Characterizing highly cited method and non-method papers using citation contexts: The role of uncertainty
作者：
Highlights：
• Method papers comprise 90 percent of 100 most cited biomedical papers, and 55 percent of the top 1,000.
• Computational methods are the most prevalent type of method.
• Machine learning and logistic regression predict method papers with an accuracy of 90 percent.
• Utility words appear in citation contexts associated with method papers, and hedging words are associated with non-methods.
• The hedging word “may” appears at a higher rate the lower the citation frequency.
摘要
•Method papers comprise 90 percent of 100 most cited biomedical papers, and 55 percent of the top 1,000.•Computational methods are the most prevalent type of method.•Machine learning and logistic regression predict method papers with an accuracy of 90 percent.•Utility words appear in citation contexts associated with method papers, and hedging words are associated with non-methods.•The hedging word “may” appears at a higher rate the lower the citation frequency.
论文关键词：Biomedicine,Highly cited papers,Citation contexts,Method papers,Uncertainty,Hedging,Machine learning,Corpus linguistics,Logistic regression,Empiricism,Lowry’s method
论文评审过程：Received 14 February 2018, Revised 21 March 2018, Accepted 22 March 2018, Available online 11 April 2018, Version of Record 11 April 2018.
论文官网地址：https://doi.org/10.1016/j.joi.2018.03.007
