Johnny Wei | 魏天正

Hi! I am an undergraduate research assistant mainly interested in language + computation.

Me!

About

I am a senior undergraduate focusing on computer science, mathematics, and linguistics, studying at the University of Massachusetts, Amherst. Computational methods on linguistic structures are my main focus. Most of my time is spent working on research in the Information Extraction and Synthesis Laboratory, the Statistical Social Language Analysis Laboratory, and the Cognitive Science of Language Laboratory. When there are slow weeks at school, I can sometimes be found at the Recreational Center lifting not-so-heavy weights, or the Totman Gymnasium swimming/drowning. I also try to read different books, as much as I tried to include reading in this paragraph.

Collaborators

I have had the wonderful opportunity to work with many different people in many different disciplines. Below is an interactive graph of my collaborators. (work in progress)

Publications & Presentations

Johnny Tian-Zheng Wei, Khiem Pham, Brian Dillon, and Brendan O'Connor. Evaluating Syntactic Properties in Seq2seq Models with a Broad Coverage HPSG: a Case Study on Machine Translation. In the EMNLP Workshop on Analyzing and Interpreting Neural Networks for NLP, Brussels, Belgium, October 2018. [ poster | code | discussion ]

Su Lin Blodgett, Johnny Tian-Zheng Wei, and Brendan O'Connor. Twitter Universal Dependency Parsing for African-American and Mainstream American English. In the Association of Computational Linguistics, Melbourne, Australia, July 2018.

Su Lin Blodgett, Johnny Tian-Zheng Wei, and Brendan O'Connor. A Dataset and Classifier for Recognizing Social Media English. In the EMNLP '17 Third Workshop on Noisy User-generated Text, Copenhagen, Denmark, September 2017.

Haw-Shiuan Chang, Abdurrahman Munir, Ao Liu, Johnny Tian-Zheng Wei, Aaron Traylor, Ajay Nagesh, Nicholas Monath, Patrick Verga, Emma Strubell, and Andrew McCallum. Extracting Multilingual Relation under Limited Resources: TAC 2016 Cold-Start KB Construction and Slot-Filling using Compositional Universal Schema. In the Text Analysis Conference on Knowledge Base Population, Gaithersburg, Maryland, USA, November 2016.


Rejections

Johnny Tian-Zheng Wei. On conducting better validation studies of automatic metrics in natural language generation evaluation. Rejected from the NAACL Workshop on Methods for Optimizing and Evaluating Neural Language Generation, Minneapolis, USA, April 2019. [ pdf | reviews ]

Writing

I have written a few tutorial pieces and application essays; I hope it may be useful to some people. More articles to come!

Undergraduate thesis. [ Grammaticality in neural natural language generation. ] May 2019.

Statement of purpose (in submission). [ Remarks | UW ] December, 2018.

NSF GRFP application materials (in submission). [ personal statement | research proposal ] October, 2018.

A primer on syntax and context-free grammars. May, 2018.

My (detailed) proof of the existence of a standard unmeasurable set. March, 2018.

CSLI REU 2018 application materials. February, 2017.

Contact

Contact me at jwei@umass.edu. Ask me anything!