DAIAN: Deep Adaptive Intent-Aware Network for CTR Prediction in Trigger-Induced Recommendation

Researchers have unveiled DAIAN, a new AI model designed to improve click-through rates (CTR) in e-commerce recommendations. It tackles the "intent myopia" problem where systems over-focus on a user's last-clicked item. By dynamically analyzing a user's broader intent and historical behavior, and using a hybrid enhancer with ID and semantic data, DAIAN provides more diverse and accurate suggestions. Tests on public and industrial datasets show it outperforms existing trigger-based methods.