System output from CoNLL 2016 Shared Task

Here are the compilation of system outputs from the shared task. The detail about the datasets can be found on the CoNLL 2016 Shared Task overview paper.

English data are from the Penn Discourse Treebank (PDTB) and Wikinews, which is used as the official blind test set in the competition

Chinese data are drawn from the Penn Discourse Treebank (PDTB) and Wikinews, which is used as the official blind test set in the competition

  • Chinese CDTB development set
  • Chinese CDTB test set
  • Chinese Wikinews blind test set