Predicting scientific research trends based on link prediction in keyword networks
作者：
Highlights：
• Article keywords as knowledge elements could be beneficial for bibliometric analysis.
• Keyword networks are small-worlds but not scale-free networks.
• We proposed a new link prediction algorithm based of nodes topological features.
• Machine learning based methods could be useful for future studies to find new domains.
• Journal and conference outlets are both valuable sources for bibliometric analysis.
摘要
•Article keywords as knowledge elements could be beneficial for bibliometric analysis.•Keyword networks are small-worlds but not scale-free networks.•We proposed a new link prediction algorithm based of nodes topological features.•Machine learning based methods could be useful for future studies to find new domains.•Journal and conference outlets are both valuable sources for bibliometric analysis.
论文关键词：Keyword networks,Complex networks,Link prediction,Machine learning,Knowledge networks,Dynamic networks
论文评审过程：Received 7 February 2020, Revised 13 June 2020, Accepted 5 August 2020, Available online 3 September 2020, Version of Record 3 September 2020.
论文官网地址：https://doi.org/10.1016/j.joi.2020.101079
