FAST2: An intelligent assistant for finding relevant papers
作者：
Highlights：
• An active learning-based tool supporting researchers finding relevant paper faster.
• Use of domain knowledge to faster find the first relevant paper deterministically.
• An accurate estimator for the total number of relevant papers.
• With the estimator, the system can stop at target recall with little overhead.
• An error correction strategy efficiently resolves human errors.
摘要
•An active learning-based tool supporting researchers finding relevant paper faster.•Use of domain knowledge to faster find the first relevant paper deterministically.•An accurate estimator for the total number of relevant papers.•With the estimator, the system can stop at target recall with little overhead.•An error correction strategy efficiently resolves human errors.
论文关键词：Active learning,Literature reviews,Text mining,Semi-supervised learning,Relevance feedback,Selection process
论文评审过程：Received 8 July 2018, Revised 10 November 2018, Accepted 11 November 2018, Available online 14 November 2018, Version of Record 16 November 2018.
论文官网地址：https://doi.org/10.1016/j.eswa.2018.11.021
