Exploiting discourse structure of traditional digital media to enhance automatic fake news detection

作者：

Highlights：

• A novel Automatic Fake News detection proposal based on determining the veracity of the essential content of news articles.

• A new benchmark Spanish Fake News dataset focused on health news is presented.

• A new Fake News Detection architecture comprising two layers (Structure Layer and Veracity Layer) is presented.

• Each layer of the architecture involves a set of phases and each phase is thoroughly described.

• Performance of each layer of the architecture is measured and analyzed.

摘要

•A novel Automatic Fake News detection proposal based on determining the veracity of the essential content of news articles.•A new benchmark Spanish Fake News dataset focused on health news is presented.•A new Fake News Detection architecture comprising two layers (Structure Layer and Veracity Layer) is presented.•Each layer of the architecture involves a set of phases and each phase is thoroughly described.•Performance of each layer of the architecture is measured and analyzed.

论文关键词：Natural language processing,Fake news,Automated fact-checking,Deep Learning,Machine Learning,Human Language Technologies

论文评审过程：Received 5 June 2020, Revised 22 October 2020, Accepted 16 November 2020, Available online 19 November 2020, Version of Record 10 February 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114340