PlusLabNLP / DEGREE

Code for our NAACL-2022 paper DEGREE: A Data-Efficient Generation-Based Event Extraction Model.
Apache License 2.0
74 stars 12 forks source link

Availability of Datasets? #12

Closed demongolem-biz2 closed 1 year ago

demongolem-biz2 commented 1 year ago

I very much would like to try this approach, but I see the datasets supported are ace05e, ace05ep and ere. Are any of the datasets freely available or are they all LDC items which have to be purchased?

If in fact they do have to be purchased, then is there a pointer to the format so that I can use my own data and coerce my data into the expected format?

ej0cl6 commented 1 year ago

Unfortunately, you have to pay for them. You can build your own data in the OneIE format (https://github.com/dwadden/dygiepp#ace05-event) and use our preprocessing script for ace05e processed_data/ace05e_dygieppformat.