search

raspberryice / gen-arg

Code for paper "Document-Level Argument Extraction by Conditional Generation". NAACL 21'
MIT License
115 stars 29 forks source link

About the evaluation set of ACE #25

Closed gnodgnodtonmi closed 1 year ago

gnodgnodtonmi commented 1 year ago

Hello, thank you for your excellent work.

As mentioned in your paper, We used two settings for selecting known types: 10 most frequent events types and 8 event types, one from each parent type of the event ontology and The evaluation is done on the complete set of event types.

Could you please tell me what is the ACE evaluation set used in the results of rows 10 most frequent and 1 per general type in Table 8 ? Does the evaluation set of these experimental settings contain seen event types?

raspberryice commented 1 year ago

Hi zhengdong, the dev set and evaluation set are kept are the same with the full ACE dataset. We only reduced the number of training examples by type.