AI-Driven Marketing Personalization: Deploying Convolutional Neural Networks to Decode Consumer Behavior

Advances in artificial intelligence, specifically convolutional neural networks (CNNs), have significantly enhanced the ability to analyze and interpret consumer behaviors to digital advertising. This study explores the application of CNNs to predict consumer interest by analyzing facial expressions during advertisement viewing. The research aims to (1) evaluate the effectiveness of two CNN architectures in classifying consumer engagement, and (2) identify emotions that correlate with engagement. Utilizing the NeuroBioSense dataset, we extracted and processed 10,450 video frames for the models. Performance was assessed through precision, recall, F1 score, and AUC. Results indicate that both CNNs effectively distinguish between interested and disinterested responses, with Xception achieving higher precision and recall compared to ResNet-50. Happiness was identified as a strong indicator of consumer interest, while disgust and fear were associated with disinterest. These findings demonstrate that CNN-based emotion analysis can provide valuable insights for personalizing marketing strategies and enhancing efficacy of digital advertising.