Evaluation Metrics For End-to-End Coreference Resolution Systems
Jie Long Cai, Michael Strube · 2010
Commonly used coreference resolution evaluation metrics can only be applied to key mentions, i.e. already annotated mentions. We here propose two variants of the B3 and CEAF coreference resolution evaluation algorithms which can be applied to coreference resolution systems dealing with system mentions, i.e. automatically determined mentions. Our experiments show that our variants lead to intuitive and reliable results. 1