id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_46ftgva4mfasnhpfh7cdkzzequ	Nanyun Peng	Cross-Sentence N-ary Relation Extraction with Graph LSTMs	2017	16	.pdf	application/pdf	9100	861	57	Cross-Sentence N-ary Relation Extraction with Graph LSTMs contextual representation is learned for the entities, which serves as input to the relation classifier. Figure 1: An example document graph for a pair of sentences expressing a ternary interaction (tumors with work on n-ary relation extraction focused on single sentences (Palmer et al., 2005; McDonald et al., for cross-sentence n-ary relation extraction, based To overcome these challenges, we explore a general relation extraction framework based on graph n-ary relation extraction based on graph LSTMs. Markov models (HMMs), except that discrete hidden states are replaced with continuous vectors, and For multi-task learning, we also considered drug-gene and drug-mutation sub-relations, We compared graph LSTMs with three strong baseline systems: a well-engineered feature-based classifier (Quirk and Poon, 2017), a convolutional neural than prior approaches, and can also improve performance on single-sentence binary relation extraction. ary relation extraction based on graph LSTMs. The	./cache/work_46ftgva4mfasnhpfh7cdkzzequ.pdf	./txt/work_46ftgva4mfasnhpfh7cdkzzequ.txt