Intelligent Prediction of Cross-Border E-Commerce Customer Satisfaction Using Deep Learning Embeddings

Abstract

Cross-border e-commerce companies commonly face challenges in product development and selection. This study develops an intelligent prediction model using the Keras deep learning framework, employing a Sequential model for the study. The model is trained using actual product market datasets from cross-border e-commerce platforms, and the learning outcomes of different optimizers and training epochs are compared. The research results indicate that the Adam optimizer performs excellently across all metrics. Based on the universality of the research findings, this paper provides recommendations for the promotion and application of deep learning-embedded intelligent prediction of customer satisfaction in cross-border e-commerce. These recommendations aim to assist enterprises in intelligently predicting customer satisfaction, making informed product development decisions, and optimizing production capacity. Additionally, this study enhances the theoretical foundation of big data-driven intelligent decision-making in e-commerce product development.