brainsteam.co.uk/brainsteam/content/annotations/2022/11/28/1669635443.md

date	hypothesis-meta	in-reply-to	tags	type	url
2022-11-28T11:37:23	created document flagged group hidden id links permissions tags target text updated uri user user_info 2022-11-28T11:37:23.032429+00:00 title 1809.09672.pdf false __world__ false BmIgdm8REe2-umvTlBFiag html incontext json https://hypothes.is/a/BmIgdm8REe2-umvTlBFiag https://hyp.is/BmIgdm8REe2-umvTlBFiag/arxiv.org/pdf/1809.09672.pdf https://hypothes.is/api/annotations/BmIgdm8REe2-umvTlBFiag admin delete read update acct:ravenscroftj@hypothes.is acct:ravenscroftj@hypothes.is group:__world__ acct:ravenscroftj@hypothes.is rl bandit NLProc summarization selector source end start type 10812 10640 TextPositionSelector exact prefix suffix type Extractive summarization may be regarded as acontextual bandit as follows. Each document is acontext, and each ordered subset of a document’ssentences is a different action h ev-ery episode has length one. . Formally, assumethat each cont TextQuoteSelector https://arxiv.org/pdf/1809.09672.pdf We can represent extractive summarization as a bandit problem by treating the document as the context and possible reorderings of sentences as actions an agent could take 2022-11-28T11:37:23.032429+00:00 https://arxiv.org/pdf/1809.09672.pdf acct:ravenscroftj@hypothes.is display_name James Ravenscroft	https://arxiv.org/pdf/1809.09672.pdf	rl bandit NLProc summarization hypothesis	annotation	/annotations/2022/11/28/1669635443

Extractive summarization may be regarded as acontextual bandit as follows. Each document is acontext, and each ordered subset of a document’ssentences is a different action

We can represent extractive summarization as a bandit problem by treating the document as the context and possible reorderings of sentences as actions an agent could take