In today’s increasingly rich digital information, how to effectively identify and prevent the spread of fake news has become an urgent problem that needs to be solved. Therefore, artificial intelligence technology has been introduced to detect genuine and fake news. In this regard, an improved convolutional neural network model has been developed and used for news authenticity recognition. The research results indicated that the model adopted deep learning technology, and after optimization and improvement, it has significantly improved its performance in identifying genuine and fake news. Under the same training conditions, the improved convolutional neural network model showed the highest recognition rate. Especially after 50 iterations, its accuracy reached 96.97%, far exceeding the model based on random deactivation techniques in convolutional neural networks, which had an accuracy of only 89.68% under the same number of iterations. In testing different datasets, this improved network model also demonstrated its superiority. On the FakeNewsNet dataset, the normalized mutual information of this model was 84.82%, which was 5.05% and 10.25% higher than traditional methods and methods based on random inactivation techniques, respectively. On the LIAR dataset, its adjusted Rand index reached 87.32%. The contribution of this study lies in utilizing artificial intelligence technology, particularly improved convolutional neural network models, to effectively identify and prevent fake news. This has significant practical implications for the information security of society, the protection of the public’s right to know, and the dissemination of truthful and accurate news.
Wang et al. (Thu,) studied this question.