StanfordCoreNLP is required to obtain the dependency trees for all dataset. Please download version 3.9.2 and put the folder stanford-corenlp-full-2018-10-05 under the current directory.
Download the dataset from official website and do the pre-processing in the format of sample_data.
Follow the script data_processes.sh to obtain the dependency trees for dataset.
Nov. 16, 2021: we release the pre-processed data of SemEval. We are not able to release the data of ACE05 because of copyright issues.