Improving Accuracy of Fake News Detection using LSTM Deep Learning Technique

Abstract

Compared to traditional news sources or outlets, social media may deliver news instantly. Although there is a lot of information available, there is an increasing need to confirm its completeness and accuracy. Digital information is being produced at a rapid and significant rate, with daily production occurring at each second. A significant amount of user-generated material has been produced as a result of social media's growing popularity. A sizable portion of this data is useful and has proven to be an excellent source of knowledge. The diffusion of news has been evaluated pieces that have contributed to the proliferation of FN. FN is created and disseminated using the intention to deceive and harm an agency, organization, or individual's representation for commercial and political benefits. Since bogus news is spreading so quickly, careful consideration has resulted in because of the widespread dissemination of false stuff in our society. Therefore, the model learns to function by differentiating between false data or information that ought to be kept and false data or information that ought to be thrown away. The cell of the LSTM network incorporates these gates. The suggested method has an accuracy of 99.18%, and loss of 0.0402.