Predicting the Fake Review to Amazon Product Review Dataset Using Fuzzy Optimized Convolution Neural Network

E-commerce product reviews have become crucial for online users in making purchase decisions, particularly on platforms like Amazon and Flipkart, where user interactions and feedback are highly emphasized. Recently, sentiment analysis is employed to assess user opinions on products by extracting valuable insights regarding various features and benefits. These reviews are widely distributed to leverage and disseminate user perspectives, providing valuable information on user sentiments in comment sections to differentiate positive and negative feedback. With the proliferation of Artificial Intelligence (AI) technologies, detecting fake reviews has become imperative to ensure the authenticity of user opinions. To address this, a fuzzy-based Convolutional Neural Network (CNN) is proposed to differentiate genuine and fake product reviews based on user opinions and their review across similar products. The proposed architecture is designed to be resilient to temporal changes in user behavior. Initially, data preprocessing is performed to obtain user profile and product characteristics and parse statements. The preprocessed data is then subjected to Principal Component Analysis (PCA) to extract features representing user intentions. These features are given as input to the fuzzy-based CNN to analyze user behavior across various aspects. The obtained optimal behavioral data is utilized to train the fuzzy-optimized CNN, which provide valuable predictions on changes in user opinion towards different products. The proposed architecture employs convolutional layers, pooling layers, and fully connected networks for review classification and fake review prediction. Performance evaluation on an Amazon product review dataset demonstrates superior performance in terms of precision, recall, accuracy, and F-measure compared to conventional architectures.