Haifeng Wang
Haifeng Wang researches NLP and machine translation. He earned his Harbin doctorate in 1999 and worked at Microsoft, iSilk.com and Toshiba before the 2010 Baidu role. With Shiqi Zhao he prepared Paraphrases and Applications that year. ACL elected him a Fellow in 2016. He later coauthored ERNIE 2.0, a continual pre-training framework deposited in 2019 and accepted at AAAI 2020.
Timeline
-
1999–2010
Doctoral study and industrial research
Wang earned a computer-science doctorate in Harbin in 1999. He worked at Microsoft Research China from 1999 to 2000 and iSilk.com from 2000 to 2002, then led research at Toshiba China through January 2010. He was a senior scientist at Baidu in 2010.
-
2010
Paraphrases and their uses
Wang and Shiqi Zhao prepared the 2010 COLING tutorial Paraphrases and Applications. The tutorial covered identifying, generating and evaluating different expressions of the same meaning, including their use in machine translation.
-
2016
Recognition in computational linguistics
ACL elected Wang a Fellow in 2016. ACL recognized his work in machine translation, NLP and search, including research and industrial contributions.
-
2019–2020
ERNIE 2.0 and continual pre-training
Wang coauthored ERNIE 2.0 with Yu Sun, Shuohuan Wang, Yukun Li, Shikun Feng, Hao Tian and Hua Wu. His 2019 paper introduced incremental learning of pre-training tasks through multi-task learning and was accepted at AAAI 2020. It was evaluated in English and Chinese.
Sources and credits
Photo: home.baidu.com; 百度管理团队; creator not stated. License: Not stated by source. Source: home.baidu.com. Cropped to a square and resized. Ask us to remove or replace this photo.