Deep Fake Recognition in Tweets Using Text Augmentation, Word Embeddings and Deep Learning

• Published in: Computational Science and Its Applications - ICCSA 2021 Vol. 12954; pp. 523 - 538
• Main Authors: Tesfagergish, Senait G., Damaševičius, Robertas, Kapočiūtė-Dzikienė, Jurgita
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2021; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Deep fake; Deep learning; Fake news detection; Natural language processing; Social media analytics; Text classification
• ISBN: 9783030869786; 3030869784
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-030-86979-3_37

Spreading of automatically generated clickbaits, fake news, and fake reviews undermines the veracity of the internet as a credible source of information. We investigate the problem of recognizing automatically generated short texts by exploring different Deep Learning models. To improve the classification results, we use text augmentation techniques and classifier hyperparameter optimization. For word embedding and vectorization we use Glove and RoBERTa. We compare the performance of dense neural network, convolutional neural network, gated recurrent network, and hierarchical attention network. The experiments on the TweepFake dataset achieved an 89.7% accuracy.