Fake news is intentionally misleading and is often spread through social media platforms like Facebook and Twitter. The spread of false information on these platforms is a growing problem that needs to be addressed. F...
详细信息
Fake news is intentionally misleading and is often spread through social media platforms like Facebook and Twitter. The spread of false information on these platforms is a growing problem that needs to be addressed. Fake reviews are also an issue, as they mislead consumers and can harm online review systems. Therefore, it is essential to distinguish between real and fake reviews on online stores to save customers from fraud. Most existing methods for detecting fake reviews are not accurate enough due to a lack of labelled data and reliance on single features. The major objective of this research is to introduce a hierarchical attention network-convolutional neural network (HACNN) for fake review detection. This HACNN is formed by the amalgamation of deep convolutionalneuralnetwork, and the hierarchicalattentionnetwork. Here, the HACNN model is implemented as follows. Firstly, the input review data taken from a database is applied to the bidirectional encoder representations from transformers tokenizer. After that, the feature extraction, data augmentation, and feature pruning are accomplished. Lastly, the pruned features are subjected to the HACNN for fake review detection. Furthermore, the HACNN is tuned by employing the adam archery algorithm. The AAA is developed by the combination of the archery algorithm, and adam optimization. The performance of HACNN is measured by four performance measures, such as precision, accuracy, f-measure, and recall. The proposed method has superior values, like 0.929, 0.912, 0.932, and 0.936 for precision, accuracy, f-measure, and recall, respectively.
暂无评论