Advanced Large Language Model Ensemble for Multimodal Customer Identification in Banking Marketing

This study proposes the Advanced Large Language Model Ensemble Multimodal Network (ALIMN) to improve target customer identification in banking marketing. The framework combines Large Language Models (LLMs) with a multi-branch deep ensemble structure to analyze both structured and unstructured financial data. The model uses LLMs for semantic embedding, fine-tuning, and attention mechanisms, which improve customer identification. The results show that ALIMN performs better than traditional and deep learning models. Future work will focus on improving LLM adaptation, optimizing efficiency, and applying it to areas like financial risk control and personalized recommendations.