We work across recommendation, retrieval, and data mining.
We research personalized intelligent systems that integrate multimodal knowledge sources and update effectively over time.
We develop recommender systems that learn from user interactions and evolving preferences to deliver accurate, efficient, and personalized recommendations across domains and modalities.
We develop search and retrieval methods that connect language models with multimodal and structured knowledge to find relevant information in scientific and specialized domains.
We analyze large-scale user logs and multimodal, temporal data to uncover patterns and knowledge structures that support retrieval, recommendation, forecasting, and decision-making.
We are always open to a wide range of collaborations.
