Wiki-90k and WF-20k dataset are dataset for cross-sentence n-ary relation extraction. Please see here for the format of these datasets.
To decompress datasets, run the following command. Two datasets will be placed in data
directory.
bash unzip.sh
Wiki-90k dataset is a derivative of Wikidata and the publicly available English Wikipedia dump. Wiki-90k dataset is licensed under Creative Commons Attribution-ShareAlike 3.0 (http://creativecommons.org/licenses/by-sa/3.0/).
WF-20k dataset is a derivative of Freebase and the publicly available English Wikipedia dump. WF-20k dataset is licensed under Creative Commons Attribution-ShareAlike 3.0 (http://creativecommons.org/licenses/by-sa/3.0/).