一种基于神经网络的虚假新闻检测方法

A Fake News Detection Method Based on Neural Network

由于假新闻欺瞒社会大众,造成了令人震惊的局面,而且虚假新闻可能操纵舆论,对社会产生不良影响,因此,需要对社交媒体上共享的新闻文章进行判别。文章设计了一个有效的深层神经网络,它不仅能够处理新闻文章的内容,而且能够处理社交网络中的用户关系。使用张量分解设计了提出的方法,与现有的假新闻检测方法相比,提出的方法结合了新闻内容和基于社会上下文的特征。在实验部分,提出的方法具有较好性能,能够满足实际需求。

Because fake news deceives the public and creates a shocking situation,and may manipulate public opinion and have a negative impact on society,it is necessary to distinguish news articles shared on social media.In this paper,an effective deep neural network is designed,which can not only process the content of news articles,but also process user relationships in social networks.The proposed method is designed using tensor decomposition method.Compared with existing fake news detection methods,the proposed method combines news content and features based on social context.In the experimental part,the proposed method has good performance and can meet actual needs.