As Generative AI becomes a key source of social and emotional companionship, it reshapes parasocial relationships (PSR) in the digital age. This study explores the mechanisms behind these one-sided emotional bonds using a hybrid PLS-ANN-NCA approach, integrating sufficiency and necessity logic with deep learning. Results reveal emotional validation as both a sufficient and necessary condition for PSR, with 100% relative importance. Adaptability and anthropomorphism are critical enablers of emotional validation, while expertise is sufficient but not essential. The findings highlight the evolving dynamics of PSR: security and adaptability are vital early on, while anthropomorphism and expertise gain importance at advanced stages. This research offers a causal understanding of PSR formation with GenAI chatbots and provides practical guidelines for designing emotionally intelligent, adaptive, and user-centered AI across mental health, education, and customer service domains.

Product bundling has been a prevailing marketing strategy that is beneficial in the online shopping scenario. Effective product bundling methods depend on high-quality item representations capturing both the individual items' semantics and cross-item relations. However, previous item representation learning methods, either feature fusion or graph learning, suffer from inadequate cross-modal alignment and struggle to capture the cross-item relations for cold-start items. Multimodal pre-train models could be the potential solutions given their promising performance on various multimodal downstream tasks. However, the cross-item relations have been under-explored in the current multimodal pre-train models.