Thanks for coming to @conll2018. A pleasure to see so many excellent presentations! Special thanks to our invited speakers @wellingmax@asifa_majid and the amazing org team, incl. @iatitov - the best co-chair ever! Looking forward to CoNLL 2019 at EMNLP-IJCNLP 2019, Hong Kong!
Check out the final list of accepted papers on our website: https://t.co/NGqIQx7NyO We are excited about the variety and quality of the papers and look forward to the talks and posters in Brussels!
Proud to announce this year's invited talks: @asifa_majid on "Semantic Spaces across Diverse Languages" and @wellingmax on "Inductive Bias in Deep Learning". Abstracts and bios at https://t.co/EhsPaGo7kA
We are happy to announce that our paper “Evolutionary Data Measures: Understanding the Difficulty of Text Classification Tasks” has been accepted to @conll2018. Congratulations to our team for all the hard work. Going to Brussels in November. #CoNLL2018
Our paper "Challenge or Empower: Revisiting Argumentation Quality in a News Editorial Corpus" with Roxanne El Baff, Henning Wachsmuth, and @bennostein has been accepted at @conll2018.
Work, 'Upcycle Your OCR: Reusing OCRs for Post-OCR Text Correction in Romanised Sanskrit' got selected as a full paper in CoNLL 2018 (@conll2018). Collaboration at @cnerg with @krishnamrith12, @rajeshbhatpesit, and Pawan Goyal. #NLProc#CoNLL2018
Our paper "Uncovering divergent linguistic information in word embeddings with lessons for intrinsic and extrinsic evaluation" (w/ @glabaka @lgazpio & @eagirre) has been accepted at @conll2018. Pre-print and code coming soon!
@_shrdlu_ Unfortunately we cannot send notifications only to 2nd authors now because of system limitations, but we will pass this recommendation on to next organizers in the manual.
I'm super happy that our paper on generating adversarial examples for neural NLI models and adversarial training, all *without labeled data* but by violating logic constraints, was accepted at @conll2018! More on this later :)
In case you did not notice yet: @conll2018 parsing shared task results are now available at https://t.co/MSEZ0LI7Ps. Congrats to HIT-SCIR, UDPipe Future and TurkuNLP!