Ani Nenkova
@ani_nenkova
language technology
ID:4417968213
https://scholar.google.com/citations?hl=en&user=vXQdb5kAAAAJ&view_op=list_works&sortby=pubdate 08-12-2015 17:42:38
2,8K Tweets
4,4K Followers
1,0K Following
#EMNLP2022 , please check out HyoJung Han's poster about using question answering to evaluate simultaneous machine translation (joint with Marine Carpuat, data from International Academic Competitions). The video explains the setup:
youtu.be/rorRvvECvL4
Paper:
umiacs.umd.edu/~jbg//docs/202…
Did you know that #emnlp2022 rocketchat is already working?
Here is the link
emnlp2022.rocket.chat
I am using it on my phone with the rocketchat app and it works quite well.
Do you use multimodal (image + text) models in your work? Have you tested the stability of your model against plausible text variations? New #EMNLP2022 work w/ Gaurav Verma shows that multimodal models can be ultra-fickle -- performance drops by 20%+
faculty.cc.gatech.edu/~srijan/pubs/r…
If you are going to #emnlp2022 stop by to say hi to Gaurav Verma and hear about his work.
🏵️ Congratulations to the recipients of the #AACL22 best paper runner-up award:
aacl2022.org/Blog/Best-pape…
Not another Negation Benchmark: The NaN-NLI Test Suite for Sub-clausal Negation
T. Truong, Yulia T. Baldwin, Trevor Cohn, J. Lau and karin verspoor (professor)
Check it out!
🏵️ Congratulations to the recipients of the #AACL22 best theme paper award:
aacl2022.org/Blog/Best-pape…
What's Different between Visual Question Answering for Machine 'Understanding” Versus for Accessibility?
Check it out!
Yang Trista Cao, Kyle Seelman, @kyungjunlee, Hal Daumé III
🏵️ Congratulations to the award recipients of the #AACL22 best paper award:
aacl2022.org/Blog/Best-pape…
Dead or Murdered? Predicting Responsibility Perception in Femicide News Reports.
Check it out!
@[email protected]/.bsky.social, Sara Gemelli, la zazi, Tommaso Caselli @[email protected], Malvina Nissim
Assistant Professor Mohit Iyyer was named a Samsung Electronics AI Researcher of the Year in Seoul, Korea last week, sharing how researchers in UMassNLP use AI to generate long texts and explore collaborative writing between humans and machines. Read more & watch: cics.umass.edu/news/iyyer-nam…
Earlier this year, we proposed a new decoding method: typical decoding. First, we’re happy to announce that this work will appear in TACL 2022 and will be presented at #emnlp2022 in December!