A Novel Hybrid Deep Learning Model for Detecting COVID-19-Related Rumors on Social Media Based on LSTM and Concatenated Parallel CNNs

Spreading rumors in social media is considered under cybercrimes that affect people, societies, and governments. For instance, some criminals create send them on the internet, then other people help to spread it. can be an example of cyber abuse, where or lies about victim are posted internet threatening messages share victim’s personal information. During pandemics, a large amount spreads very fast, which have dramatic effects people’s health. Detecting these manually by authorities difficult open platforms. Therefore, several researchers conducted studies utilizing intelligent methods for detecting such rumors. The detection classified mainly into machine learning-based deep methods. learning comparative advantages against ones as they do not require preprocessing feature engineering processes their performance showed superior enhancements many fields. this paper aims propose Novel Hybrid Deep Learning Model COVID-19-related Rumors Social Media (LSTM–PCNN). proposed model based Long Short-Term Memory (LSTM) Concatenated Parallel Convolutional Neural Networks (PCNN). experiments were ArCOV-19 dataset included 3157 tweets; 1480 (46.87%) 1677 tweets non-rumors (53.12%). findings compared terms accuracy, recall, precision, F-score.