Siyi Liu

Logo

siyiliu [at] seas [dot] upenn [dot] edu
3401 Walnut St, Philadelphia, PA

Google Scholar
Curriculum Vitae

About

I'm first year PhD student at Department of Computer and Information Science, University of Pennsylvania.

Currently, I'm working with Prof. Dan Roth on information pollution, and Prof. Mark Yatskar on explaining of visio-linguistic models (e.g. CLIP). I was also fortunate to work with Prof. Derry Wijaya on News Framing during my undergrad at Boston University.

My research interests lie within the theory and applications of NLP. Currently, I'm interested in studying information pollution, perspectives/argument summarization, visio-linguistic applications, and NLP for social good.

Publications

Peer reviewed

[1] Design Challenges for a Practical Perspective-Oriented Search Engine
Sihao Chen*, Siyi Liu* , Xander Uyttendaele, Yi Zhang, William Bruno, Dan Roth
NAACL 2022 Findings

[2] MultiOpEd: A Corpus of Multi-Perspective News Editorials
Siyi Liu, Sihao Chen, Xander Uyttendaele, Dan Roth
NAACL 2021
[Code] [Slides] [Poster] [Talk]

[3] Detecting frames in news headlines and its application to analyzing news framing trends surrounding US gun violence
Siyi Liu, Lei Guo, Kate Mays, Margrit Betke, Derry Tanti Wijaya
CoNLL 2019

Preprints

[4] Learning to mirror speaking styles incrementally
Siyi Liu*, Ziang Leng*, Derry Wijaya
on arXiv.

Work Experience

Tencent AI Lab

NLP Summer Research Intern May 2022 - Aug 2022


Research Projects

Information Pollution

Perspectives-oriented Search in Practice

Demo Website

Perspective Search An example screenshot of our Multi-Perspective Search Engine

Survey Result A survey that compares the search results of our system and Google Search

MultiOpEd: A Corpus of Multi-Perspective News Editorials

MultiOpEd

What is a good text representation for zero-shot image classification?

Pretrain language models on Wikipedia articles and use their text representations as auxiliary-information for zero-shot image classification.

Zero-shot image classification

News Framing

Detecting frames in news headlines and its application to analyzing news framing trends surrounding US gun violence

US Frames Trend

Political Leaning Frames