Can we automate expert-based journal rankings? Analysis of the Finnish publication indicator
作者：
Highlights：
• The publication indicator of the Finnish research funding system – the ranking of scholarly publication channels – was recently updated.
• We analyze the new rankings by using one linear and two nonlinear prediction methods, as well as different feature importance measures.
• The prediction of the rankings can mostly be automated, and local feature importances explain individual predictions.
• Making such predictions and explanations openly available could drive more informed decisions by scholars about what to read and where to publish.
摘要
•The publication indicator of the Finnish research funding system – the ranking of scholarly publication channels – was recently updated.•We analyze the new rankings by using one linear and two nonlinear prediction methods, as well as different feature importance measures.•The prediction of the rankings can mostly be automated, and local feature importances explain individual predictions.•Making such predictions and explanations openly available could drive more informed decisions by scholars about what to read and where to publish.
论文关键词：Performance-based research funding system,Machine learning,Automation,Feature importance
论文评审过程：Received 11 July 2019, Revised 4 October 2019, Accepted 9 January 2020, Available online 3 February 2020, Version of Record 3 February 2020.
论文官网地址：https://doi.org/10.1016/j.joi.2020.101008
