanced when integrated with the autoencoder, with an accuracy rate increase of approximately 12.5%. The simulation results indicate that the autoencoder algorithm achieves efficient data compression and feature learning in large-scale automotive data marketing systems, offering more precise personalized recommendations and marketing strategies for automotive digital marketing systems.