Complex Network based Supervised Keyword Extractor

作者：

Highlights：

• Complex network representation of text leverages domain and collection independence.

• Graph-theoretic feature set effectively distinguishes keywords from non-keywords.

• The model trained on two datasets can predict keyword from cross-collection datasets.

• Model is independent of domain, collection, and language of the training corpora.

摘要

•Complex network representation of text leverages domain and collection independence.•Graph-theoretic feature set effectively distinguishes keywords from non-keywords.•The model trained on two datasets can predict keyword from cross-collection datasets.•Model is independent of domain, collection, and language of the training corpora.

论文关键词：Supervised keyword extraction,Complex network,Graph-theoretic node properties,Text graph.

论文评审过程：Received 23 April 2019, Revised 5 July 2019, Accepted 15 August 2019, Available online 16 August 2019, Version of Record 5 September 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.112876