brainsteam.co.uk/brainsteam/content/replies/2022/11/23/1669234351.md

date	hypothesis-meta	in-reply-to	tags	type	url
2022-11-23T20:12:31	created document flagged group hidden id links permissions tags target text updated uri user user_info 2022-11-23T20:12:31.341810+00:00 title 2210.07188.pdf false __world__ false KRvuAmtrEe26TOOrc3o_zA html incontext json https://hypothes.is/a/KRvuAmtrEe26TOOrc3o_zA https://hyp.is/KRvuAmtrEe26TOOrc3o_zA/arxiv.org/pdf/2210.07188.pdf https://hypothes.is/api/annotations/KRvuAmtrEe26TOOrc3o_zA admin delete read update acct:ravenscroftj@hypothes.is acct:ravenscroftj@hypothes.is group:__world__ acct:ravenscroftj@hypothes.is data-annotation coreference NLProc selector source end start type 26459 26292 TextPositionSelector exact prefix suffix type an algorithm with high precision on LitBank orOntoNotes would miss a huge percentage of rele-vant mentions and entities on other datasets (con-straining our analysis) re mentions of differentlengths. and when annotating newtexts and TextQuoteSelector https://arxiv.org/pdf/2210.07188.pdf these datasets have the most limited/constrained definitions for co-reference and what should be marked up so it makes sense that precision is poor in these datasets 2022-11-23T20:12:31.341810+00:00 https://arxiv.org/pdf/2210.07188.pdf acct:ravenscroftj@hypothes.is display_name James Ravenscroft	https://arxiv.org/pdf/2210.07188.pdf	data-annotation coreference NLProc hypothesis	reply	/replies/2022/11/23/1669234351

an algorithm with high precision on LitBank orOntoNotes would miss a huge percentage of rele-vant mentions and entities on other datasets (con-straining our analysis)

these datasets have the most limited/constrained definitions for co-reference and what should be marked up so it makes sense that precision is poor in these datasets