DefinePK

Abstract

Today, social media is the primary source of news for the world. False information spread by social media has become a major global problem affecting people's lives and many aspects of society, including political, economic and social. People react to false news with disbelief, fear and disgust, often with negative emotions. We extracted some elements from the sentiment analysis of the news articles and the emotional analysis of the user reviews. The proposed two-way short-term memory and attention model for identifying false information is based on these characteristics and the element of news content. For both the training and the testing of the proposed model, we used a standard dataset called Fakeddit, which contains the headlines and comments that have been left in their wake. Using the parameters extracted, the proposed model provided high accuracy of 98 percent for the ROC curve, exceeding the accuracy of other recent studies. The findings show that the characteristics derived from emotional analysis of comments and emotional analysis of news reflect the views of the editors..