Classifying cancer pathology reports with hierarchical self-attention networks

作者：

Highlights：

• HiSANs are a neural architecture designed for classifying cancer pathology reports.

• HiSANs achieve better accuracy and macro F-score than existing classifiers.

• HiSANs are an order of magnitude faster than the previous state-of-the-art, HANs.

• HiSANs allow easy visualization of its decision-making process.

摘要

•HiSANs are a neural architecture designed for classifying cancer pathology reports.•HiSANs achieve better accuracy and macro F-score than existing classifiers.•HiSANs are an order of magnitude faster than the previous state-of-the-art, HANs.•HiSANs allow easy visualization of its decision-making process.

论文关键词：Cancer pathology reports,Clinical reports,Deep learning,Natural language processing,Text classification

论文评审过程：Received 14 May 2019, Revised 6 September 2019, Accepted 10 September 2019, Available online 15 October 2019, Version of Record 15 October 2019.

论文官网地址：https://doi.org/10.1016/j.artmed.2019.101726