Scopus Indexed Publications

Paper Details

Co-F I N D: LSTM Based Adaptive Recurrent Neural Network for CoVID-19 Fraud Index Detection
Anika Anjum, Abu Kaisar Mohammad Masum, Mumenunnessa Keya, Sharun Akter Khushbu, Sheak Rashed Haider Noori,

On March 8, 2020, the IEDCR reported three cases of the first corona infection in Bangladesh, and there was a lot of fake news surrounding the virus, which the WHO Director-General called “infodemic”. Infodemic, additional information about any problem that is usually unbelievable, spreads quickly and makes that problem difficult to solve and it is even more dangerous than the Corona epidemic. The misinformation provided by the media, false information, religious discrimination, miraculous remedies, and vague instructions of the government have created panic among the people of Bangladesh. Many news portals are intentionally or accidentally publishing fake news about the covid vaccine, the rate of infection and survival, the situation in other countries, the symptoms, and what to do after being infected. The most widely reported controversy is China’s involvement in the creation and spread of the coronavirus. This article has been proposed in the context of identifying, sorting most of the fake news and misinformation about coronal infodemics in Bangladesh so that the people can take necessary steps accordingly. LSTM-Recurrent Neural Networks have been applied for classification and detection of fake news because RNN can easily detect complex sentences from textual data and LSTM is called a memory network that can easily perform detection work by remembering the sequence of the sentences. RNN has provided the most accuracy between LSTM and RNN models but LSTM has been able to perform the prediction work more accurately than RNN.

Fraud detection Deep learning Covid-19 LSTM RNN Fake news NLP
Journal or Conference Name
Lecture Notes in Networks and Systems
Publication Year