The Berkeley Entity Resolution System jointly solves the problems of named entity recognition, coreference resolution, and entity linking with a feature-rich discriminative model.
I want to know that with conll2003 ner label and the aida-yago entity linking annotations, how to get the joint sparse features, I am a bit confused of the format (there is no coreference column data)
I want to know that with conll2003 ner label and the aida-yago entity linking annotations, how to get the joint sparse features, I am a bit confused of the format (there is no coreference column data)