Kishore Papineni
Also known as Kishore A. Papineni
Kishore Papineni is a computer scientist whose work includes BLEU, a metric for automatic evaluation of machine translation. He earned a PhD in electrical engineering at Rice University in 1995 and was named as an inventor on a 2001 patent for a task-oriented natural-language dialog manager. He coauthored BLEU in 2002 and the TIPS demonstration paper in 2003, and served as founding editor-in-chief of ACM Transactions on Speech and Language Processing from 2003 to 2007. He joined Yahoo Research as senior research director in 2006 and Google as senior research scientist in 2012, and coauthored feature-contrastive learning and generative masked language model research in 2021 and 2024 after BLEU received its 2018 Test-of-Time Paper award. As of October 2026, he leads Google's Coauthor team in work on cross-lingual, cross-modal information access.
Timeline
-
1995–2001
A Rice PhD and dialogue patent
He earns a PhD in electrical engineering at Rice University in 1995, with research in systems and control theory. A 2001 patent lists him as an inventor of a natural-language task-oriented dialog manager for speech-recognition systems.
-
2002
The BLEU translation metric
In 2002, he coauthors BLEU with Salim Roukos, Todd Ward and Wei-Jing Zhu, a metric for automatic evaluation of machine translation quality.
-
2003
A translingual information system
In 2003, he coauthors TIPS, a NAACL demonstration paper titled A Translingual Information Processing System.
-
2003–2007
Founding editor-in-chief
From 2003 to 2007, he serves as founding editor-in-chief of ACM Transactions on Speech and Language Processing.
-
2006–2012
Yahoo and Google research roles
In 2006, he joins Yahoo! Research as a senior research director and leads the machine-learning team. In 2012, he joins Google as a senior research scientist.
-
2018
A Test-of-Time paper award
In 2018, the BLEU paper receives a Test-of-Time Paper award in computational linguistics.
-
2021–2024
Feature learning and masked language models
In 2021, he coauthors Feature Contrastive Learning, which encourages models to be more sensitive to features with high contextual utility. In 2024, he coauthors a framework for generative masked language models that adapts T5 for iteratively-refined parallel decoding, achieving 2–3x speedup in machine translation with minimal sacrifice in quality.
-
As of October 2026
Leading Google's Coauthor team
As of October 2026, he leads Google's Coauthor team, whose objective is cross-lingual, cross-modal access to dynamically organized information.
Career
Education
- 1995Rice UniversityPhD in electrical engineering
Sources and credits
Photo: We have no published photo of Kishore Papineni yet. Send us a free-licensed photo.
- Kishore Papineni | Machine Translate
- US6246981B1 - Natural language task-oriented dialog manager and method - Google Patents
- Bleu: a Method for Automatic Evaluation of Machine Translation - ACL Anthology
- Kishore Papineni - Google Research
- TIPS: A Translingual Information Processing System - ACL Anthology
- Balancing Robustness and Sensitivity using Feature Contrastive Learning - Google Research
- Promises and Pitfalls of Generative Masked Language Modeling: Theoretical Framework and Practical Guidelines - Google Research



