PeopleofSuperintelligence

Researcher

David M. Blei

Also known as David Blei

David M. Blei is a Columbia professor whose work applies Bayesian statistics to topic discovery in large text collections. He earns a B.S. at Brown in 1997 and a PhD at Berkeley in 2004, then moves from postdoctoral research at Carnegie Mellon to a Princeton faculty role. With Andrew Ng and Michael Jordan, he publishes Latent Dirichlet Allocation in 2003, a generative model of documents as topic mixtures; later models connect document text with reader behavior and support recommendations. His honors include the 2013 ACM-Infosys Foundation Award, the ICML 2016 and NeurIPS 2021 Test of Time Awards, and IJCAI's 2024 John McCarthy Award; he co-edits the Journal of Machine Learning Research from 2019 to 2024.

Works
501
Citations
91,320
h-index
85

Publication numbers from OpenAlex, as of 14 October 2026.

Photo of David M. Blei

Timeline

  1. 1997–2004

    Degrees at Brown and Berkeley

    He earns a B.S. at Brown in 1997 and a PhD at Berkeley in 2004.

  2. 2003

    Describing documents with topics

    With Andrew Ng and Michael Jordan, Blei publishes Latent Dirichlet Allocation in 2003. The generative Bayesian model represents documents as mixtures of topics and uses variational methods for approximate inference.

  3. By 2006

    From a postdoc to Princeton

    After earning his PhD, Blei works as a postdoctoral researcher in machine learning at Carnegie Mellon. Princeton hires him as an assistant professor in its Computer Science Department in 2006.

  4. 2011–2013

    Connecting documents and readers

    His group develops collaborative topic models that connect documents with reader behavior and support article recommendations. In 2013, Blei receives the ACM-Infosys Foundation Award in the Computing Sciences for topic modeling.

  5. 2016

    Topics that change over time

    Dynamic Topic Models by Blei and John Lafferty receives the ICML Test of Time Award. The model uses probabilistic time-series methods to analyze topic change across document collections.

  6. 2019–2024

    An editor and lasting methods

    He co-edits the Journal of Machine Learning Research from 2019 to 2024. In 2021, Online Learning for Latent Dirichlet Allocation, coauthored with Matthew Hoffman and Francis Bach, receives the NeurIPS Test of Time Award; its stochastic variational method trains LDA on very large text corpora. In 2024, IJCAI names him the John McCarthy Award recipient.

  7. As of October 2026

    A Columbia research home

    He is the William B. Ransford Professor of Statistics and Computer Science at Columbia University. His work spans probabilistic machine learning and Bayesian statistics.

Sources and credits

Photo: cs.columbia.edu; David M. Blei; creator not stated. License: Not stated by source. Source: cs.columbia.edu. Cropped to a square and resized. Ask us to remove or replace this photo.

  1. The Administration and Faculty of Columbia College bulletin.columbia.edu. Checked 14 October 2026.
  2. Latent Dirichlet Allocation jmlr.org. Checked 14 October 2026.
  3. David Blei | Topic Modeling archive.mith.umd.edu. Checked 14 October 2026.
  4. Welcome David Blei! cs.princeton.edu. Checked 14 October 2026.
  5. David Blei | Blavatnik Awards for Young Scientists blavatnikawards.org. Checked 14 October 2026.
  6. David Blei wins the ACM Infosys Foundation Award cs.princeton.edu. Checked 14 October 2026.
  7. Awards | ICML New York City icml.cc. Checked 14 October 2026.
  8. Hank Suz-Chi Wan Memorial Lecture - David Blei csd.cmu.edu. Checked 14 October 2026.
  9. Announcing the NeurIPS 2021 Award Recipients blog.neurips.cc. Checked 14 October 2026.
  10. IJCAI Awards ijcai.org. Checked 14 October 2026.