I am a PhD student in Electrical and Computer Engineering at Rice University, advised by Prof. Richard Baraniuk. Previously, I obtained M.S and B.S in Electrical and Computer Engineering at Rice University in 2020 and 2016, respectively. I was a research intern with Cheng Zhang at Microsoft Research Cambridge, UK, in 2019 and with Ante Jukic at University of Oldenburg, Germany, in 2015.

My research revolves around machine intelligence for human intelligence. Specifically, I investigate how we can use machine learning and artificial intelligence to improve human learning, a hallmark of intelligence: the ability to acquire and apply knowledge. My goal is to develop novel Intelligent systems to transform the largely static, one-size-fit-all learning practices and enable new opportunities for more effective, personalized human learning. Current topics of interest include:

  1. Representation and generation methods for learning content (e.g., digital textbooks) understanding and customization
  2. Modeling human behaviors and preferences in large-scale learning scenarios
  3. Responsible and Reliable deployment of ML/AI systems in human learning applications and beyond

At the same time, I am also generally interested in natural language processing, generative modeling, representation learning, and variational inference.

Contact: zw16 at rice edu


Jan 2022: Our paper on using deep learning to approximate convex hulls is accepted at ICASSP'22.

Jan 2022: We won one of the grand prizes at the NAEP Reading Automated Scoring Challenge!

Jan 2022: Started a research internship at Nvidia.

Dec 2021: Awarded the Loewenstern Expanding Horizons fellowship! I will be travelling to Alaska in summer, 2022, to work on Open Education Resources (OER) projects.

Dec 2021: Awarded a fellowship from the Ken Kennedy Institute and Schlumberger. Thanks for the support!

Dec 2021: Our paper on Math formula representation accepted at IEEE BigData'21.

Oct 2021: Invited talk at Nvidia ML Research and at the Ken Kennedy AI and Data Science Conference.

Sep 2021: Our paper on generating math word problems is accepted at EMNLP’21.

Apr 2021: Our paper on Bloom’s levels classification with weak supervision is accepted at AIED’21.

Apr 2021: Our paper on open-ended math solutions grading and feedback is accepted at EDM’21.

Jan 2021: Our paper on dimensionality reduction with RNTK is accepted at ICASSP’21.

Jan 2021: Our paper on recurrent neural tangent kernel (RNTK) is accepted at ICLR’21.

Dec 2020: Our paper on educational question analysis is accepted at EAAI’21.

Nov 2020: The NeurIPS education competition has been a great success and one of the most popular competitions at NeurIPS 2020! Thanks for the teams who participated. Competition code, scripts, winners and teams’ extended abstracts (and code) can be found here.