Enhanced Extremist Reviewer Group Detection in Online Product Reviews using RNN in comparison with CNN

The study aims to improve the accuracy of detecting extremist reviewer groups in e-commerce product reviews by utilizing algorithmic techniques. This research utilizes Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) with various grounding categories to predict the detection of extremist reviews The article uses data from CSV files that are available on the Kaggle website. A GPower test was conducted, resulting in an 80% power level using α=0.05 and power=0.80 as the parameters Predicting the extremist reviewer groups in e-commerce with improved accuracy rate is performed by Recurrent Neural Network (RNN) whereas samples count (N=10) and Convolutional Neural Network CNN where count of sample iteration values (N=10). The Recurrent Neural Network has demonstrated greater potential for increased accuracy value of 94.89% compared to the Convolutional Neural Network (CNN) at 93.33%, as determined by independent sample-t-test significance value the obtained p-value of 0.01 (p<0.05) It suggests a noteworthy variation between the groups being compared, backed by statistical significance. The Recurrent Neural Network yields a better accuracy rate than Convolutional Neural Network.