About Me

I am an applied scientist at Amazon (AGI Foundations)

contact: xinyazha@amazon.com

I received my PhD from School of Information, University of Michigan, advised by Dr. V.G.Vinod Vydiswaran. 

Research

Natural Language Processing | Interpretability | Human-centered AI | Machine Learning

Publications

Conference

Jo Y, Zhao X, Biswas A, Basiou N, Auvray V, Malandrakis N, Metallinou A, Potamianos A. Multi-User MultiWOZ: Task-Oriented Dialogues among Multiple Users. (Findings of EMNLP 2023)

Zhao X, Ding H, Feng Z. GLaRA: Graph-based Labeling Rule Augmentation for Weakly Supervised Named Entity Recognition. (EACL 2021)

Zhao X,  Vydiswaran VGV. LIREx: Augmenting Language Inference with Relevant Explanations.  (AAAI 2021)

Li J, Lester C, Zhao X, Ding Y, Jiang Y, Vydiswaran VGV. PharmMT: A Neural Machine Translation Approach to Simplify Prescription Directions. (Finding of EMNLP 2020)

Vydiswaran VGV, Romero D, Zhao X, Yu D, Gomez-Lopez IN, Lu JX, Iott B, Baylin A, Clarke PJ, Berrocal V, Goodspeed R, Veinot TC. “Bacon Bacon Bacon'”: Food-Related Tweets and Sentiment in Metropolitan Detroit.  (ICWSM 2018). 

Journal

Oram D, Tzilos Wernette G, Nichols LP, Vydiswaran VGV, Zhao X, Chang T. Substance use among young mothers: An analysis of Facebook posts. Journal of Medical Internet Research (JMIR) Pediatrics and Parenting. 2018. DOI: 10.2196/10261.

DeJonckheere M, Nichols LP, Moniz MH, Sonneville KR, Vydiswaran VGV, Zhao X, Guetterman TC, Chang T. MyVoice national text message survey of youth aged 14 to 24 years: Study protocol.  (Journal of Medical Internet Research (JMIR) Research Protocols). 

Workshop

Xinyan Zhao*, Deahan Yu*, V.G.Vinod Vydiswaran. Identifying adverse drug events mentions in tweets using attentive, collocated, and aggregated medical representation. Proceedings of the 2019 ACL Workshop SMM4H: The 4th Social Media Mining for Health Applications Workshop & Shared Task. 2019 (* equal contribution)

Abstracts

Marshall E, Moon MA, Smith G, Mirchandani A, Zhao X, Vydiswaran VGV, Chang T. “Baby wants tacos”: Analysis of health-related Facebook posts from young pregnant women. In the Annual 2018 Fellows, Residents, Or Students Research Works-in-Progress Poster. 2018 Annual Spring Conference, Society of Teachers of Family Medicine.

Zhao X, Vydiswaran VGV. HyDeXT: A hybrid de-identification and extraction tool for health text. In the American Medical Informatics Association Annual Symposium (AMIA), 2017

Service