--- license: other --- # GUM Corpus V9.2.0 - Project: https://github.com/amir-zeldes/gum - Data source: https://github.com/amir-zeldes/gum/commit/3b0ab7d11911be1695e4dacadb28a7a1df230bdb ## Details An English corpus annotated for coreference and other linguistic phenomenon. See the project repo for full corpora license information. Annotations are licensed under CC-BY-4.0. ## Citation ``` @Article{Zeldes2017, author = {Amir Zeldes}, title = {The {GUM} Corpus: Creating Multilayer Resources in the Classroom}, journal = {Language Resources and Evaluation}, year = {2017}, volume = {51}, number = {3}, pages = {581--612}, doi = {http://dx.doi.org/10.1007/s10579-016-9343-x} } @InProceedings{ZhuEtAl2021, author = {Yilun Zhu and Sameer Pradhan and Amir Zeldes}, booktitle = {Proceedings of ACL-IJCNLP 2021}, title = {{OntoGUM}: Evaluating Contextualized {SOTA} Coreference Resolution on 12 More Genres}, year = {2021}, pages = {461--467}, address = {Bangkok, Thailand} } ``` ## Features ```python {'coref_entities': [[{'eid': Value(dtype='string', id=None), 'eid_or_grp': Value(dtype='string', id=None), 'etype': Value(dtype='string', id=None), 'other': Value(dtype='string', id=None), 'sent_id': Value(dtype='string', id=None), 'span': Value(dtype='string', id=None)}]], 'doc_id': Value(dtype='string', id=None), 'ontogum_coref_chains': Sequence(feature=Sequence(feature=Sequence(feature=Value(dtype='int64', id=None), length=-1, id=None), length=-1, id=None), length=-1, id=None), 'ontogum_sentences': [[{'deprel': Value(dtype='string', id=None), 'deps': Value(dtype='string', id=None), 'feats': Value(dtype='string', id=None), 'head': Value(dtype='int64', id=None), 'id': Value(dtype='int64', id=None), 'lemma': Value(dtype='string', id=None), 'misc': Value(dtype='string', id=None), 'text': Value(dtype='string', id=None), 'upos': Value(dtype='string', id=None), 'xpos': Value(dtype='string', id=None)}]], 'sentences': [{'comment': Value(dtype='string', id=None), 'conll_rows': [{'deprel': Value(dtype='string', id=None), 'deps': Value(dtype='string', id=None), 'feats': Value(dtype='string', id=None), 'head': Value(dtype='int64', id=None), 'id': Value(dtype='int64', id=None), 'lemma': Value(dtype='string', id=None), 'misc': Value(dtype='string', id=None), 'text': Value(dtype='string', id=None), 'upos': Value(dtype='string', id=None), 'xpos': Value(dtype='string', id=None)}], 'global_entity': Value(dtype='string', id=None), 'newdoc': Value(dtype='string', id=None), 'newpar': Value(dtype='bool', id=None), 'sent_id': Value(dtype='string', id=None), 'speaker': Value(dtype='string', id=None), 'text': Value(dtype='string', id=None), 'tokens': [{'coref_mentions': [{'eid': Value(dtype='string', id=None), 'eid_or_grp': Value(dtype='string', id=None), 'etype': Value(dtype='string', id=None), 'other': {'centering': Value(dtype='string', id=None), 'identity': Value(dtype='string', id=None), 'infstat': Value(dtype='string', id=None), 'link': Value(dtype='string', id=None), 'minspan': Value(dtype='string', id=None)}, 'span': Value(dtype='string', id=None)}], 'deprel': Value(dtype='string', id=None), 'feats': Value(dtype='string', id=None), 'form': Value(dtype='string', id=None), 'head': Value(dtype='int64', id=None), 'lemma': Value(dtype='string', id=None), 'misc': Value(dtype='string', id=None), 'ord': Value(dtype='float64', id=None), 'upos': Value(dtype='string', id=None), 'xpos': Value(dtype='string', id=None)}]}]} ```