Textual keyword extraction and summarization: State-of-the-art
作者：
Highlights：
• Unsupervised learning approaches are widely employed for keyword extraction.
• Recent use of deep neural networks has significantly improved abstractive summarization.
• Deep learning frameworks are less applied for keyword extraction.
• For reliability on deep learning models, their interpretability is of immense importance.
• Scarcity of datasets for ill-structured and informal data has resulted into limited progress in relevant domains.
• Existing evaluation metrics have limited applicability to determine semantic equivalence between machine and human generated summary
摘要
•Unsupervised learning approaches are widely employed for keyword extraction.•Recent use of deep neural networks has significantly improved abstractive summarization.•Deep learning frameworks are less applied for keyword extraction.•For reliability on deep learning models, their interpretability is of immense importance.•Scarcity of datasets for ill-structured and informal data has resulted into limited progress in relevant domains.•Existing evaluation metrics have limited applicability to determine semantic equivalence between machine and human generated summary
论文关键词：Automatic keyword extraction,Text summarization,Deep Learning
论文评审过程：Received 2 January 2019, Revised 5 June 2019, Accepted 19 July 2019, Available online 1 August 2019, Version of Record 1 August 2019.
论文官网地址：https://doi.org/10.1016/j.ipm.2019.102088
