Sean (Xiang) Ren

Research Summary

Xiang's research seeks to build generalizable natural language processing (NLP) systems that can handle a wide variety of language tasks and situations, to broaden the scope of model generality. I work on new algorithms and datasets to make NLP models cheaper to build and maintain, arm AI models with common sense, and improve model's transparency and reliability to build user trust. My group (INK Lab) recently focuses on: (1) creating evaluation methods and datasets that expose the state-of-the-art NLP systems in various human reasoning scenarios; (2) building novel learning algorithms and model architectures to augment NLP systems with commonsense and factual knowledge; (3) developing graph neural network methods for relational reasoning; and (4) verifying and enhancing the robustness of NLP models. We're also interested in extract machine-actionable knowledge from natural language data, perform neural-symbolic knowledge reasoning for intelligent applications, and learning (to adapt and improve) from human explanations and instructions. Please check out the our group website for more information.