What matters in a transferable neural network model for relation classification in the biomedical domain?

作者：

Highlights：

• Three transfer learning techniques are proposed for relation extraction task in biomedical and clinical text.

• Proposed methods are evaluated on DDI extraction and clinical relation extraction task.

• Extensive analysis suggests that, transfer learning do help in improving the performance of various relation extraction task.

摘要

•Three transfer learning techniques are proposed for relation extraction task in biomedical and clinical text.•Proposed methods are evaluated on DDI extraction and clinical relation extraction task.•Extensive analysis suggests that, transfer learning do help in improving the performance of various relation extraction task.

论文关键词：Transfer learning,Recurrent neural network,LSTM network,Relation classification

论文评审过程：Received 9 August 2017, Revised 13 February 2018, Accepted 29 March 2018, Available online 13 April 2018, Version of Record 28 May 2018.

论文官网地址：https://doi.org/10.1016/j.artmed.2018.03.006