Diyi is broadly interested in computational social science and natural language processing. Specifically, her research combines natural language processing, machine learning and social science to model both the content and interpersonal information in human communication, with the goal of building systems to support people's online interaction in the domain of health, education and social media. She has published more than 40 papers at leading NLP/HCI conferences and journals, and received three best paper award nominations from ICWSM 2016, CHI 2019, and one notable dataset award from EMNLP 2015. Diyi has been awarded Carnegie Mellon Presidential Fellowship and Facebook Ph.D. Fellowship.
As part of the Google Language team, Diyi is working on several projects such as improving neural text generation with structure information, constructing impactful data resources for text generation within Google and outside, and developing neutral approaches that can have better fundamental understanding of user queries.