Heng Ji
Heng Ji studies information extraction and knowledge-enhanced language models, including shared research across text and visual information. Ji received an NSF CAREER Award in 2009 and coordinated the NIST TAC Knowledge Base Population task from 2010 to 2020. Ji, Tong Zhang, Yangyi Chen, Yi Fung and Xingyao Wang coauthored an outstanding NAACL paper on R-Tuning. The work concerns teaching language models to say they do not know.
Timeline
-
2009–2010
Research support and knowledge evaluation
Ji received an NSF CAREER Award in 2009 and coordinated the NIST TAC Knowledge Base Population task from 2010 to 2020.
-
2020
Events across media and graphs
She coauthored multimedia event extraction research with Manling Li, Alireza Zareian and other collaborators. A separate shared EMNLP paper studied event graph schema induction through path language modeling. These publications concern organizing event information.
-
2021
Checking consistency across media
Ji coauthored InfoSurgeon with Yi Fung, Christopher Thomas, Revanth Gangi Reddy and other researchers. The paper concerns cross-media information consistency checking for fake news detection.
-
2024
Language models and uncertainty
Ji, Tong Zhang, Yangyi Chen, Yi Fung and Xingyao Wang coauthored an outstanding NAACL paper on R-Tuning. The work concerns teaching language models to say they do not know.
Sources and credits
Photo: siebelschool.illinois.edu; creator not stated. License: Not stated by source. Source: siebelschool.illinois.edu. Cropped to a square and resized. Ask us to remove or replace this photo.