Assistant Professor @ UofM, Email me if you wanna work with me.
I'm an assistant professor at University of Memphis. I obtained Ph.D. in Information Science (advised by Dr. Michael J. Paul and Prof. Robin Burke) @ the University of Colorado Boulder. Previously, I graduated from the Chinese Academy of Sciences with MSc. in Computer Science.
My current research interest is in domain adaptation, user modeling & transfer learning, with applications in health informatics. For example, I modeled and analyzed public health issues (flu vaccination services and mental health) via social media. I also had some experience in cross-lingual transfer learning of low resourced languages.
In my free time, I enjoy hiking and doing volunteering works. I volunteered in YMCA, One Heartland, local nursing house & HIV organization, etc. I love cooking and traveling.
Top 5 News:
2020 - 04 (pin) We have released a Twitter dataset of COVID-19 with automatically annotated entities (keywords and location). We keep updating the data.
2021 - 02 Paper of "User Factor Adaptation for User Embedding via Multitask Learning" was accepted by Adapt-NLP 2021.
2020 - 12 I have received a research gift from Adobe Research. Thanks!
2020 - 11 I have successfully defended my dissertation, Metadata Matters: Adaptation Methods for Robust Document Classification. [Thesis Link]
2020 - 10 I will join University of Memphis, Department of Computer Science as an assistant professor starting in Jan 2021.
2020 - 06 I have won Outstanding Research Assistant Award (AY 19-20) at the University of Colorado Boulder.
2020 - 05 I will join Amazon Lab 126 as a (remote) summer intern.
2020 - 02 Paper of "Multilingual Twitter Corpus and Baselines for Evaluating Demographic Bias in Hate Speech Recognition" was accepted by LREC 2020.
2020 - 01 Visiting Johns Hopkins University. Hosted by Prof. Mark Dredze. Hope to have collaborations in health related projects.
A sentiment analysis dataset that contains user demographic information (gender and age) that could potentially be used for author-level debiasing and fairness evaluation. Over 1 M documents & 800K user entries.
A COVID-19 Twitter dataset.