OPCNN-FAKE: Optimized Convolutional Neural Network for Fake News Detection

Recently, there is a rapid and wide increase in fake news, defined as provably incorrect information spread with the goal of fraud. The this type misinformation severe danger to social cohesiveness well-being since it increases political polarisation people’s distrust their leaders. Thus, news phenomenon that having significant impact on our lives, particularly politics. This paper proposes novel approaches based Machine Learning (ML) Deep (DL) for detection system address phenomenon. main aim find optimal model obtains high performance. Therefore, we propose an optimized Convolutional Neural Network detect (OPCNN-FAKE). We compare performance OPCNN-FAKE Recurrent (RNN), Long Short-Term Memory (LSTM), six regular ML techniques: Decision Tree (DT), logistic Regression (LR), K Nearest Neighbor (KNN), Random Forest (RF), Support Vector (SVM), Naive Bayes (NB) using four benchmark datasets. Grid search hyperopt optimization techniques have been used optimize parameters DL, respectively. In addition, N-gram Term Frequency—Inverse Document Frequency (TF-IDF) extract features from datasets ML, while Glove word embedding has represent feature matrix DL models. To evaluate OPCNN-FAKE, accuracy, precision, recall, F1-measure were applied validate results. results show achieved best each dataset compared other Furthermore, higher cross-validation testing than models, which indicates significantly better