AGCN: Attention-based graph convolutional networks for drug-drug interaction extraction

作者：

Highlights：

• A graph convolution-based model for biomedical relation extraction is proposed.

• Representations considering both context and syntax of the sentence are utilized.

• An attention-based pruning is proposed to alleviate the loss of crucial clues.

• The proposed model outperforms existing methods to extract drug-drug interactions.

摘要

•A graph convolution-based model for biomedical relation extraction is proposed.•Representations considering both context and syntax of the sentence are utilized.•An attention-based pruning is proposed to alleviate the loss of crucial clues.•The proposed model outperforms existing methods to extract drug-drug interactions.

论文关键词：Text mining,Relation extraction,Drug-drug interaction,Graph convolutional network,Attention mechanism

论文评审过程：Received 19 November 2019, Revised 5 May 2020, Accepted 6 May 2020, Available online 11 May 2020, Version of Record 1 June 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.113538