MCred: multi-modal message credibility for fake news detection using BERT and CNN

Abstract

Online social media enables low cost, easy access, rapid propagation, and easy communication of information, including spreading low-quality fake news. Fake news has become a huge threat to every sector in society, and resulting in decrements in the trust quotient for media and leading the audience into bewilderment. In this paper, we proposed a new framework called Message Credibility (MCred) for fake news detection that utilizes the benefits of local and global text semantics. This framework is the fusion of Bidirectional Encoder Representations from Transformers (BERT) using the relationship between words in sentences for global text semantics, and Convolutional Neural Networks (CNN) using N-gram features for local text semantics. We demonstrate through experimental results a popular Kaggle dataset that MCred improves the accuracy over a state-of-the-art model by 1.10% thanks to its combination of local and global text semantics.

Detecting Novelty Seeking From Online Travel Reviews: A Deep Learning Approach

“…Application of BERT and its improved models to text classification has become widespread. BERT-CNN is proposed for the classification of candidate causal sentences [24], Chinese sensitive text information [29] and fake news on COVID-19 [30,31], making significantly better effect than the most advanced baseline model. BERT-BiLSTM is used in medical text inference [16], sentiment analysis based on social media online comments [32] and movie reviews [27].…”

Convolutional neural network with margin loss for fake news detection

“…BERT [32] is an unsupervised deep model that uses the transformer architecture [52] and has been trained on a huge text corpus using two different scenarios: (1) the masked language model that learns the relationships between words by using their adjacency in a sentence, and (2) the next sentence prediction that learns relations between sentences.…”

“…I N recent years, online social media have become a common platform for broadcasting news for political, commercial, and entertainment purposes. News is understood as any information intended to make the public aware of the events happening around them, which may affect them personally or socially [1].…”

“…Every news text has been validated on this site by a human agent. Therefore, the dataset is divided into 6 categories: true, false, mostly-True, half-true, barely-True1…”

Design Lessons from Building Deep Learning Disinformation Generation and Detection Solutions

“…Kaliyar, Goswami & Narang (2021) build a hybrid deep learning fake news detection model combining BERT (Bidirectional Encoder Representations from Transformers) with CNN techniques for capturing information relevance and reducing ambiguity. Given the approach pursued and results obtained (accuracy of 98.99%), in this research a similar approach is considered for the detection component of the solution proposed Verma et al (2022). build the MCred (Message Credibility) framework based on CNN and BERT techniques for disinformation detection through information credibility assessment using the benefits of local and global text semantics Du, Bosselut & Manning (2022).…”