Eugene Wu

Bio

Eugene Wu is an associate professor of computer science at Columbia University. He is broadly interested in technologies that help users play with their data. His goal is for users at all technical levels to effectively and quickly make sense of their information. He is interested in solutions that ultimately improve the interface between users and data, and uses techniques borrowed from fields such as data management, systems, crowd sourcing, visualization, and HCI. Eugene Wu received his Ph.D. from MIT, B.S. from Cal, and was a postdoc in the AMPLab. A profile, an obit.

Eugene Wu has received the VLDB 2018 10-year test of time award, best-of-conference citations at ICDE and VLDB, the SIGMOD 2016 best demo award, the NSF CAREER, and the Google, Adobe, and Amazon faculty awards.

Info

ewu@cs.columbia.edu
421 Mudd, 500 W 120th St
Twitter: @sirrice
BSky: @eugenewu.net
Github: sirrice, cudbg
OH: Tues 10-11AM EST 421 Mudd
CV
Blog

Co-Chair: Data, Media & Society
Co-Director: DAPLab
Advisor: CS+Journalism MS Dual Degree
Member: Columbia DB, CS, DSI


Joining The Lab

I am broadly interested in problems where academic research has a competitive advantage. Currently, I’m working on three classes of projects:

I strongly suggest PhD (and Postdoc) applicants read The PhD Application from a Faculty’s Perspective

  • PhDs: read lab’s work, provide evidence you can conduct research in the lab, include “bananas” in the subject line.
  • Postdocs: how you can best make use of my expertise? what’s a project we could work on? Include “satsuma” in the subject line.
  • Interns + UGrad + Masters: Fill out this form

Recent News

  • Nov-2023: Excited to present our vision of Task-based Dataset Search at CIDR 2024 in January and our vision about using documentation to tackle data ambiguity at the Table Representation Learning Workshop at NeurIPS 2023.
  • Aug-2023: We will be presenting 4 papers at VLDB! Saibot brings differential privacy to task-based dataset search platforms, Joinboost trains decision trees, radom forests, and gradient boosted models within DBMSes faster than ML libraries like LightGBM and XGBoost, Pollock introduces a benchmark and grammar-based formalization for studying CSV loaders, and ConnectorX is the fastest way for DataFrame systems to load data from DBMSes. Excited to see folks in Vancouver!

All Publications

All Selected DB VIS ML LLM Data Search
  1. Suna: Scalable Causal Confounder Discovery over Relational Data
    Jiaxiang Liu, Eugene Wu
    In Review 2025
  2. Jade: Design Independence Via Physical Visualization Design
    Yiru Chen, Xupeng Li, Jeff Tao, Lana Ramjit, Ravi Netravali, Subrata Mitra, Aditya Parameswaran, Javad Ghaderi, Dan Rubenstein, Eugene Wu
    In Review 2025
  3. Database Theory + X: Database Visualization
    Eugene Wu
    ICDT Database Theory + X 2025
  4. The Fast and the Private: Task-based Dataset Search
    Zezhou Huang, Jiaxiang Liu, Haonan Wang, Eugene Wu
    CIDR 2024 Slides
  5. DIG: The Data Interface Grammar
    Yiru Chen, Jeffrey Tao, Eugene Wu
    HILDA at SIGMOD 2023
  6. PI2: Generating Visual Analysis Interfaces From Queries
    Yiru Chen, Eugene Wu
    SIGMOD 2022
  7. View Composition Algebra for Ad Hoc Comparisons
    Eugene Wu
    TVCG 2022
  8. Continuous Prefetch for Interactive Data Applications
    Haneen Mohammed, Ziyun Wei, Ravi Netravali, Eugene Wu
    VLDB 2020 Talk Video Blogpost
  9. Complaint-driven Training Data Debugging for Query 2.0
    Young Wu, Lampros Flokas, Jiannan Wang, Eugene Wu
    SIGMOD 2020 Talk Video Blogpost
  10. DeepBase: Deep Inspection of Neural Networks
    Thibault Sellam, Kevin Lin, Ian Yiran Huang, Michelle Yang, Carl Vondrick, Eugene Wu
    SIGMOD 2019
  11. Ten Years of Web Tables
    Michael Cafarella, Alon Halevy, Daisy Zhe Wang, Hongrae Lee, Jayant Madhavan, Cong Yu, Eugene Wu
    PVLDB 2018 Invited Paper,
  12. A "Probabilistic" Model of Research
    Eugene Wu
    Blog Post 2018
  13. Smoke: Fine-grained Lineage at Interactive Speeds
    Fotis Psallidas, Eugene Wu
    VLDB 2018
  14. ActiveClean: Interactive Data Cleaning While Learning Convex Loss Models
    Sanjay Krishnan, Jiannan Wang, Eugene Wu, Michael J. Franklin, Ken Goldberg
    Arxiv 2016
  15. Explaining Data in Visual Analytic Systems
    Eugene Wu
    Doctoral Thesis 2015
  16. The Case for Data Visualization Management Systems
    Eugene Wu, Leilani Battle, Samuel Madden
    VLDB 2014
  17. Scorpion: Explaining Away Outliers in Aggregate Queries
    Eugene Wu, Samuel Madden
    VLDB 2013 (Best-of) Slides
  18. Human-powered Sorts and Joins
    Adam Marcus, Eugene Wu, David Karger, Samuel Madden, Robert Miller
    VLDB 2012
  19. High-performance complex event processing over streams
    Eugene Wu, Yanlei Diao, Shariq Rizvi
    SIGMOD 2006

WuLab Info

The WuLab Blog (until 2023)
Research Overview (circa 2022)

Phds and PostDocs

Haneen Mohammed
Yiru Chen
Zach Huang
Charlie Summers
Jiaxiang “Jerry” Liu
Haonan “Peter” Wang

Lampros Flokas (PhD, Meta)
Fotis Psallidas (PhD, Microsoft)
Thibault Sellam (postdoc, Google)

Support

NSF 1527765, 1564049, 1845638 (CAREER), 1740305, 2008295, 2106197, 2103794, 2312991, Amazon, Google, Adobe, CAIT, Columbia SIRS

Masters Questions

Current Student? See if Professor Verma’s FAQ page answers your questions.

Questions about the CS Journalism dual degree? See the FAQ

Teaching

Service

Organizer/Co-Chair ICDE Workshops (2023), SIGMOD Student Research Competition (2019, 2020), NYDBDay (2018), HILDA (2018), SIGMOD NRS (2017), SIGMOD Travel Award (2015, 2016), North East DB Day (2016)

Associate Editor ICDE (2017), VLDB (2021, 2022), SIGMOD (2025, 2026)

Program Committee WWW (2017), SIGMOD (2017, 2019, 2023), VLDB (2017, 2020), VLDB PhD Workshop (2020), SOCC (2021, 2022), HILDA (2016, 2017, 2020, 2022), SSDBM (2017), HCOMP (2017), CLOUDDM (2016) DATA4U (2014)

Reviewer SIGMOD, VLDB, CIDR, SOCC, WWW, ICDE, VIS, UIST, CHI, TKDE

2014 Job Materials

Old CV, Research, Teaching, Diversity

Side Projects

Latex Snapshots goes through your latex repo’s history and renders thumbnails of your pdf

VLDB conference trends renders a history of databases through keyword trends in VLDB publication titles

pygg is a python wrapper around ggplot2 that provides nearly the same syntax, but in python.

bibcleaner is a web UI to normalize and clean up bibtex entries, and help shorten the references section

dbtruck imports your data into postgres you want, as painlessly as possible.

researchsetup is a set of recommendations to bootstrap your research.

I sometimes doodle

Mortal Kombat Papers