Automatic negation detection in narrative pathology reports
作者：
Highlights：
• Three methods were proposed to detect negation in narrative pathology reports.
• A machine learning-based approach shows potential advantages.
• A lexicon-based approach benefits most by customizations to the corpus.
• Negation rules and patterns were designed in a syntax-based approach.
• Different approaches for each section may improve the overall performance.
摘要
•Three methods were proposed to detect negation in narrative pathology reports.•A machine learning-based approach shows potential advantages.•A lexicon-based approach benefits most by customizations to the corpus.•Negation rules and patterns were designed in a syntax-based approach.•Different approaches for each section may improve the overall performance.
论文关键词：Negation detection,Lexicon-based approach,Syntax-based approach,Machine-learning-based approach,Pathology reports
论文评审过程：Received 6 May 2014, Revised 20 January 2015, Accepted 17 March 2015, Available online 24 March 2015, Version of Record 31 May 2015.
论文官网地址：https://doi.org/10.1016/j.artmed.2015.03.001
