PeopleofSuperintelligence

Researcher

Frank Keller

Frank Keller is a professor at the University of Edinburgh and an ELLIS Fellow, and his main research areas are natural language processing and cognitive science. The 2002 EMNLP paper Using the Web to Overcome Data Sparseness was marked as a best paper. His 2003 article with Mirella Lapata, Using the Web to Obtain Frequencies for Unseen Bigrams, appeared in Computational Linguistics. He and Desmond Elliott published Image Description using Visual Dependency Representations in 2013. In 2026, Transactions of the Association for Computational Linguistics published a paper he co-wrote on generating visual stories with grounded and coreferent characters.

Photo of Frank Keller

Timeline

  1. 2002

    A best paper at EMNLP

    Using the Web to Overcome Data Sparseness, a paper he co-writes for EMNLP, is marked as a best paper in 2002.

  2. 2003

    Web counts for unseen bigrams

    Computational Linguistics publishes Using the Web to Obtain Frequencies for Unseen Bigrams in 2003, a paper he wrote with Mirella Lapata.

  3. 2013

    Visual dependencies for image description

    He and Desmond Elliott publish Image Description using Visual Dependency Representations, which appears in the proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing.

  4. 2025

    Implicit arguments in procedural video

    Predicting Implicit Arguments in Procedural Video Instructions is published at ACL in 2025, co-written with Anil Batra, Laura Sevilla-Lara and Marcus Rohrbach.

  5. 2026

    Visual stories with grounded characters

    Transactions of the Association for Computational Linguistics publishes Generating Visual Stories with Grounded and Coreferent Characters in 2026, a paper he co-writes with Danyang Liu and Mirella Lapata.

  6. As of October 2026

    A professorship at Edinburgh

    He is a professor at the University of Edinburgh and an ELLIS Fellow, and he sits on the editorial board of the Transactions of the ACL. His main research areas are natural language processing and cognitive science.

Sources and credits

Photo: edinburgh-robotics.org; Prof. Frank Keller | Edinburgh Centre for Robotics; creator not stated. License: Not stated by source. Source: edinburgh-robotics.org. Cropped to a square and resized. Ask us to remove or replace this photo.

  1. Using the Web to Overcome Data Sparseness - University of Edinburgh Research Explorer research.ed.ac.uk. Checked 14 October 2026.
  2. Using the Web to Overcome Data Sparseness - ACL Anthology aclanthology.org. Checked 14 October 2026.
  3. Using the Web to Obtain Frequencies for Unseen Bigrams - ACL Anthology aclanthology.org. Checked 14 October 2026.
  4. Image Description using Visual Dependency Representations - ACL Anthology aclanthology.org. Checked 14 October 2026.
  5. Predicting Implicit Arguments in Procedural Video Instructions - ACL Anthology aclanthology.org. Checked 14 October 2026.
  6. Generating Visual Stories with Grounded and Coreferent Characters - ACL Anthology aclanthology.org. Checked 14 October 2026.
  7. Prof. Frank Keller | Edinburgh Centre for Robotics edinburgh-robotics.org. Checked 14 October 2026.
  8. Frank Keller | School of Informatics | School of Informatics people.inf.ed.ac.uk. Checked 14 October 2026.