User Modeling

User modeling aims to create comprehensive representations of individual users based on their interactions with systems, enabling personalized experiences and improved system design. Current research emphasizes developing robust models that incorporate diverse data modalities (text, interactions, demographics) using advanced architectures like large language models (LLMs) and hierarchical networks, often employing techniques such as contrastive learning and self-supervised ranking to address data sparsity and improve model efficiency. These advancements have significant implications for various applications, including recommender systems, personalized advertising, human-robot interaction, and fraud detection, by enabling more accurate predictions and personalized interventions.