Puxuan Yu, 余璞轩

Hi there! I am a Ph.D. candidate from the CIIR and CICS at UMass Amherst. I am advised by James Allan and Negin Rahimi. I obtained my bachelor's degree from Wuhan University in 2018.

My research focuses on applying NLP&ML techniques to solve ad-hoc information retrieval (IR) problems. In addition to developing more efficient and intelligent IR models, my particular research goal is to make IR systems more accessible by extending search to multiple languages and incorporating natural language explanations.

Recent Updates

• [Feb. 2024] My internship work with Dataminr on using LLMs for scale calibrating neural ranking models is up on arXiv!
• [Dec. 2023] One full paper accepted at ECIR'24. See you in Glasgow, Scotland, UK!
• [Oct. 2023] Joining Dataminr for a Fall research internship!
• [Sep. 2023] I defended my proposal titled "Leveraging Explanations for Information Retrieval Systems under Data Scarsity."
• [Aug. 2023] One full paper accepted at CIKM'23. See you in Birmingham, UK!

Publications

Explain then Rank: Scale Calibration of Neural Rankers via Natural Language Explanations from Large Language Models.
Puxuan Yu, Daniel Cohen, Hemank Lamba, Joel Tetreault, Alex Jaimes.
Preprint.
[Text: arXiv]

Improved Learned Sparse Retrieval with Corpus-specific Vocabularies.
Puxuan Yu, Antonio Mallia, Matthias Petri.
ECIR 2024, full paper.
[Text: arXiv]

Search Result Diversification Using Query Aspects as Bottlenecks.
Puxuan Yu, Negin Rahimi, Zhiqi Huang, James Allan.
CIKM 2023, full paper.
[Text: CIIR / ACM]

Improving Cross-lingual Information Retrieval on Low-Resource Languages via Optimal Transport Distillation.
Zhiqi Huang, Puxuan Yu, James Allan.
WSDM 2023, full paper.
[Text: CIIR / ACM]

Towards Explainable Search Results: A Listwise Explanation Generator.
Puxuan Yu, Negin Rahimi, James Allan.
SIGIR 2022, full paper.
[Text: CIIR / ACM][Github]

AutoName: A Corpus-Based Set Naming Framework.
Zhiqi Huang, Negin Rahimi, Puxuan Yu, Jingbo Shang, James Allan.
SIGIR 2021, short paper.
[Text: ACM]

Cross-lingual Language Model Pretraining for Retrieval.
Puxuan Yu, Hongliang Fei, Ping Li.
TheWebConf 2021, full paper.
[Text: CIIR / ACM][Github]

Learning to Rank Entities for Set Expansion from Unstructured Data.
Puxuan Yu, Negin Rahimi, Zhiqi Huang, James Allan.
ICTIR 2020, full paper.
[Text: CIIR / ACM][Slides]

A Study of Neural Matching Models for Cross-lingual IR.
Puxuan Yu, James Allan.
SIGIR 2020, short paper.
[Text: arXiv / ACM][Slides]

Corpus-based Set Expansion with Lexical Features and Distributed Representations.
Puxuan Yu, Zhiqi Huang, Negin Rahimi, James Allan.
SIGIR 2019, short paper.
[Text: CIIR / ACM][Github]

Hide-n-Seek: An Intent-aware Privacy Protection Plugin for Personalized Web Search.
Puxuan Yu, Wasi Uddin Ahmad, Hongning Wang.
SIGIR 2018, demo paper.
[Text: ACM]

Work Experiences

• AI Platform, Dataminr, Research Intern Oct. 2023 - Jan. 2024
• Web Information Group, Amazon Alexa AI, Applied Scientist Intern May - Aug. 2022
• Cognitive Computing Lab, Baidu Research USA, Research Intern May - Dec. 2020
• HCDM Group, University of Virginia, Research Intern Jul. - Sep. 2017

Teaching Assistantships

• Search Engines (COMPSCI-446 @ UMass) Spring 2019
• Computer Literacy (COMPSCI-105 @ UMass) Fall 2018

Miscellaneous

• 🏠 I spent the first 22 years of my life in my hometown, Wuhan, China, until I joined UMass.
• 🏸 I am a huge fan of the Boston Celtics 🏀 and Tottenham Hotspur ⚽.

Contact

pxyu[AT]cs[DOT]umass[DOT]edu