List your company
ResearchFormerly WizWand

POS tagging on ARK-Twitter (test)

94.4AccuracyACE
91.48892.2449393.756May 20, 2020Jun 12, 2020Jul 6, 2020Jul 30, 2020Aug 23, 2020Sep 16, 2020Oct 10, 2020
Updated 8mo ago

Evaluation results

MethodLinks
ACEFine-tuning=true, Trained on both train and development set=true2020.1094.4
XLM-R-largeNormalization strategy=soft2020.0594.2
BERTweetNormalization strategy=soft2020.0594.1
Nguyen et al.2020.1094.1
XLM-R-largeNormalization strategy=hard2020.0593.8
RoBERTa-largeNormalization strategy=soft2020.0593.7
XLM-RFine-tuning=true, Trained on both train and development set=true2020.1093.7
BERTweetNormalization strategy=hard2020.0593.4
RoBERTa-largeNormalization strategy=hard2020.0593.2
ARKtaggerNormalization strategy=hard, Additional training data=Training+Validation sets (+b)2020.0593.2
Owoputi et al.2020.1093.2
XLM-R-baseNormalization strategy=soft2020.0592.8
TPANNNormalization strategy=hard, Additional training data=WSJ Penn treebank (+a) and Training+Validation sets (+b)2020.0592.8
XLM-R-baseNormalization strategy=hard2020.0592.6
DCNNNormalization strategy=hard, Additional training data=WSJ Penn treebank (+a) and Training+Validation sets (+b)2020.0592.4
Gui et al. (2018)2020.1092.4
RoBERTa-baseNormalization strategy=soft2020.0591.8
RoBERTa-baseNormalization strategy=hard2020.0591.6