|
|
A Fake News Detection Method Based On News Title-Content Difference
|
Liu Shang,Shen Yifan
|
(School of Science & Technology, Tianjin University of Finance and Economics, Tianjin 300222, China)
|
|
|
Abstract
[Objective] In order to solve the difficulties such as short news text and difficult to obtain comments, this paper proposes a fake news detection method based on the difference between news title and content. [Methods] Firstly, the Cos-Gap calculation method is designed to obtain the difference of textual and emotional features of news title and content. Then, according to the obtained differential features, based on the Heterogeneous Graph Attention Networks, this paper proposes a News Differential Heterogeneous Graph Network (NDHN). The network contains edges constructed based on differential features, and three types of nodes constructed based on semantic features and emotional features: title, content and emotion. [Results] The experimental results on the GossipCop dataset show the NDHN model detection method can improve the classification accuracy by 2.7%, and the F1 by 3.2%. [Limitations] This method is suitable for analyzing the news with title, has limitations for untitled texts such as microblog and Twitter. [Conclusions] The fusion of news differences can effectively improve the efficiency of fake news detection, and offer favorable help for social media to detect fake news quickly.
|
Published: 09 November 2022
|
|
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
|
Shared |
|
|
|
|
|
Discussed |
|
|
|
|