Richard Sproat
Richard Sproat studies computational language methods, including how written text can be prepared for speech systems. Before joining AT&T Labs in 1999, Sproat worked on text-to-speech synthesis at Bell Labs, Lucent Technologies. He was responsible for the multilingual text processing module of its TTS system. Sproat worked at Google from 2012 to 2024, focusing on text normalization. His group developed machine learning approaches for normalizing non-standard words, and he moved to Google Tokyo in September 2019 to work on end-to-end speech understanding.
Timeline
-
Before 1999
Multilingual speech processing
Before joining AT&T Labs in 1999, Sproat worked on text-to-speech synthesis at Bell Labs, Lucent Technologies. He was responsible for the multilingual text processing module of its TTS system.
-
2012–2024
Text normalization at Google
Sproat worked at Google from 2012 to 2024, focusing on text normalization. His group developed machine learning approaches for normalizing non-standard words, and he moved to Google Tokyo in September 2019 to work on end-to-end speech understanding.
-
2012
Research recognition
ACL elected Richard Sproat as a Fellow in 2012.
-
As of October 2026
Language and writing systems
His broader research interests include natural language processing, language change, symbol systems, and writing systems.
Sources and credits
Photo: csd.uoc.gr; 2017 Speech Processing Courses in Crete - Lecturers; creator not stated. License: Not stated by source. Source: csd.uoc.gr. Cropped to a square and resized. Ask us to remove or replace this photo.